Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Human-building integration based on biometric signal analysis: investigation of the relationships between human comfort and IEQ in a multi-occupancy condition
(USC Thesis Other)
Human-building integration based on biometric signal analysis: investigation of the relationships between human comfort and IEQ in a multi-occupancy condition
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
i
Human-Building Integration Based on Biometric Signal Analysis
Investigation of the Relationships Between Human Comfort and IEQ in a Multi-
Occupancy Condition
By
Jie Liu
Presented to the
FACULTY OF THE
SCHOOL OF ARCHITECTURE
UNIVERSITY OF SOUTHERN CALIFORNIA
In partial fulfillment of the
Requirements of degree
MASTER OF BUILDING SCIENCE
August 2019
ii
ACKNOWLEDGMENTS
I would like to thank all the participants for providing their valuable data for this research. And I also would like to
thank my family for always supporting me throughout the thesis research and my kind friends for being with me and
encouraging me. I would really appreciate Prof. Joon-Ho Choi, Prof. Yolanda Gil, and Prof. Marc Schiler for their
patient instruction and encouragement.
iii
COMMITTEE MEMBERS
Chair
Joon-Ho Choi
Associate Professor
USC School of Architecture
Joonhoch@usc.edu
Second Committee Member
Yolanda Gil
Research Professor
USC School of Engineering
gil@isi.edu
Third Committee Member
Marc Schiler
Professor
USC School of Architecture
marcs@usc.edu
iv
ABSTRACT
The comfort of occupants in office buildings is an important element of work productivity. A higher comfort level of
occupants could lead to lower stress levels. Research on the relationships between an individual occupant’s comfort
and indoor environmental quality (IEQ) in a single workspace is mature. However, studies on the relationships
between the comfort level of a crowd and IEQ in a multioccupancy condition is being reconsidered. As biometric
signals can be processed and represent the stress condition of the human body, it is meaningful to determine the
relationships between IEQ and human comfort by integrating biometric signals. Four different survey groups were
selected based on the building attributes. Data collection consisted of three parts: biometric signal measurement, which
was conducted using wearable sensors; IEQ measurements, which was measured by HOBO sensors; and satisfaction
surveys. All the collection methods were simultaneously performed on-site to ensure that the survey feedback from
occupants could reflect concurrent indoor environment conditions. For data analysis, varying methods of statistical
analysis were adopted. Through these methodologies, the model could identify the relationships between the
parameters, which are IEQ and biometric signals, and significant influencing factors on every collected IEQ parameter:
thermal comfort, visual comfort, indoor air quality, and acoustic comfort.
The research outcomes could provide a method of increasing human comfort in a multioccupancy condition. In
addition, the higher thermal comfort level could increase worker productivity, which is meaningful for future
development of the open office environment.
KEY WORDS: IEQ; Biometric Signals; Human Comfort; Multioccupancy; Statistical Analysis
v
HYPOTHESIS
- Individual biometric data can be associated with specific IEQ elements.
- Biometric crowd analysis helps identify any unpleasant environmental conditions.
RESEARCH GOAL
- Enhance thermal comfort in a multioccupancy condition by collecting and analyzing data through biometric
signals from the crowd.
RESEARCH OBJECTIVES
- Investigate individual physiological characteristics under the same environmental condition.
- Explore the relationship between biometric signals and indoor environmental quality (IEQ) conditions in a
multioccupancy space.
- Identify overall IEQ satisfaction as a function of biometric crowd data using statistical analysis methods.
vi
TABLE OF CONTENTS
ACKNOWLEDGMENTS ........................................................................................................................................... ii
COMMITTEE MEMBERS ....................................................................................................................................... iii
ABSTRACT ................................................................................................................................................................. iv
HYPOTHESIS .............................................................................................................................................................. v
RESEARCH GOAL ..................................................................................................................................................... v
RESEARCH OBJECTIVES ....................................................................................................................................... v
1. INTRODUCTION .................................................................................................................................................... 1
1.1 Problems with Indoor Environmental Quality (IEQ) .................................................................................... 1
1.2 IEQ Research Tool: Postoccupancy Evaluation (POE) ................................................................................. 2
1.3 Importance of Biometric Signals ...................................................................................................................... 2
1.4 Application of Statistics in Building Science ................................................................................................... 3
1.5 Scope of the Thesis ............................................................................................................................................. 3
1.6 Structure of the Thesis ...................................................................................................................................... 3
2. BACKGROUND AND LITERATURE REVIEW ................................................................................................ 4
2.1 Importance of Indoor Environmental Quality ............................................................................................... 4
2.2 POE with IEQ .................................................................................................................................................... 4
2.3 Biometric Signals with IEQ .............................................................................................................................. 5
2.3.1 Heart Rate .................................................................................................................................................... 5
2.3.2 Stress Level .................................................................................................................................................. 6
2.3.3 Electrodermal Activity ................................................................................................................................. 7
2.3.4 Skin Temperature ......................................................................................................................................... 7
2.4 Statistical Analysis ............................................................................................................................................. 8
2.5 Summary ............................................................................................................................................................ 8
3. METHODOLOGY ................................................................................................................................................... 9
3.1 IEQ Measurement ........................................................................................................................................... 11
3.1.1 Sensor Selection ........................................................................................................................................ 11
3.1.2 Sensor Arrangement .................................................................................................................................. 13
3.2 Biometric Signal Measurement ...................................................................................................................... 14
3.2.1 Sensor Selection ........................................................................................................................................ 14
3.2.2 Sensor Arrangement .................................................................................................................................. 15
3.3 Satisfaction and Sensation Survey ................................................................................................................. 18
3.4 Survey Locations and Groups ........................................................................................................................ 19
3.5 Data Analysis ................................................................................................................................................... 23
3.5.1 Correlation ................................................................................................................................................. 24
3.5.2 Stepwise Regression .................................................................................................................................. 25
3.5.3 Analysis Tool: Minitab .............................................................................................................................. 25
4. DATA PREPROCESSING .................................................................................................................................... 26
4.1 Raw Data .......................................................................................................................................................... 26
4.2 Format Conversion and Data Extraction ...................................................................................................... 26
4.3 Data Integration ............................................................................................................................................... 29
4.4 Data Cleaning ................................................................................................................................................... 30
4.5 Summary .......................................................................................................................................................... 31
5. RESULTS AND DISCUSSION ............................................................................................................................. 32
5.1 Demographic Information of Test Subjects .................................................................................................. 32
5.2 Location A: Studio ........................................................................................................................................... 32
5.2.1 Impact by Gender ...................................................................................................................................... 35
5.2.2 Impact by Age ........................................................................................................................................... 37
vii
5.3 Location B: Classroom .................................................................................................................................... 38
5.3.1 Impact by Gender ...................................................................................................................................... 41
5.3.2 Impact by Age ........................................................................................................................................... 42
5.4 Location C: Office Zone C1 ............................................................................................................................ 44
5.4.1 Impact by Gender ...................................................................................................................................... 47
5.4.2 Impact by Age ........................................................................................................................................... 51
5.5 Location C: Office Zone C2 ............................................................................................................................ 54
5.5.1 Impact by Gender ...................................................................................................................................... 57
5.5.2 Impact by Age ........................................................................................................................................... 60
5.6 Comparison of Different Locations ............................................................................................................... 64
6. CONCLUSIONS AND FUTURE WORK ........................................................................................................... 73
6.1 Conclusions ...................................................................................................................................................... 73
6.2 Limitations ....................................................................................................................................................... 73
6.3 Future Work .................................................................................................................................................... 74
6.4 Summary .......................................................................................................................................................... 74
References ................................................................................................................................................................... 75
Appendix A ................................................................................................................................................................. 77
Appendix B .................................................................................................................................................................. 78
Appendix C ................................................................................................................................................................. 79
Appendix D ................................................................................................................................................................. 80
1
1. INTRODUCTION
This study focuses on investigating the methods of analyzing collected data to explore the impact of indoor
environment on building occupants in a multioccupancy condition to enhance their overall thermal comfort. To
investigate how the indoor environment affects occupants’ thermal comfort in a building, postoccupancy evaluation
(POE) and biometric signals of occupants are collected in this study. Also, the researcher established a data-driven
model to analyze and optimize the indoor thermal condition based on the collected data. This chapter introduces the
basic terms involved in this thesis and the importance and current issues of IEQ, POE, biometric signals, and statistical
analyses.
1.1 Problems with Indoor Environmental Quality (IEQ)
According to the United States Environmental Protection Agency (US EPA), people spend approximately 90% of
their time staying indoors, which includes offices, schools, and residential buildings (Shan, Melina, and Yang 2018).
Therefore, it is important to have good IEQ. In their time indoors, they spend much time in office buildings. For
occupants of office buildings, productivity is an important indicator during working hours. Meanwhile, IEQ is
certainly related to the productivity of occupants (Syima Mahbob et al. 2011), so improving and optimizing better
indoor environment quality and a more comfortable working space is significant to occupants. IEQ includes thermal
quality, lighting quality, acoustic quality, and air quality. All these parameters influence occupants’ sensation and
perception, which will definitely affect their productivity. For example, a high-temperature space or excessively bright
lighting condition would distract the occupants’ attention from work. Moreover, providing a work space with better
IEQ could not only enhance the human comfort and productivity but also contribute to the energy saving of the office
buildings. Therefore, for these reasons, it is significantly meaningful to consider and monitor IEQ.
As an important factor influencing the efficiency and health of users, occupants’ comfort has served a valuable purpose
in considering the environment design of the building. Many designers and engineers have already pointed out its
importance. The following four factors should be considered when a building is designed to provide a more
comfortable and better-quality environment: thermal quality, lighting quality, acoustic quality, and indoor air quality.
As one of the factors under IEQ, thermal quality directly affects the comfort of users. However, the same set of criteria
cannot be used in all occupants in the work environment, as thermal comfort is a subjective judgment. And because
of this subjectivity, measuring thermal comfort would be a problem. Since people have already understood the
importance of thermal comfort in architecture design, some industrial organizations put forward their standards, in
which the minimum acceptable values of users are given as a reference, such as ASHRAE. ANSI/ASHRAE standard
55 has been widely used as a guideline in measuring thermal quality.
Visual comfort has a significant impact on the work productivity of occupants, especially when the work is inseparable
from the computer nowadays. The lighting from computers and from office lights both belong to artificial lighting.
Besides artificial lighting, daylighting from outdoors should be considered. Artificial lighting and daylighting are two
main factors when considering visual quality. By making people’s eyes feel comfortable, good lighting quality can let
people focus more on content and have higher efficiency. In contrast, an unduly bright lighting condition would cause
discomfort to the eyes, making it difficult for people to concentrate on their work. People have different satisfaction
levels and perceptions for daylighting and artificial lighting, so it is essential to explore a balance and create a better
working environment.
Acoustic quality is also important for a user’s indoor environment satisfaction. When the acoustic level is controlled
to a comfortable extent, occupants are not easily distracted, and they have higher efficiency. Acoustic quality control
is more important in office buildings, especially in the working environment of multioccupant conditions. Because
people continuously hear various types of noises, such as those from mechanical systems and other people’s
conversations, it is necessary to control noise within a comfortable range. Although acoustic perception is subjective,
overall quality can be improved by controlling some objective acoustic noise sources. For instance, noise from
mechanical systems can be reduced by adopting better sound insulation of walls and ceilings.
2
As a target of IEQ, indoor air quality is closely related to people’s health, making it necessary to monitor and control.
The harm of indoor air environment on the human respiratory system could be reduced by controlling the content of
particulate matter (PM) and total volatile organic compounds (TVOC) in the air. Effective measures include adopting
an HVAC system with high-efficiency filters, enhancing ventilation, using more natural ventilation as possible, and
using nontoxic building decoration materials and furniture. By doing so, there could be better air flow that inhibits the
build-up of toxic substances released by the materials. Besides the effects on health, good indoor air quality can also
increase work productivity. A high concentration of CO2 makes people sleepy and sluggish and causes attention issues,
which probably leads to low productivity, even headaches (Schaffner 2009). Thus, controlling the concentration of
CO2 could also allow people to be more focused and work more efficiently.
1.2 IEQ Research Tool: Postoccupancy Evaluation (POE)
Postoccupancy evaluation (POE) was created to evaluate buildings in a systematic process after it has been built and
occupied for a certain period (Learning from Our Buildings 2002). Since buildings sometimes do not perform as they
are designed, POE has been a widely used measurement to correct problems caused by the building being built out of
design. The POE method is mainly used as an IEQ research tool to better understand occupants’ satisfaction with the
building’s indoor environment. The evaluation usually involves surveys as the main method, such as questionnaires
about occupant satisfaction. Occupants are asked to answer questions about their perception of the indoor environment,
such as lighting condition, acoustic condition, air quality, and thermal condition.
However, this research method has several limitations. Some previous research on IEQ solely relied on survey data.
Because of the subjectivity of the survey data, it will reduce the accuracy of conclusions to conduct a study analysis
based only on the survey. It hardly guarantees that the survey data of every participant could objectively reflect a
comfortable condition, because people’s senses are complicated and could easily be affected by other uncontrollable
factors. Many researchers have noticed and identified this limitation of the POE method. To avoid this, it is essential
to understand the real effects of IEQ on occupants’ perception by integrating with real-time IEQ data.
With recent developments in measuring technology, there are more advanced technologies and methods that could be
adopted such as using sensors to finish on-site IEQ measurements. In a study conducted by Göçer, Karahan, and Oygür,
POE was adopted for each participant with occupant mobility tracking (Göçer et al. 2018). In this case, the subjective
survey data can be analyzed with objective mobility data. Thus, it could be helpful to better understand how the IEQ
condition affects occupants’ satisfaction. In this paper, the author aimed to investigate the relationships between indoor
environment and occupants’ satisfaction. Therefore, a more effective POE was conducted by collecting both user
satisfaction survey data and the objective on-site measurement data on IEQ, which could be collected through specific
IEQ sensors.
1.3 Importance of Biometric Signals
Biometric signals are indicators and parameters of a human body characteristics which could be measured and
recorded by sensors. Different analyses of physiological biometric data would be adopted depending on different
purposes of biometric applications. In recent years, biometric signals have been widely applied in fields other than the
biological field, such as India’s national ID program (“India’s Aadhaar Database to Offer Face Recognition” 2018).
The project was to develop a biometric dataset of each person, such as fingerprint, iris scan, and face photo, that the
government uses to provide retailers in a secure manner. These data are transmitted and authenticated in an encrypted
form through the Internet so that the use of such data will not be restricted by specific locations.
Besides the above application, biometric signals can also be collected to observe some of the physiological
characteristics of the human body and then used as objective measures of it. In a study conducted by Choi and Loftness,
biometric signals were used as an index to thermal sensation. Choi and Loftness selected skin temperature to assess
thermal sensation since the human body regulates skin temperature to adapt to the environment’s (Choi and Loftness
2012). However, the experiment in this study was conducted in a chamber. The sensors need to be stuck on test points
on the human body, which is proper for lab experiments rather than real working or studying conditions. The wired
thermistors would somehow disturb the participants and make them hardly pay attention to the tasks at hand. Thus, to
have data from a more realistic environment condition, it is important to get a more wearable sensor. In a study
conducted in 2016, wearable devices were used to collect users’ physiological and behavioral characteristics so that
3
they could better access digital services (Blasco et al. 2016). A number of benefits are assumed by wearable devices,
including higher portability and comfort compared with traditional devices that can achieve similar functions. In recent
years, wearable devices have been extensively applied in products, and a majority of them have been well recognized
by users, such as the Apple Watch and the Samsung Galaxy Watch. These watches allow users to have a deeper
understanding of their health through continuous monitoring of their physical indexes. Measurements of certain bio-
signals are the focus of this paper, including heart rate, electrodermal activity, and skin temperature. These indexes
are collected with two different wearable devices, which can be simply worn on the wrist like a watch and thus will
ensure users are not affected by these sensors when they are working.
1.4 Application of Statistics in Building Science
As a data analysis method, statistics can be applied in various fields. In the field of building science, the relationships
between occupants’ dissatisfaction with the IEQ of buildings can be identified through statistical analysis so that
targeted improvements of the building can be implemented. For example, in a study conducted in 2016, the researchers
adopted statistical analysis techniques, which are significance analysis and Spearman’s correlation, by using Statistical
Package for the Social Science (SPSS) software (Mallawaarachchi, De Silva, and Rameezdeen 2016). After statistical
analysis, the researchers found seven factors which have a statistically significant correlation with IEQ. Besides this
application, statistics is also applicable to the prediction of individual thermal comfort preferences. Kim et al.
developed a chair with fan and heat strips, through which the user could adjust the temperature based on their cooling
or heating preferences (Kim et al. 2018). The researchers collected the data on people’s behaviors while sitting in the
chair and then built an individual model to predict their individual thermal preferences. Therefore, the role of statistics
in building science is of high significance. Statistical methods are also adopted in this study’s analysis of relationships
among IEQ, occupant satisfaction, and biometric signals.
1.5 Scope of the Thesis
This research pays attention to the influence of IEQ factors including temperature, air quality, lighting, and acoustics
on occupants’ satisfaction in a multioccupancy condition. A POE survey was conducted in three locations in Los
Angeles several times by collecting occupants’ satisfaction data and on-site IEQ measurements. Statistical analysis
was then adopted to explore the relationships between IEQ and user satisfaction. To deeply understand occupants’
satisfaction, human comfort was evaluated using four factors: thermal conditions, air quality, acoustics, and visual
comfort. Moreover, basic information about the participants, such as gender, age, job satisfaction, and health condition,
were also considered.
1.6 Structure of the Thesis
Chapter 1 introduces basic information about IEQ, POE, and biometric signals. It also illustrates the problems and
limitations of previous studies in investigating the relationships between IEQ and occupants’ satisfaction. Chapter 2
discusses the background and literature review. It explores what other scholars did to improve and reduce the
limitations of POE study, the importance of biometric signals to building environment, and how the scholars apply
physiological biometrics to building science. Chapter 3 demonstrates the methodology of this study, which contains
sensor selection, data collection, and data analysis. Chapter 4 shows the results of aggregated data, including the
satisfaction and IEQ curves of each survey group. Chapter 5 illustrates the statistical analysis process by using Minitab.
It discusses the relationships among changes in different occupants’ perception with changes in IEQ data. Finally,
Chapter 6 draws conclusions from data analyzed in Chapter 5 and discusses aspects that need to be improved for future
work.
4
2. BACKGROUND AND LITERATURE REVIEW
Many scholars have already studied the influence of IEQ on occupants through POE or other methods. The goal of
most researches is to optimize IEQ to make people more comfortable. This chapter discussed the previous research
on IEQ optimization based on POE and other possible parameters related with human comfort, including the
integration of physiological parameters (biometric signals) and building science. It also introduces the statistical
analysis methods adopted by the related research.
2.1 Importance of Indoor Environmental Quality
Many previous studies have focused on IEQ, aiming to improve the quality of the indoor environment and making
occupants feel more comfortable during their stay in the building. IEQ is usually composed of several parameters:
thermal conditions, air quality, lighting, and acoustic quality. Each of these parameters will affect the perceptions and
satisfaction of occupants, which is directly related to work productivity.
Mallawaarachchi, De Silva, and Rameezdeen discussed the relationships between IEQ and occupants’ productivity in
green buildings. The authors pointed out that in previous studies, many scholars have found that those who were more
satisfied with their overall working environmental quality were more efficient. Previous studies have mostly focused
on the relationship between a single parameter of IEQ and occupants’ productivity, for example, the relationship
between thermal comfort and occupants’ productivity, as well as the relationship between indoor air quality and
occupants’ satisfactions. In this case, the author decided to explore the relationship between the overall industrial
environmental quality and the occupants’ satisfactions. Among them, the overall IEQ included temperature, humidity,
ventilation, indoor air quality, daylighting and lighting quality, thermal comfort, and access to views. Moreover,
having good IEQ was a very important evaluation index in many green building certification systems, such as
Leadership in Energy and Environmental Design (LEED) and Energy Star. This was because IEQ was considered to
affect the health and well-being of occupants when evaluating green buildings. Therefore, it is meaningful to study
the relationship between identified parameters of IEQ and occupants’ satisfaction (Mallawaarachchi, De Silva, and
Rameezdeen 2016).
Shan, Melina, and Yang studied the influence of the IEQ of educational buildings on occupants. This paper mentioned
that people spent about 90% of their time in buildings, including office buildings, industrial buildings, educational
buildings, and residential buildings. Therefore, IEQ must satisfy the occupants and provide a comfortable environment
and higher work efficiency. In the report Health, Wellbeing and Productivity in Offices, “green” was usually
understood and interpreted as having a low carbon footprint and efficient building energy consumption. However,
buildings with these two characteristics were not necessarily those that can improve the productivity and health of
occupants. Green buildings often place too much emphasis on reducing the impact of buildings on the environment,
but they often neglected the impact of buildings on occupants (Health, Wellbeing and Productivity in Offices: The
Next Chapter for Green Building 2014). Many previous studies focused on the impact of the IEQ of office buildings
on occupants. There were relatively fewer studies on educational buildings. Since the occupants of educational
buildings were mainly students, the impact of IEQ on their well-being and health was also worth studying (Shan,
Melina, and Yang 2018).
2.2 POE with IEQ
POE is a complicated evaluation of a building after it is built and occupied to improve its performance and occupants’
comfort. Many scholars collect POE and IEQ data together and use them to analyze and improve IEQ because
satisfaction can be known from specific IEQ indicators through POE, such as temperature, acoustics, and humidity.
Mallawaarachchi, De Silva, and Rameezdeen chose a green office building in Sri Lanka and used a questionnaire to
collect survey data from 65 occupants. The questions in the questionnaire were closed-ended, and the participants
were asked to answer and evaluate their perception using a five-point scale. At the same time, the authors also
conducted semistructured interviews. However, the interview data were not analyzed separately but were used to
verify the results of survey data analysis through comparison. Through correlation analysis of the survey data, the
authors obtained the seven most significant factors of IEQ in this surveyed green building. Among these seven factors,
5
there was a positive correlation between air quality and acoustical partitioning, and between system control and
occupants’ productivity (Mallawaarachchi, De Silva, and Rameezdeen 2016).
Kim et al. investigated the differences in perception and satisfaction of different genders of occupants on IEQ. The
authors mentioned that according to previous studies, building occupants usually had significantly different feelings
and reactions to the same indoor environment. However, although many scholars have conducted research on IEQ,
few have focused on the influence of gender. In the research on the relationship between indoor environment,
especially indoor air quality, and occupants’ health symptoms, females have been found to have more health issues.
For example, females were more likely to suffer from sick building syndrome (SBS) symptoms than males, such as
fatigue, headache, and dry eyes. Moreover, females were more sensitive to cold or hot conditions than males. The
authors conducted a POE at the Center for the Built Environment (CBE) in the University of California, Berkeley.
CBE was developed as a survey tool based on the web platform, which was used to evaluate occupants’ satisfaction
of seven main indicators of IEQ, as well as overall satisfaction and overall productivity. Through t-test and logistic
regression analysis of all participants, the authors found that female occupants were more likely to complain about
IEQ and feel more dissatisfaction than their male colleagues. However, there were no obvious differences between
males and females regarding overall satisfaction (Kim et al. 2013).
Hwang and Kim studied and analyzed the impact of IEQ factors on the comfort and health of occupants’ through POE
and applied these perception and behavior data from occupants as basic data in building energy system management.
Based on this purpose, the authors carried out a two-year monitoring of the indoor environment and collection of
questionnaire survey data in an office building in Seoul, South Korea. The authors mentioned that POE was a very
valuable tool, which can be used to evaluate whether the performance of the completed building meets the original
design goals as well as what needs to be improved. These are all drawn from the information by collecting data on the
requirements and satisfaction of occupants with the current building’s condition. And these evaluation results could
also be applied to the improvement of existing buildings and new constructions in the future. This study conducted a
POE survey using five aspects: thermal comfort, indoor air quality, acoustics, lighting, and overall IEQ. Overall IEQ
was monitored five times from February 2008 to May 2009, and each measurement lasted 10 days. Questionnaires
were also administered simultaneously with these IEQ data acquisitions. It was presented to each participant in the
form of a survey website. Initially, the questionnaire was distributed to 20 people as a trial. Based on the feedback and
results received, the questionnaire was modified to more effectively evaluate the perception of occupants on IEQ. The
final data was analyzed through SPSS 12.0 (Hwang and Kim 2013).
In previous studies, many scholars conducted research on improving and optimizing IEQ according to POE, such as
the studies mentioned above. However, although POE could be adopted as a tool to improve indoor environment, it is
not comprehensive and accurate enough. Since indoor environment and human comfort are dynamic, it is not
convincing to analyze the satisfaction of occupants with indoor environment based only on survey data. Thus, it is
necessary to analyze environmental satisfaction by exploring other relationships between IEQ satisfaction and the
human body, such as bio-signals.
2.3 Biometric Signals with IEQ
As indexes that can objectively reflect physiological changes in the human body, skin temperature, heart rate, and
electrodermal activity (EDA) reflect these changes in different principles. When investigating people’s satisfaction
with IEQ, their body states may be different depending on different degrees of satisfaction. Therefore, it is of great
research value to study the relationships between IEQ and bio-signals.
2.3.1 Heart Rate
Heart rate refers to the beating speed of the heart measured via beats per minute. Heart rates vary with different
influences and stimulations from the external environment. In other words, heart rate can also be used as a reference
value to measure the extent of the impact of the external environment on the human body. If the external environment
has a greater impact and stimulation on the human body, the heart rate will change. If the influence of the external
environment on the human body is small, the heart rate will be in a stable range. This is a normal fluctuation in the
response of the cardiac muscle to the body. Therefore, when the heart rate becomes too high or is in a constantly high
6
level, it can be explained that some issues have occurred. One of the situations that can cause this phenomenon is
stress, which is a very common situation in one’s daily life. However, it has a certain subjectivity, that is, the same
thing that may make one person feel stressful will have the same effect on others. This is because everyone has
different appreciations of the environment and other things, so they will have different stress reactions. When a
stimulus passes, the heart rate will drop again, and the body will return to a resting state. However, when the body is
in one or more stimuli for a long time, it will be harmful to one’s health because the body secretes more hormones to
speed up the work of the cardiac muscle. Over time, this can lead to heart failure or heart disease. Therefore, it is
significantly necessary for human health to control and reduce the stimulation brought about by the external
environment as much as possible.
Nayak et al. measured three different human body factors and collected thermal survey data at different indoor
temperatures and then analyzed which of the three could best correlate with human performance. The three parameters
measured were electroencephalography, skin temperature, and heart rate, which was measured using Polar H7 Smart
Chest Transmitter in one-second intervals. The experimental temperature was randomly changed to either 22.2°C or
30°C. Seven volunteers were asked to participate in a 155-minute experiment in an experiment room. Also, before
and after each session, the participants were asked to take a short thermal survey. After all data were collected, the
data was analyzed using the linear regression model and the least absolute shrinkage and selection operator (LASSO)
(Nayak et al. 2018). According to the studies mentioned above, many scholars regard heart rate as an essential
parameter related to human comfort, so it is meaningful to consider heart rate as an important biosignal to be measured
in this study.
2.3.2 Stress Level
Stress level is a parameter which measures people’s mental and emotional conditions. Excessive stress would cause
harm on one’s organs. For example, it will cause head, stomach, and chest pain. It may also cause digestive problems.
It would change people’s heart rate and blood pressure. Some institutions have developed sensors which could
calculate stress levels based on measured heart rate variability (HRV). HRV reflects how the heart modulates its
rhythm. It could observe and interpret the interplay between the sympathetic and parasympathetic nervous systems
(Karim, Hasan, and Ali 2011). It could be explained as a parameter that describes the time differences between two
neighboring wave peaks of the heart rate. As shown in Figure 2.1, the time spent from a wave peak to the next is called
R-R intervals, and the change in R-R intervals is the heart rate variability (Figure 2.2). The variation is not random; it
changes depending on the external complex stimulations affecting the heart (Acharya et al. 2006). Human comfort
was reflected as the changes on homeostasis affected by changes in the environment, which then influences HRV
(Berntson et al. 1997). Thus, observing and analyzing heart rate variation has become a popular method. In building
science, the stress condition of occupants in a building is valuable for study because one of the targets of building
science is to make occupants less stressed when they are staying in the building. Garmin Corporation developed a
wearable sensor, Vivosmart 3, to compute stress levels according to measured HRV based on Garmin’s own formulas.
The stress level ranges from 0 to 100, where 0 to 25 is a resting state, 26 to 50 is low stress, 51 to 75 is medium stress,
and 76 to 100 is high stress (“Heart Rate Variability and Stress Level” n.d.).
Figure 2.1 Heart Rate Curve
7
Figure 2.2 Heart Rate Variability (HRV)
2.3.3 Electrodermal Activity
Electrodermal activity (EDA) is a characteristic of the human body which can lead to persistent changes in electrical
characteristics of the skin. It is regarded as a solution for continuously monitoring autonomic nervous system (ANS)
activity caused by stress because the sympathetic branch of the ANS, which determines EDA, mainly controls and
adjusts the stress condition (Boucsein 2012). The EDA sensor usually measures the resistance of the human skin in
high frequency; when the human body is in different environments, the state of the sweat glands and the amount of
perspiration would also be different. As the amount of perspiration affects the electrical resistance of the skin, then
the skin resistance measured by the EDA sensor changes constantly when the environment changes.
Chrisinger and King conducted a study to explore the relationships between biometric signals and thermal comfort to
improve the building environment. A total of 14 adults participated in this investigation, and they were required to
wear a wrist sensor collecting three-axis accelerometry, skin temperature, blood volume pressure, heart rate, heartbeat
interbeat interval, and electrodermal activity. They were also requested to use a smartphone-based application called
Discovery Tool (DT), to which the participants uploaded and shared images and audio about their satisfaction or
complaints about the environment and where every participant could vote for positive or negative opinions. After
collecting the data from DT and the wearable sensor, statistical analysis was conducted by the authors to investigate
the relationship between the images/audios and biometric data (Chrisinger and King 2018).
Therefore, EDA could be used to observe the impact of the surrounding environment on the human body. Similar to
heart rate, when the human body is under stress, the amount of perspiration in the sweat glands will change, leading
to significant changes in EDA values. Bornoiu and Grigore studied the relationship between EDA and human stress
and argued that EDA could be used as a predictor of stress in humans (Bornoiu and Grigore 2014).
2.3.4 Skin Temperature
Similar to heart rate and EDA, skin temperature is also used as an objective indicator to monitor the impact of the
environment on the human body. The researchers can observe the impact of the same stress on different human bodies
through the combination of EDA, heart rate, and skin temperature.
Choi and Loftness explored the possible relationships between skin temperature and thermal sensation by conducting
a correlation analysis. The author performed an experiment on 26 participants in an environmental chamber, and the
experiment for each participant lasted two hours. During the experiment, the authors monitored the skin temperature
of several body segments on participants and collected survey data on their thermal sensation. This paper mentioned
that core body temperature hardly changes because the human body would adjust itself to the environment as much
as possible within a certain temperature range. Compared with core body temperature, skin temperature was more
suitable as an indicator of human thermal comfort. Through correlation analysis of each body segment and the thermal
8
sensation of the human body, it was concluded that skin temperature at the wrist could provide more valuable data to
reflect the human body’s thermal sensation than other body segments (Choi and Loftness 2012).
Èppe pointed out that occupants are in the thermal comfort zone when the heat transition between the human body
and the environment is balanced and the sweat rate is in a comfort range for occupants. This process depends only on
metabolism (Èppe 2002). Therefore, human skin temperature is often considered and measured as a necessary
physiological parameter related to thermal comfort (Liu et al. 2011).
As mentioned above, by observing heart rate, EDA, and skin temperature, researchers can study, control, and reduce
stress generated by the stimulation of the environment on the human body, thus providing a more comfortable and
healthier environment.
2.4 Statistical Analysis
Mallawaarachchi, De Silva, and Rameezdeen conducted a Spearman’s correlation analysis of survey data using SPSS
v.20. The authors calculated Spearman’s coefficient of correlation between the variables and used it to analyze whether
a correlation exists between them, whether it was a positive or negative correlation, and whether the correlation was
weak or strong. Finally, seven significant factors of IEQ in this surveyed green building were found (Mallawaarachchi,
De Silva, and Rameezdeen 2016). Kim et al. performed t-test and logistic regression analysis on satisfaction survey
data from more than 3,000 participants. Then the authors compared and analyzed the different satisfaction levels of
males and females with each IEQ factor and obtained the factors that had the greatest impact on males or females. The
authors presented the results of the regression as odds ratios (OR), which are ratios of a single result appearing in one
group over another group. An OR greater than 1 meant that this outcome was more likely to appear in the experimental
group than the reference group, and vice versa (Jungsoo Kim et al. 2013).
Ramos et al. mentioned that in a study of indoor hygrothermal conditions, it was found that correlation analysis can
be used as a data mining technique to analyze data. The relationship between two variables could be obtained by
calculating the correlation coefficient ρ. This correlation coefficient is neither greater than 1 nor less than -1. A
correlation coefficient close to 0 indicates that there is no correlation between the two variables. The positive
correlation coefficient indicates that the two variables move in same direction, that is, when one variable increases,
the other variable also increases, and the closer the coefficient is to 1, the stronger the correlation between the two
variables. A negative correlation coefficient indicates that the two variables move in the opposite direction, and the
closer the coefficient is to - 1, the stronger the correlation between the two variables. The authors also mentioned
another data mining technique that can be used to analyze IEQ, which is multiple regression analysis. It could calculate
the relationship between several independent variables and a dependent variable and obtain the following equation:
Y = a0 + a1X1 + a2X2 + … + anXn
where X is the independent variables and Y is the dependent variable. The calculation to obtain the above equation is
very complicated, so it is best to complete it through computer calculation (Ramos et al. 2016).
2.5 Summary
This chapter mainly discusses the background research focusing on the three parts of the collected data and the method
of analyzing the data. First, as an important indicator of building performance, IEQ influences occupants’ productivity,
health, and well-being. And the satisfaction of occupants with building IEQ could be obtained by POE, which is a
subjective data regarding occupants’ perceptions, while bio-signals are objective data regarding occupants’
perceptions of IEQ. Through the correlation and regression analysis of human bio-signals, POE data, and IEQ data,
the relationships between them were determined, which are discussed in the next chapters.
9
3. METHODOLOGY
This thesis aims to explore the impact of IEQ on human comfort by introducing biometric signals. The goal of this
research is to improve human comfort in a multioccupancy environment.
Before data collection, this study selected two buildings in Southern California as the experimental sites for on-site
measurement. To learn about the relationship between IEQ and human comfort, this research collected three types of
data: IEQ, questionnaire, and biometric signal data. The first set of data is IEQ data, which was mainly collected from
four aspects: lighting, acoustic, temperature, and CO2. Three different types of sensors were used to collect them. The
second group of data comes from a questionnaire, which is used to investigate occupants’ perceptions and acceptance
of the executives of IEQ at the same time. The third group of data is biometric signals. This part collected heart rate,
EDA, and skin temperature using two wearable sensors. After data collection, the statistical analysis method was used
to analyze the three types of datasets mentioned above. Also, the relationship between these data and how IEQ affects
occupants’ comfort in a multioccupancy condition was analyzed and conclusions drawn. Figure 3.1 illustrates the
overall workflow of the research study.
10
Figure 3.1 Overall Workflow
11
3.1 IEQ Measurement
This study collected three types of data, of which IEQ data objectively reflects specific indoor environmental
conditions including acoustic, lighting, thermal, and air quality. Thus, when selecting sensors, these parameters should
be the important considerations.
3.1.1 Sensor Selection
Based on the above, IEQ data needs to be collected from acoustic, lighting, thermal, and air quality. The researcher
needed to choose from sensors in the market based on the four criteria. However, there is no sensor that could measure
all these four factors. Then the researcher selected different types of sensors to measure the four parameters according
to data interval, charging time/battery life, storage size, measure range, and raw data accessibility. For data interval,
to make the survey results more accurate, it needs to be in one-minute intervals at least. The reason is that IEQ data
varies rapidly, and the data could fluctuate in one minute. For battery life, longer is better, as the sensors need to be
set and record data in the experiment location for up to eight hours. Meanwhile, the charging time of each sensor is
ideally short. For storage size, it is more appropriate to have the data stored in its memory instead of real-time
synchronization, as it is easier for researchers to export the data from its memory after finishing the experiment than
having an app or computer terminal with the sensor throughout the whole experiment. For measurement range, it is
necessary to make sure that each sensor could measure more than the normal range of IEQ values based on ASHRAE
standards 55, 62.1, and 90.1 (“ANSI/ASHRAE Standard 55-2017 Thermal Environmental Conditions for Human
Occupancy” 2017; “ANSI/ASHRAE Standard 62.1-2013 Ventilation for Acceptable Indoor Air Quality” 2013;
“ASHRAE Standard 90.1 Energy Standard for Buildings Except Low-Rise Residential Buildings” 2010). Table 3.1
shows the summary of IEQ standards. Moreover, raw data which can be exported from the sensors as CSV files is
essential. Some devices in the market which could measure with high frequency can only export in low frequency due
to the privacy policy of the corporation. Therefore, the author selected sensors according to these important criteria.
Table 3.2 below shows the list of sensors for comparison.
Table 3.1 Summary of IEQ Standards
Parameters Guideline
Indoor temperature (℃) Between 19 and 28 ℃ (ASHRAE)
CO2 level (ppm) less than 1000 ppm (ASHRAE)
Workstations illuminance level (lux) Between 200 and 500 lux (ASHRAE)
Acoustic decibel (dBA) Less than 40 dBA (ASHRAE)
Table 3.2 Information of All IEQ Sensors for Comparison
Model Function Interval
Charging
time
Battery
life
Storage size
Measure
range
Raw data access
HOBO
MX2202
Lighting;
Temp
Adjustable × 1 year
96,000
measurements
0-167731
lux;
-4-158F
√ App
HOBO
U12-012
Lighting;
Temp
Adjustable × 1 year
43,000
measurements
1-48437
lux;
-4-158F
√ USB
HOBO
UA-002-64
Lighting;
Temp
Adjustable × 1 year
52,000
measurements
0-320000
lux;
-4-158F
√Communication
device needed
HOBO
MX1102
Temp;
CO2
Adjustable × 1 year
84,650
measurements
32-122F;
0-5000
ppm
√ App
Extech
Instruments
Acoustic Adjustable × 3 days
129,920
measurements
30-130
dBA
√ USB
12
USB Sound
Level Data
Logger
PCE-SDL
1
Acoustic Adjustable × 3 days
129,920
measurements
32-130
dBA
√ USB
PCE-322A Acoustic 2 Hz × 1 day
32,700
measurements
33-130
dBA
√ USB
This study selected three different types of sensors: HOBO MX1102 (Figure 3.2), HOBO MX2202 (Figure 3.3), and
PCE-SDL 1 (Figure 3.4), with each type having three pieces of sensors. The first sensor is the HOBO MX1102, which
was used to collect indoor temperature and CO2 concentration. It can measure temperatures in the range of 0ºC to
50ºC and CO2 concentrations in the range of 0 to 5,000 parts per million (ppm) (“HOBO MX1102 Carbon Dioxide |
Onset HOBO Data Logger” n.d.). The interval of data collection is one minute, and the collected data will be
automatically stored in its memory. When the data collection is completed, it can be synchronized and exported as a
CSV file at one time using an app developed by the HOBO Corporation called HOBO Mobile (Figure 3.5). This app
could configure the sensors and read the data when they are connected via Bluetooth. The second sensor is the HOBO
MX2202, which was used to collect the illumination level of the indoor environment and can measure values in the
range of 0 to 167,731 lux (“HOBO MX2202 Temperature/Light | Onset HOBO Data Loggers” n.d.). The interval for
the data collection was set to one minute. Similarly, its data can be automatically saved in memory, and the CSV file
can be exported at one time after the end of collection also through HOBO Mobile. The third sensor is the PCE-SDL
1, which was used to collect sound level data. Its measurement range is 30 to 130 dBA (decibel A-weighting) (“Class
2 Data-Logging Sound Level Meter PCE-SDL 1 | PCE Instruments” n.d.). The interval for the data collection was set
to one minute as well. Its data can also be automatically stored in memory and exported as a CSV file once collection
is complete through Sound Datalogger (Figure 3.6), which is developed by PCE Instruments. The sensor could set
intervals, download, read, and export data using this software. Each group of surveys collected data continuously for
four working days. As mentioned above, since all these three sensors can store at least one week of data in memory,
the data can be exported once via USB after finishing a set.
Figure 3.2 HOBO MX1102
Figure 3.3 HOBO MX2202
13
Figure 3.4 PCE-SDL 1
Figure 3.5 HOBO Mobile App and Sample Raw Data
Figure 3.6 Sound Datalogger Software and Sample Raw Data
3.1.2 Sensor Arrangement
There are three pieces of sensors for each type, so sensor placement in the area where data is collected is important.
Every sensor was placed in the same open environmental condition. However, because of the mismatch in quantity
between the biometric sensor and the IEQ sensor, this study fails to configure a set of IEQ sensors for each subject’s
workstation, which resulted in a small range of differences in IEQ data of subjects in different seats. Despite the fact
that this limitation cannot be completely eliminated, its effect shall be weakened by selecting subjects in adjacent seats.
In addition, in the subsequent data analysis, different seats could be combined to interpret the analysis results to
enhance the practical value of the conclusions. For example, the data collected near a window may be affected by the
outdoor temperature and daylighting compared with the other data collected by sensors far from the windows, and
seats close to the tea room suffers from more noise than other seats.
14
3.2 Biometric Signal Measurement
To study the physiological condition, or stress condition, of the human body, four parameters—EDA, skin temperature,
heart rate, and stress level—was collected in this study. Thus, these parameters should be considered when selecting
the sensors.
3.2.1 Sensor Selection
When selecting the sensor, in addition to EDA, skin temperature, and heart rate, its simplicity and portability should
also be considered. Because data collection is conducted in a real working environment instead of an experiment in a
lab, it is necessary to consider minimizing the impact of measurement on the normal work and study conditions of
occupants when selecting sensors. Most sensors that can measure EDA, skin temperature, and heart rate are mainly
chest, finger, and wrist strap types. The researcher chose the wearable wrist strap type, which makes the wearer more
comfortable, allowing them to focus on the tasks at hand and not be too affected by the sensor. However, when further
selecting a wrist strap sensor, it is difficult to find a suitable sensor to measure all the required parameters, which are
EDA, skin temperature, and heart rate, because of the limitation of the devices’ function on the market. Besides these
parameters, other characteristics of devices, including charging time, battery life, storage size, and raw data
accessibility, also need to be taken into account. Charging time should be short because all 12 devices need to be
charged before each round of survey. Battery life is preferred to be longer, which is at least one day, because the
experiment would last for one day, and the devices could be charged mostly once per day to keep the experiment
going. For storage size, it should store at least one-day data in its memory for faster data export after one group
collection and before the next group collection. For raw data accessibility, all the sensors need to have the accessibility
to export raw data in CSV or XLS format. Therefore, the author selected sensors according to these important criteria.
Table 3.3 shows the list of wearable sensors for comparison.
Table 3.3 Information of All Wearable Sensors for Comparison
Model Function Interval
Charging
time
Battery
life
Storage size
Raw data
access
Garmin
Vívosmart 4
Heart rate 1 Hz 2 hours 7 days 14 days √ App & USB
Garmin
Vivosmart 3
Heart rate 1 Hz 2 hours 7 days 14 days √ App & USB
Scosche
Rhythm24
Heart rate 1 Hz 1 hour 1 day 13 hours √ App
Biostrap Heart rate 10 Hz 90 mins 2 days 3 days √ App
Garmin
Vivomove HR
Heart rate 1 Hz 90 mins 5 days 14 days √ App
Tempdrop Skin temp 4 Hz 1 hour 1 day 1 day √ App
Ava bracelet
Skin temp;
Heart rate
1 Hz;
1 Hz
2 hours 1 day 13 hours √ App
Empatica
Embrace
Skin temp;
EDA
1 Hz;
4 Hz
2 hours
15
hours
14 hours √ App
Based on the above, this study selected two wearable watches, Garmin Vivosmart 3 (Figure 3.7) and Empatica
Embrace (Figure 3.8). Empatica Embrace was used to measure EDA and skin temperature (“Embrace Seizure
Monitoring | Smarter Epilepsy Management | Embrace Watch | Empatica” n.d.). Battery life for this device is
approximately 15 hours, and charging time takes about one to two hours. Up to 14 hours of data can be stored in
memory without syncing via Bluetooth. EDA’s measurement frequency is 4 Hz while skin temperature frequency is
1 Hz. Garmin Vivosmart 3 was used to measure heart rate and stress level (“Vívosmart® 3 | Fitness Activity Tracker
| GARMIN” n.d.). Battery life for this device is about seven days, and charging time takes about two hours. The
memory can store up to two weeks of data. The interval of raw data of the heart rate and stress level that can be
exported after measurement is one minute. However, because of Garmin’s requirements and restrictions, users are not
allowed to directly export raw data in CSV format; only FIT files can be exported. The researcher converted FIT files
into CSV format through a self-written Python program that extracts the required heart rate and stress level data as
well.
15
Figure 3.7 Garmin Vivosmart 3
Figure 3.8 Empatica Embrace
3.2.2 Sensor Arrangement
To set the Embrace and Vivosmart 3 before measuring data, both needed the user to register an account for each device.
Therefore, the researcher registered 12 e-mail address and used them to register Garmin’s account for each Vivosmart
3, which was unnecessary for the Embrace because each subject/device’s username and password could be generated
instantly through the research portal system developed by Empatica as shown in Figure 3.9 (“What Is the Empatica
Research Portal? – Empatica Support” 2019). Each sensor was then numbered by pasting a sticker in the order of 1 to
12, which helps to accurately know the corresponding account name and number of each device (Figure 3.10).
16
Figure 3.9 Empatica Research Portal
Figure 3.10 Numbered Sensors (Assigned to One Participant)
Each participant was assigned two wearable sensors, Garmin Vivosmart 3 and Empatica Embrace, before each round
of data collection. All participants were asked to wear both sensors on their wrist during the work hours in the office
(Figure 3.11). They were not expected to wear the watches out of the office because if the data was recorded outside
work, it cannot be used and analyzed with the other data collected indoors. They could wear the watches on their
preferred wrist side but need to wear them a little tightly to ensure the sensors behind the watch could touch the skin
and measure. All the detailed information was presented and listed with simple instructions on the two watches to
make data collection clearer for every participant. Before the start of the experiment, each participant was expected to
download the Empatica Alert (Figure 3.12) app developed by Empatica on their mobile phones and log in with the
accounts and passwords generated by the research portal, to ensure that the collected data can be synchronized to the
cloud in real time through Bluetooth without any additional operation of the Vivosmart 3. The reason is that data can
be stored in the Vivosmart 3 for up to two weeks in its memory while the Embrace can only store it for less than two
days. In addition, when the Embrace’s memory is full, one to two hours need to be spent exporting data from a device,
17
and there are a total of 12 devices, which is a great challenge to the researcher’s access to raw data. Nevertheless, this
can be solved by requiring each participant to upload data to the cloud through the app. When the storage memory is
full, the Vivosmart 3 will continuously measure and overwrite the oldest files with new data while the Embrace will
stop recording. After the experiment is completed, the researcher will synchronize and export the data stored in the
Vivosmart 3 through USB connection while the data collected by the Embrace will have already been uploaded to the
cloud through the app. All the sensors would be collected after a set of surveys and recharged for the next group
collection.
Figure 3.11 Wear Embrace and Vivosmart 3 during data collection
Figure 3.12 Empatica Alert App
After sensor selection, this study adopted five sensors to measure air temperature, CO2, lighting, acoustics, heart rate,
stress level, EDA, and skin temperature. Table 3.4 below summarizes the specification information about all the
adopted sensors, including IEQ and bio-signal measurement sensors.
Table 3.4 Summary of Adopted Sensors for Measurement
IEQ Measurement Sensors Bio-Signals Measurement Sensors
Model
HOBO MX1102
HOBO MX2202
PCE-SDL 1
Garmin Vivosmart3
Empatica Embrace
Measured
Parameter
- Air temperature
- CO2 level
- Illuminance - Sounds
- Heart rate
- Stress level
- EDA
- Skin temperature
Range
- 0° to 50°C
- 0 to 5,000 ppm
- 0 to 167,731 lux - 30 to 130 dBA
- 0 to 200 bpm
- 0 to 100
- 0.01 μSiemens to 100
μSiemens
- -40 to 115°C
Resolution - 0.024°C at 25°C - 1 lux - 0.1 dBA
- 1 bpm
- 1 unit of stress level
- 1 digit ~900 pSiemens
- 0.02°C
Accuracy
- ±0.21°C from 0° to
50°C
- ±50 ppm ±5% of
reading at 25°C
- ±10% typical for
direct sunlight
- ± 1.4 dB
- ±0.2°C within 36-
39°C
18
Interval
(Setting)
- 1 minute
- 1 minute
- 1 minute - 1 minute
- 1 Hz
- 1 Hz
- 4 Hz
- 1 Hz
Charging
Time
× (lithium battery) × (lithium battery)
× (lithium
battery)
2 hours 2 hours
Battery Life 1 year 1 year 3 days 7 days 15 hours
Storage Size
84,650
measurements
96,000
measurements
129,920
measurements
14 days 14 hours
Raw Data
Accessibility
√ App √ App √ App & USB √ App & USB √ App
3.3 Satisfaction and Sensation Survey
To understand the occupants’ satisfaction with and perception of their indoor environment, this study conducted a
questionnaire for the experimental subjects. While carrying out on-site measurement, the paper-based survey was
distributed to each participant. They were asked to fill out the questionnaire once per hour during the experiment,
which was conducted for five consecutive workdays. They only needed to fill it out when they worked in their
workstation, and they were not asked to do so when they are out of the office. The questionnaire includes demographic
questions, physical conditions, satisfaction, and sensation level of the ambient environment condition. Table 3.5 below
demonstrates the questionnaire on the first day and rest of the four days. The reason for designing them in two versions
is that some questions, such as gender and age, did not need to be asked every day. It is necessary to ask about sickness
every day because if the participant has a fever, their skin temperature would be different from that in normal
conditions. The reason for asking about bedtime and wake-up time is that they affect the stress level or productivity
for that day. It is valuable to know that productivity and stress conditions are partly influenced by sleep quality.
Similarly, it is useful to know the job satisfaction of each participant because if the occupant is not satisfied, this may
affect work productivity. Figure 3.13 shows the final format of the occupants’ satisfaction survey.
Table 3.5 Occupants’ Satisfaction Survey Contents
Category Question
Q01 Demographic/
Human factor
(only for 1
st
day)
What is your age?
Q02 What is your gender?
Q03 How satisfied are you with your job?
Q04
Physical condition
Do you have a fever or other sickness today?
Q05 When was your bed time yesterday and wake up time today?
Q06 Indoor
temperature
How do you feel about the indoor temperature? (sensation)
Q07 How satisfied are you with the indoor temperature?
Q08
Indoor air quality
How do you feel about the indoor air quality? (sensation)
Q09 How satisfied are you with the indoor air quality?
Q10 Indoor lighting
condition
How do you feel about the lighting condition (daylight & artificial)? (sensation)
Q11 How satisfied are you with the lighting condition (daylight & artificial)?
Q12
Acoustic
How do you feel about the acoustic condition? (sensation)
Q13 How satisfied are you with the acoustic condition?
Q14 Overall
satisfaction
How satisfied are you with the overall indoor environmental quality today?
Q15 How satisfied are you with your productivity today?
19
Figure 3.13 Occupants’ Satisfaction Survey (Final Format)
For satisfaction and sensation level, the questionnaire adopted a five-point scale instead of a seven- or three-point
scale (Table 3.6) shown below.
Table 3.6 Summary of Scale Points of the Occupants’ Satisfaction Survey
7-point scale
-3
Strongly
unsatisfied
-2
Unsatisfied
-1
Slightly
unsatisfied
0
Neutral
1
Slightly
satisfied
2
Satisfied
3
Strongly
satisfied
5-point scale
-2
Unsatisfied
-1
Slightly unsatisfied
0
Neutral
1
Slightly satisfied
2
Satisfied
3-point scale Unsatisfied Neutral Satisfied
Generally, the more points in the scale, the higher the accuracy because it is more probable to lose information from
questions with fewer options. Thus, a three-point scale cannot provide enough information. The reason for adopting
the five-point scale is that a classification model needed to be built for different IEQ and biometric data in future steps
after finishing the current study, and it will predict occupants’ satisfaction level. In this case, the output accuracy of
adopting the five-point scale could be quantified as 1/5 while the output accuracy of the seven-point scale is 1/7.
Therefore, this study adopted a five-point scale to improve the accuracy of the classification model.
3.4 Survey Locations and Groups
This study collected data from three survey groups selected from two locations at the University of Southern California
School of Architecture and one location in an office building. The three locations were decided based on their indoor
Occupant Satisfaction Survey
Date: _________________ Name: _________________
1. What is your age?
18-29 30-39 40-49 50-59 60 and above
2. What is your gender?
Male Female Other _________________
3. How satisfied are you with your current working task?
Unsatisfied Slightly unsatisfied Neutral Slightly satisfied Satisfied
4. Do you have a fever or other sickness today?
No Fever Other ___________________
5. When was your bed time yesterday and wake up time today?
Bed time: __________________
Wake up time: __________________
*Answer for sensation questions:
Indoor Temperature Indoor air quality
Cool Slightly cool Neutral Slightly warm warm Bad Not bad Neutral Nice Good
-2 -1 0 1 2 -2 -1 0 1 2
Indoor lighting condition Acoustic
Dark Slightly dark Neutral Slightly bright Bright Noisy Slightly noisy Neutral Slightly quite Quite
-2 -1 0 1 2 -2 -1 0 1 2
*Answer for satisfaction questions:
Unsatisfied Slightly unsatisfied Neutral Slightly satisfied Satisfied
-2 -1 0 1 2
Time
Indoor temperature Indoor air quality Indoor lighting condition Acoustic
sensation satisfaction sensation satisfaction sensation satisfaction sensation satisfaction
2:00pm
2:30pm
3:00pm
3:30pm
4:00pm
4:30pm
5:00pm
5:30pm
6:00pm
How satisfied are you with the overall indoor environmental quality this afternoon? (-2 to 2)
How satisfied are you with your productivity this afternoon? (-2 to 2)
20
environment. They also have similar climate conditions, which means the survey data would not vary too much among
different survey groups. The study selected three survey groups from the two buildings based on their indoor
environment. As this research studies the relationship between IEQ and human comfort in a multioccupancy condition,
it is necessary to consider their indoor environment when selecting experimental subjects. In this case, the independent
office could not meet the requirements of the experiment while an open environment was selected in each of the two
buildings. Because of the limited number of human body sensors, each group of experiments has a maximum of 12
participants. To obtain more data and make the conclusion more accurate, the survey was conducted a total of seven
times in three different locations. This study obtained a total of seven sets of experimental data from 78 experiment
subjects. However, as the workstations of participants in each experimental group are located in different seats in the
office area, their data will be affected to a certain extent. For example, a workstation near the window will receive
more daylighting than one that is farther away. The temperature between the closest workstation to the louver of the
HVAC system and the farther ones would also be slightly different, so it is important to consider the workstation
locations of each survey group according to these factors when analyzing the data. Table 3.7 below shows the time,
location, and subjects of each conducted survey. Figures 3.14, 3.15, and 3.16 show the image of each location. Figures
3.17 and 3.18 show the layout of subjects in Location A; Figure 3.19 shows the layout of subjects in Location B; and
Figure 3.20 shows the layout of subjects in Location C. The blue text in these three figures illustrates the seat of each
subject.
Table 3.7 Time, Location, and Subjects of Each Survey
Building Open Environment Survey Number of Subjects Survey Time
Building A
Location A (studio) Survey A 12 02/01/2019
Location B (basement classroom) Survey C 8 02/07/2019
Location A (studio) Survey B 10 02/08/2019
Building B Location C (office)
Survey D 12 02/12/2019
Survey E 12 02/13/2019
Survey F 12 02/14/2019
Survey G 12 02/15/2019
Figure 3.14 Image of Location A
21
Figure 3.15 Image of Location B
Figure 3.16 Image of Location C
Figure 3.17 Layout of Subjects in Location A (Survey A)
22
Figure 3.18 Layout of Subjects in Location A (Survey B)
Figure 3.19 Layout of Subjects in Location B (Survey C)
23
Figure 3.20 Layout of Subjects in Location C (Surveys D, E, F, G)
3.5 Data Analysis
According to the above, the data was collected from IEQ, biometric signals, and satisfaction and sensation surveys.
After finishing the seven groups’ data collection, bio-signal data including EDA, skin temperature, and heart rate were
synced and exported through the app or USB. Meanwhile, IEQ data including temperature, air quality, lighting, and
acoustic data were also exported through the app or USB into a CSV file. All the participants’ data from one survey
group will be pooled into a single Excel dataset. Minitab was used for data analysis, which mainly conducted statistical
analysis, including correlation and regression.
During data collection, some subjects of the same survey group were measured more than once on different days. This
is because the weather in those days varies, some sunny and some rainy, so even if it is the same group of participants
in the same place, varying weather may have different effects on the data. In the data analysis, more interesting
conclusions may be drawn. The second purpose is to increase the number of experimental groups because when doing
data analysis, data collected on different days were analyzed separately. If there are more experimental groups, it
would be more likely to draw universal conclusions in the comparative analysis of different condition groups. These
conclusions are applicable to different places, occupants, and weather conditions. As to why different days should be
analyzed separately rather than together, it is because different weather and physical sensations and other factors will
affect people’s perception. Therefore, this study only analyzed all participants continuously under the same
environmental condition. For example, if a person’s comfortable temperature is 75ºF, and it is warm outside today,
even if he stays indoors he will still find 75ºF a little warm. However, when it is raining, he may feel that 75ºF indoors
is a little cold. Based on the above reasons, this study divided all the collected data into seven groups by day for
separate analysis. Figure 3.21 shows the analysis and data framework in this article.
163’– 4”
113’–4”
24
Figure 3.21 Data Analysis Framework
3.5.1 Correlation
Correlation analysis refers to the analysis of two or more variable elements to measure their correlation degree. From
the correlation coefficient r, we can know whether the two variables are linear, the strength of the linear relationship,
and whether they are a positive or negative correlation. This study will perform correlation analysis on the two
variables of bio-signal data and occupants’ satisfaction survey, and the two variables of IEQ data and occupants’
satisfaction survey, respectively. Through analysis, the correlation coefficient r is calculated, which is between -1 and
1. If r is close to 0, there is no linear relationship between the two variables. If r is close to -1 or 1, the correlation
between the two variables is very strong. A positive r value means that when the value of y is very small, the value of
x is also very small. As the value of x increases, the value of y also increases, and there is a positive correlation between
them. A negative r value means that the y value will decrease with an increase in x value, and there is a negative
correlation between them. Figure 3.22 is an x-y graph for several different r values (Berg n.d.). Because correlation
analysis is meaningful for variables with linear relationships, a scatterplot of data should be drawn to judge their
relationship before correlation analysis. Sometimes the data will show a curved correlation. In this case, the r value
obtained through correlation analysis is meaningless.
Figure 3.22 Correlation under Different r Values
25
3.5.2 Stepwise Regression
Correlation analysis can confirm the strength of the relationship between two variables but cannot calculate the
regression line, so to find the most suitable line, regression analysis is needed. Regression analysis studies the
relationship between dependent variables (targets) and independent variables (predictors). This technique is
commonly used in predictive analysis, time series models, and finding causal relationships between variables.
Regression analysis can show the significant relationship between independent variables and dependent variables, as
well as the influence intensity of multiple independent variables on a dependent variable. There are mainly seven
regression analysis techniques: linear regression, logistic regression, polynomial regression, stepwise regression, ridge
regression, lasso regression, and elastonet. This article uses stepwise regression, which is usually used to handle
multiple independent variables. This study conducted stepwise regression analysis on IEQ data and bio-signal data.
The basic idea of stepwise regression is to introduce independent variables one by one, introduce the independent
variables with the most significant influence on y each time, test the old variables in the equation one by one, and
eliminate the variables that become insignificant from the equation one by one. The final regression equation neither
omits the variables that have significant influence on y nor contains the variables that have no significant influence on
y. Thus, a formula similar to the equation below is obtained, in which β is the correlation coefficient. The larger the
coefficient before the variable, the greater the influence of the variable on the y value.
y = β0 + β1x1 + β2x2 … + βrxr + ε
3.5.3 Analysis Tool: Minitab
Minitab (Figure 3.23) is a statistical software developed by researchers Barbara F. Ryan, Thomas A. Ryan Jr., and
Brian L. Joiner from Pennsylvania State University in 1972 (“Minitab Statistical Software - Minitab” n.d.). Minitab
is used by more than 90% of the top 100 Fortune companies, and more students worldwide are using Minitab to learn
statistics. In this study, Minitab is used to perform statistical analysis of collected data to explore the relationships
between various parameters. It provides a lot of data processing and analysis functions, such as correlation, regression,
and t-test. In addition, it can also generate an image according to the user’s selection to show the relevant
characteristics of the data.
Figure 3.23 Minitab as Statistical Analysis Tool
The combination of these two analysis tools helped draw final conclusions about the correlation between parameters.
The results and detailed analysis process of this study are discussed in Chapter 4.
26
4. DATA PREPROCESSING
As mentioned in Chapter 3, there are different types of sensors used in data acquisition, including two wearable sensors
and three IEQ sensors. The frequency of sensor recording is high, and the amount of data is large. However, in this
case, the data formats that each sensor exported are different, so the raw data needs to be in unified format and
integrated. This chapter discusses the detailed methods and steps undertaken.
4.1 Raw Data
Each group of surveys has six raw data files, including temperature and CO2, lighting, acoustics, Garmin Vivosmart
3, Empatica Embrace, and questionnaire. The Vivosmart 3 and the Embrace have 12 subfolders. There are seven
groups of surveys in total, and the raw data files of each group are put into one folder. One example folder is shown
in Figure 4.1.
Figure 4.1 Raw Data Files of Each Survey
These raw data will be taken according to their original format through different methods of conversion, extraction,
and integration. Detailed information is discussed below.
4.2 Format Conversion and Data Extraction
Since the raw data formats are different, the steps required to convert them to the final requested format are also
different. Because of some repetitive tasks in the conversion and extraction of data, the author programmed and used
several Python programs to complete part of these tasks. Table 4.1 lists the corresponding Python programs that all
raw data files need to use in the conversion and extraction steps.
Table 4.1 Format Conversion
Raw Data Format Conversion/Extraction Requested Data Format
Human
Data
Questionnaire Paper-based survey Manually input
.csv/.xlsx
Empatica .csv Empatica.py
Garmin .fit Garmin.py, Garmin2.py
IEQ Data
Acoustic .csv
IAQ, Temp .csv
27
Lighting .csv
Because the author collected the satisfaction data of the participants through a paper-based questionnaire, the data
needs to be manually entered into an Excel sheet. The data form obtained is shown in Figure 4.2.
Figure 4.2 Survey Data File
For the two wearable sensors that measured bio-signals, the author performed different steps of conversion based on
the format of the original file and the format ultimately needed. In the raw data on EDA and skin temperature measured
by Empatica, the “time” column is represented through unix timestamp (UTC), not the usual day-month-year format.
The author wrote a Python program called Empatica.py to convert the unix timestamps to human-readable dates to
make this time readable by humans rather than computers. The specific code for this program is attached in Appendix
A. The converted file is shown in Figure 4.3.
Figure 4.3 EDA/Skin Temperature Data Conversion
For the Vivosmart 3, which measured heart rate and stress level, the raw data that can be directly exported by the users
is a file in FIT format, which is a unique format belonging to Garmin. The author wrote a Python program called
Garmin.py to convert the FIT file into a directly readable CSV file. In this program, the author directly converted the
unix timestamp into a human-readable date. The specific code for this program is attached in Appendix B. The
converted file is shown in Figure 4.4.
28
Figure 4.4 Heart Rate/Stress Level Data Conversion 1
The above step completes the first step of the Garmin raw data conversion. As can be seen from the figure above, the
layout in the file completed through Garmin.py conversion is confusing; the table contains a lot of spaces, and the
negative numbers corresponding to the time points are caused by sensor leaks. Because of the large amount of data,
manual sorting takes a lot of time. The author thus wrote a program called Garmin2.py to clean up the spaces and
negative numbers in the table and rearrange it. The specific code for this program is attached in Appendix C. The
converted file is shown in Figure 4.5, which is the requested data format.
Figure 4.5 Heart Rate/Stress Level Data Conversion 2
For acoustic data, the acoustic loggers were connected to computers, and the raw data was exported through the Sound
Datalogger software developed by PCE Instruments. The format of the exported raw data file is XLS, and the times
shown in the sheet are human-readable dates. Therefore, the acoustic data files did not need to be converted.
For the IAQ, temperature, and lighting data, the loggers were connected to HOBO Mobile via Bluetooth, and the raw
data was exported through the app in CSV format. Also, the times in the raw data are human-readable dates, so they
did not need to be converted as well.
29
4.3 Data Integration
After the format conversion of individual files, these files also needed to be uniform in time and unit because they
needed to be integrated into one file to be imported to the analysis software.
First, the author made the time frequency uniform. Since the EDA acquisition frequency is 240 Hz, that is, 240´60
data points per minute, the acquisition frequency of skin temperature is 60 Hz, that is, 60´60 data points per minute.
The other two measured biometric signals (heart rate and stress level) have one data point per minute in the exported
raw data file. All IEQ measurement sensors were also uniformly set to one-minute intervals. Therefore, in the
consolidated file, the time for all data was unified to one data point per minute. For EDA and skin temperature, which
have more than one data point per minute in the raw data, the author adopted the method of averaging, which is taking
the average value of all data in one minute as the data value of that minute. This way, all the data would be consolidated
into a uniform time frequency.
Next, it is necessary to make the units consistent. Of all the variables, only skin temperature and air temperature have
the same unit of measurement. This paper selected SI units while the raw air temperature data is in Fahrenheit, which
is an IP unit, so the author converted it to Celsius.
Finally, this paper also made necessary assumptions about the survey data. As mentioned earlier, the acquisition
frequency of the survey data is hourly, that is, hourly data on IEQ satisfaction and sensation. In data integration, this
paper assumed that the sensation and the satisfaction in every minute of an hour are the same. For example, the values
of satisfaction and sensation in every minute from 15:01 to 16:00 are considered to be the same as those collected at
16:00. If this assumption was not made, only one set of data from IEQ and human body can be used every hour when
analyzing the survey data, and the rest will be wasted. The advantage of this assumption is that the collected data will
not be wasted, and every minute of data can be used for analysis.
Finally, after time frequency unification, unit unification, and satisfaction and sensation assumptions, the sheet shown
in Figure 4.6 is obtained, in which “subject” represents the number of objects collected in the experiment, and all
subjects are arranged in a sheet longitudinally in sequence, as shown in Figure 4.7.
Figure 4.6 Data File (Part) After Integrating
30
Figure 4.7 All Subjects in One Sheet
4.4 Data Cleaning
According to the previous steps, each survey group has an integrated sheet. The next step is to perform data cleaning
on this sheet.
The first step of data cleaning in this article is to clean up all incorrectly collected data, which refers to negative values.
Negative values occurred because the sensor did not stick to the skin sometimes during collection, so the data at those
times were not correctly collected. The author used Excel’s filtering function to filter out these negative values and
then deleted them. The next step is to clear values outside the normal range. All EDA values above 0 are retained in
this step. According to Choi and Loftness’s study, when the air temperature changed from 20°C to 30°C, the measured
skin temperature on the wrist of occupants was from 26°C to 35°C (Choi and Loftness 2012). Thus, for skin
temperature, values lower than 26°C or greater than 35°C were deleted. For heart rate, values lower than 50 bpm or
greater than 130 bpm were deleted.
After data cleaning, the data was placed in a master file for subsequent analysis. The data of each location is stored
separately in different sheets in the master file. The first sheet stores all studio data, the second sheet stores all the
classroom data, and the third sheet stores the office data; see Figure 4.8 below.
Figure 4.8 Master File of All Data
31
4.5 Summary
This chapter mainly discussed the data preprocessing steps before conducting data analysis, which are necessary in
this research. The original raw data files exported directly from sensors were hardly readable and usable for direct
analysis, so several programs in Python were designed to convert the format of the raw data files into Excel files. Also,
the Python program completed data reorganization, data integration, and data cleaning. The data used for analysis
should have unified measurement units and frequency. The data cleaning steps removed the unusual data points which
were outside the normal range, which would influence the outcomes. All the data after preprocessing was placed in a
master file, in which the datasets from three different locations were organized into three separate sheets in an Excel
file. The data analysis after preprocessing is discussed in the next chapter.
32
5. RESULTS AND DISCUSSION
The on-site IEQ and human body measurements and the occupants’ satisfaction survey were simultaneously
performed in three locations in Southern California. The data required statistical analysis after preprocessing. This
chapter focuses on the analysis and discussion of all collected data centered on biometric signals with other parameters.
5.1 Demographic Information of Test Subjects
Based on collected data consisting of IEQ data, bio-signal data, and survey data, this study conducted statistical
analysis to investigate and define the relationships between bio-signals, environmental data, and occupants’ survey
data. Table 5.1 below summarizes the demographic information of all datasets by human factors.
Table 5.1 Demographic Information of All Datasets
Location Survey
Age
Group
Age
Gender
Total
Female Male
A: Studio
Survey A
Junior 18-29 5 5 10
Mid-age 30-39, 40-49 1 1 2
Subtotal 6 6 12
Survey B
Junior 18-29 7 3 10
Mid-age 30-39, 40-49 0 0 0
Subtotal 7 3 10
B: Classroom Survey C
Junior 18-29 4 3 7
Mid-age 30-39, 40-49 0 1 1
Subtotal 4 4 8
C: Office
Survey
D, E, F, G
Zone C1
Junior 18-29 3 2 5
Mid-age 30-39, 40-49 1 1 2
Subtotal 4 3 7
Zone C2
Junior 18-29 3 1 4
Mid-age 30-39, 40-49 0 1 1
Subtotal 3 2 5
Total 45 33 78
As summarized in Table 5.1, a total of 78 subjects from three different locations participated in the study. For survey
A, conducted in a studio, a total of 12 occupants participated in this study, which consists of 6 males and 6 females,
ranging from 18 to 49 years old. There are 10 participants in the junior group (18 to 29 years old) and 2 participants
in the mid-age group (30 to 49 years old). For survey B, also conducted in a studio, a total of 10 participants reported
their satisfaction and sensation with IEQ conditions. There were 3 males and 7 females ranging from 18 to 29 years
old while no participants were from the mid-age group. For survey C conducted in a classroom, there was a total of 8
subjects consisting of 4 males and 4 females and ranging from 18 to 49 years old. Seven participants were in the junior
group, and one participant was in the mid-age group. For surveys D, E, F, and G conducted in an office, the participants
are same in these four survey groups, and they were separated into two zones based on their locations. For zone C1, a
total of seven participants answered the survey, and their human body data was measured. The participants consisted
of 3 males and 4 females ranging from 18 to 49 years old. There were five subjects in the junior group and two in the
mid-age group. For zone C2, a total of five occupants reported their satisfaction and sensation with IEQ. Among them,
2 are males and 3 are females, which ranged from 18 to 49 years old. There were four subjects in the junior group and
one subject in the mid-age group. Overall, the number of subjects in the junior group were more than those in the mid-
age group.
5.2 Location A: Studio
This study collected several parameters and performed statistical analysis to explore and obtain the relationship among
IEQ, physiological data, and satisfaction surveys. This section illustrates the distribution of the measured IEQ data,
including air temperature, CO2, illuminance level, and sound level. Two sets of studio data were combined for analysis
in Location A because they had a total of eight hours, which is the similar to the data measured in other locations. The
datasets of different locations are compared and analyzed in later sections.
33
The frequency distribution of indoor air temperatures is shown in Figure 5.1. According to the comfortable range for
occupants based on the recommended guidelines from standards mentioned in Chapter 3, all measured temperature
values in the studio were in the comfortable range for humans. In Figure 5.2, the CO2 condition was also in the range
that made occupants comfortable. Additionally, the comfort zone of the lighting condition recommended by ASHRAE
standards is 200 to 500 lux. However, Figure 5.3 shows that the lighting condition measured in the studio was not
entirely within the comfort zone. Only about 44% illuminance level was in the comfort range. In Figure 5.4, the whole
acoustic data measured in the studio was outside the comfort range, which is a maximum of 40 dBA. The data ranged
between 45 to 65 dBA, which is higher than the standards.
Figure 5.1 Frequency Distribution of the Indoor Air Temperature in the Studio
Figure 5.2 Frequency Distribution of CO2 in the Studio
0.2%
30.2%
69.3%
0.2%
0
50
100
150
200
250
300
22.5-23 23-23.5 23.5-24 24-24.5
Frequency
Temperature (℃)
Air Temperature
10.5%
20.1%
32.4%
34.6%
2.2%
0.2%
0
20
40
60
80
100
120
140
160
430-459 460-489 490-519 520-549 550-579 580-609
Frequency
CO2 (ppm)
CO2
34
Figure 5.3 Frequency Distribution of Lighting Conditions in the Studio
Figure 5.4 Frequency Distribution of Acoustic Conditions in the Studio
Table 5.2 shows the summary of measured IEQ data in the studio. All CO2 level and air temperature data were within
ASHRAE’s recommended standards. The mean value of indoor air temperature was 23.58℃, and the standard
deviation was 0.22, which indicates that the air temperature remained stable in the studio. The mean value of CO2
level was 505.88 ppm, and the standard deviation is 30.70, which means that the CO2 data fluctuated in the studio.
The reason may be that occupants moved around in the studio instead of staying in their seats, so the CO2 level
probably increased when many occupants gathered together. The studio had poor lighting conditions; only 44% of the
lighting data is within the range of the guidelines. The mean value is 139.93 lux, which is lower than the minimum
200 lux, and the standard deviation is 97.84, which indicates that the lighting data fluctuated significantly in the studio.
This was because of the sunlight; the measured illuminance data was high at noon and then went down in the late
afternoon. The acoustic data measured in the studio was entirely outside the guidelines. The mean value is 55.44 dBA,
which is higher than the maximum 40 dBA, and the standard deviation is 2.66, which means the acoustic data was
stable in the studio. According to the table, the distributions of acoustic and lighting conditions indicated that
occupants in the studio were in a noisy and slightly dark indoor environment.
Table 5.2 Summary of Measured IEQ Data in the Studio
IEQ factor Mean StDev Within guideline
Air temperature (℃) 23.58 0.22 100%
CO2 level (ppm) 505.88 30.70 100%
Illuminance level (lux) 139.93 97.84 44%
Acoustic (dBA) 55.44 2.66 0%
27.7%
5.1%
19.0%
4.6%
34.5%
9.0%
0
20
40
60
80
100
120
140
160
0-49 50-99 100-149 150-199 200-249 250-300
Frequency
Illuminance level (lux)
Lighting
1.8%
42.6%
50.8%
4.8%
0
50
100
150
200
250
45-50 50-55 55-60 60-65
Frequency
Sound level (dBA)
Sound Level
35
After preprocessing of all the data, stepwise regression analysis was conducted using the software Minitab. The author
performed stepwise regression analysis of IEQ with bio-signals and bio-signals with survey satisfaction/sensation,
which included IEQ with EDA, IEQ with skin temperature, IEQ with heart rate, IEQ with stress level, bio-signals with
thermal sensation, bio-signals with thermal satisfaction, bio-signals with IAQ sensation, bio-signals with IAQ
satisfaction, bio-signals with lighting sensation, bio-signals with lighting satisfaction, bio-signals with acoustic
sensation, bio-signals with acoustic satisfaction, and bio-signals with overall satisfaction. The results are summarized
and shown in the tables below. The original stepwise regression results generated by Minitab is in Appendix D.
Table 5.3 shows the summary of the impact of IEQ on bio-signals and of bio-signals on satisfaction/sensation in the
studio. The p-values obtained from the stepwise regression analysis, with most values lower than 0.05, indicated that
the analyses are statistically significant. The r-square values are enclosed in parentheses in the tables. Since this study
focused on the variation regarding how overall IEQ can be explained by EDA, skin temperature, heart rate, and stress
level, the purpose of the study was to investigate and identify the significant order of the impact on parameters. Thus,
while the r-square values listed in the table are lower than 70%, the order of contributing parameters found in this
study are still valuable. It can be seen that EDA, heart rate, and skin temperature were most affected by lighting
conditions while skin temperature was most affected by CO2 level. Thermal satisfaction was most affected by EDA;
IAQ satisfaction and overall satisfaction were most affected by heart rate, lighting satisfaction was affected by skin
temperature, and acoustic satisfaction was most affected by stress condition. It appears that almost all the different
satisfaction parameters were affected by different bio-signals.
Table 5.3 Summary of the Impact of IEQ on Bio-Signals and Bio-Signals on Satisfaction/Sensation in the Studio
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Studio
Bio-Signals
Lighting
(3.71%)
Acoustic
(4.03%)
Temp
(4.14%)
CO2
(4.31%)
EDA
CO2
(6.41%)
Temp
(8.92%)
Lighting
(10.33%)
Skin Temp
Lighting
(4.07%)
CO2
(4.96%)
Temp
(6.20%)
Heart Rate
Lighting
(15.16%)
Temp
(16.34%)
Stress Level
Studio
Satisfaction/Sensation
EDA
(1.19%)
Skin temp
(1.63%)
Thermal Satisfaction
Skin temp
(5.27%)
EDA
(6.29%)
HR
(6.50%)
Stress
(7.97%)
Thermal Sensation
HR
(0.42%)
EDA
(0.79%)
IAQ Satisfaction
EDA
(0.58%)
Stress
(1.00%)
IAQ Sensation
Skin temp
(1.00%)
EDA
(1.25%)
Lighting Satisfaction
Stress
(0.66%)
HR
(1.06%)
EDA
(1.41%)
Lighting Sensation
Stress
(1.36%)
HR
(3.84%)
EDA
(4.58%)
Acoustic Satisfaction
Stress
(1.32%)
HR
(2.64%)
Skin temp
(3.10%)
Acoustic Sensation
HR
(1.54%)
EDA
(2.12%)
Skin temp
(2.39%)
Stress
(2.60%)
Overall Satisfaction
5.2.1 Impact by Gender
Datasets were analyzed by being grouped in different genders and age groups. Table 5.4 below shows the summary
of the impact on bio-signals of IEQ by different genders in the studio. It indicates that indoor lighting conditions
36
influenced stress levels most significantly in both females and males followed by air temperature and CO2 level in
that order. It can also be seen that males’ biometric signals were significantly affected by lighting conditions. Besides
the first-ranked lighting condition, CO2 also affected females’ and males’ bio-signals as the second factor. The reason
is that the lighting and CO2 data fluctuated the most among the four IEQ parameters, so the bio-signals were easily
affected by changes in them.
Table 5.4 Summary of the Impact of IEQ on Bio-Signals by Different Genders in the Studio
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Female
CO2
(1.51%)
Lighting
(1.96%)
Acoustic
(2.22%)
EDA
Male
Lighting
(17.64%)
Temp
(18.39%)
Acoustic
(18.68%)
CO2
(18.83%)
Female
Lighting
(7.47%)
CO2
(10.05%)
Temp
(15.17%)
Skin Temp
Male
CO2
(25.30%)
Acoustic
(25.57%)
Female
Lighting
(6.97%)
CO2
(9.13%)
Temp
(10.57%)
Heart Rate
Male
Temp
(1.02%)
Lighting
(1.65%)
Female
Lighting
(18.80%)
CO2
(19.71%)
Temp
(20.23%)
Stress Level
Male
Lighting
(9.90%)
Temp
(18.64%)
CO2
(21.14%)
Table 5.5 shows the summary of the impact of bio-signals on satisfaction and sensation by different genders in the
studio. Regarding thermal sensation, thermal satisfaction, IAQ sensation, IAQ satisfaction, lighting satisfaction, and
acoustic sensation, females were most affected by skin temperature while for lighting sensation and acoustic sensation,
they were most affected by EDA. For males, EDA mostly affected their thermal sensation and acoustic satisfaction;
their IAQ sensation, IAQ satisfaction, and lighting sensation were most affected by skin temperature. And regarding
the impact on overall IEQ satisfaction, both females and males were affected by stress level. Overall, the data could
indicate that both females and males are sensitive to skin temperature.
Table 5.5 Summary of the Impact of Bio-Signals on Satisfaction/Sensation by Different Genders in the Studio
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Female
Skin temp
(11.36%)
HR
(12.49%)
Stress
(15.89%)
Thermal Sensation
Male
EDA
(10.87%)
HR
(13.74%)
Skin temp
(14.70%)
Female
Skin temp
(7.15%)
HR
(7.87%)
Thermal Satisfaction
Male
HR
(4.76%)
Skin temp
(8.97%)
EDA
(10.76%)
Stress
(12.35%)
Female
Skin temp
(6.37%)
EDA
(7.33%)
Stress
(8.51%)
HR
(8.96%)
IAQ Sensation
Male
Skin temp
(4.77%)
HR
(5.07%)
Stress
(6.68%)
Female
Skin temp
(4.15%)
HR
(5.57%)
Stress
(8.68%)
EDA
(9.03%)
IAQ Satisfaction
Male
Skin temp
(0.84%)
HR
(1.32%)
Stress
(4.10%)
Female
EDA
(4.25%)
Skin temp
(7.13%)
Stress
(9.39%)
Lighting Sensation
37
Male
Skin temp
(1.44%)
EDA
(2.85%)
HR
(3.71%)
Stress
(4.13%)
Female
Skin temp
(2.56%)
EDA
(4.54%)
Stress
(4.73%)
HR
(5.91%)
Lighting Satisfaction
Male
Stress
(0.71%)
Skin temp
(1.36%)
Female
EDA
(4.69%)
HR
(5.59%)
Stress
(8.46%)
Skin temp
(8.91%)
Acoustic Sensation
Male
Skin temp
(2.53%)
Stress
(3.90%)
HR
(7.70%)
EDA
(7.98%)
Female
Skin temp
(1.67%)
Stress
(2.45%)
HR
(4.62%)
EDA
(5.12%)
Acoustic Satisfaction
Male
EDA
(2.92%)
HR
(4.38%)
Stress
(11.16%)
Skin temp
(13.06%)
Female
Stress
(3.21%)
HR
(4.65%)
EDA
(5.42%)
Skin temp
(6.01%)
Overall Satisfaction
Male
Stress
(5.60%)
HR
(10.80%)
5.2.2 Impact by Age
The previous section illustrates the analysis of the relationships between bio-signals, IEQ, and satisfaction and
sensation by different genders. This section shows the analysis considering different age groups. Table 5.6 shows a
summary of the impact of IEQ on the bio-signals of different age groups in the studio. The blank in the table means
Minitab did not select any correlated factors from the provided factors. It can be seen that for the impact on all the
four bio-signals, the junior group was most affected by lighting condition while for the impact on EDA, the mid-age
group was most sensitive to CO2 conditions. For the effect on skin temperature, the mid-age group was most affected
by acoustic condition. From the overall table, it is difficult to compare the differences between the junior group and
the mid-age group because in the mid-age group, stepwise regression only involved selecting the impact on EDA and
skin temperature when analyzing how IEQ factors affected bio-signals.
Table 5.6 Summary of the Impact of IEQ on Bio-Signals by Different Age Groups in the Studio
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Junior
Lighting
(3.04%)
Acoustic
(3.44%)
Temp
(3.59%)
CO2
(3.73%)
EDA
Mid-age
CO2
(1.26%)
Lighting
(3.11%)
Temp
(5.33%)
Junior
CO2
(7.44%)
Temp
(10.15%)
Lighting
(11.50%)
Skin Temp
Mid-age
Acoustic
(1.16%)
Junior
Lighting
(2.51%)
Temp
(3.28%)
CO2
(4.63%)
Heart Rate
Mid-age
Junior
Lighting
(11.46%)
CO2
(13.23%)
Stress Level
Mid-age
Table 5.7 lists the summary of the impact of bio-signals on the satisfaction and sensation of different age groups in
the studio. It can be seen that for the impact on thermal sensation, thermal satisfaction, and IAQ sensation, both junior
and mid-age groups were most affected by skin temperature. Besides, skin temperature also affected the junior group’s
lighting satisfaction and the mid-age group’s IAQ satisfaction and overall satisfaction. Besides skin temperature, EDA
and heart rate affected occupants’ satisfaction as well. For instance, EDA affected the lighting sensation of both males
38
and females. From the overall table, skin temperature was the top bio-signal that relatively influenced the sensation
and satisfaction in both junior and mid-age groups.
Table 5.7 Summary of the Impact of Bio-Signals on Satisfaction/Sensation by Different Age Groups in the Studio
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Junior
Skin temp
(5.91%)
HR
(8.19%)
EDA
(9.74%)
Stress
(10.26%)
Thermal Sensation
Mid-age
Skin temp
(32.21%)
Junior
Skin temp
(1.44%)
HR
(2.32%)
EDA
(2.97%)
Thermal Satisfaction
Mid-age
Skin temp
(79.87%)
Stress
(81.09%)
Junior
Skin temp
(1.89%)
HR
(2.90%)
Stress
(5.05%)
EDA
(5.66%)
IAQ Sensation
Mid-age
Skin temp
(76.71%)
HR
(80.88%)
Stress
(83.24%)
EDA
(83.68%)
Junior
HR
(1.55%)
Stress
(2.44%)
Skin temp
(2.99%)
EDA
(3.26%)
IAQ Satisfaction
Mid-age
Skin temp
(76.71%)
HR
(80.88%)
Stress
(83.24%)
EDA
(83.68%)
Junior
EDA
(0.22%)
HR
(0.44%)
Stress
(0.69%)
Lighting Sensation
Mid-age
EDA
(6.29%)
HR
(11.63%)
Stress
(17.85%)
Skin temp
(22.48%)
Junior
Skin temp
(0.89%)
HR
(1.39%)
EDA
(1.77%)
Lighting Satisfaction
Mid-age
EDA
(6.29%)
HR
(11.63%)
Stress
(17.85%)
Skin
(22.48%)
Junior
Stress
(3.10%)
HR
(3.27%)
EDA
(3.44%)
Acoustic Sensation
Mid-age
Skin temp
(63.42%)
HR
(71.27%)
Stress
(72.86%)
Junior
Stress
(3.50%)
EDA
(4.43%)
Skin temp
(5.31%)
HR
(5.56%)
Acoustic Satisfaction
Mid-age
HR
(58.35%)
Skin temp
(65.30%)
Stress
(66.59%)
Junior
HR
(2.34%)
EDA
(2.74%)
Overall Satisfaction
Mid-age
Skin temp
(76.71%)
HR
(80.88%)
Stress
(83.24%)
EDA
(83.68%)
5.3 Location B: Classroom
IEQ measurement, bio-signal measurement, and survey collection were conducted in location B, which is a basement
classroom, for about four hours. The frequency distribution of indoor air temperature is shown in Figure 5.5.
According to the comfortable range for occupants based on the recommended guidelines, all measured temperature
values in the classroom were in the human comfort range. In Figure 5.6, the CO2 condition was also in the range that
made occupants comfortable. Figure 5.7 shows the frequency distribution of the lighting condition in the classroom.
The lighting data was entirely outside the comfort zone according to the guideline. The measured illuminance values
were lower than the minimum 200 lux. In Figure 5.8, the whole acoustic data measured in the classroom was outside
the comfort range, which is a maximum of 40 dBA. The data ranged between 45 to 70 dBA, which is higher than the
standard.
39
Figure 5.5 Frequency Distribution of Indoor Air Temperature in the Classroom
Figure 5.6 Frequency Distribution of CO2 in the Classroom
Figure 5.7 Frequency Distribution of Lighting Conditions in the Classroom
0.5%
6.1%
11.8%
30.2%
51.4%
0
20
40
60
80
100
120
22.5-23 23-23.5 23.5-24 24-24.5 24.5-25
Frequency
Temperature (℃)
Air Temperature
1.4%
5.2%
18.9%
33.5%
28.8%
12.3%
0
10
20
30
40
50
60
70
80
830-849 850-869 870-889 890-909 910-929 930-950
Frequency
CO2 (ppm)
CO2
0.5%
11.8%
87.7%
0
50
100
150
200
108-108 118-118 128-129
Frequency
Illuminance level (lux)
Lighting
40
Figure 5.8 Frequency Distribution of Acoustic Conditions in the Classroom
Table 5.8 shows the summary of measured IEQ data in the classroom. All the CO2 level and air temperature data were
within ASHRAE standards. The mean value of indoor air temperature was 24.32℃, and the standard deviation was
0.40, which indicates that the air temperature remained stable in the classroom. The mean value of CO2 level is 903.18
ppm, and the standard deviation is 22.03, which means that the CO2 data fluctuated in the classroom. The reason may
be that students listened more than discussed during the lecture, so the CO2 level probably went high when more
occupants spoke. The classroom did not have good lighting conditions since none of the lighting data was in the range
of the guidelines. The mean value was 127.60 lux, which is lower than the minimum 200 lux, and the standard
deviation is 3.80, which means the lighting data was stable in the classroom. This was because the professor needed
a projector for the lecture, and the artificial lights were set slightly darker in the classroom to make the projector work
better. Additionally, the classroom was in the basement and did not have windows, so all the measured lighting data
were actually from artificial lights. This is why the illuminance data was stable in the classroom. The acoustic data
measured in the studio was entirely outside the guidelines. The mean value was 53.92 dBA, which was higher than
the maximum 40 dBA from the acoustic comfort zone, and the standard deviation is 3.91, which means the acoustic
data was stable in the classroom. According to the table, the distributions of acoustic and lighting condition indicate
that occupants in the classroom were in a noisy and slightly dark indoor environment.
Table 5.8 Summary of Measured IEQ Data in the Classroom
IEQ factor Mean StDev Within guideline
Air temperature (℃) 24.32 0.40 100%
CO2 level (ppm) 903.18 22.03 100%
Illuminance level (lux) 127.60 3.80 0%
Acoustic (dBA) 53.92 3.91 0%
Table 5.9 summarized the impact of IEQ on bio-signals and of bio-signals on satisfaction/sensation in the classroom.
EDA and heart rate were most affected by lighting condition while skin temperature and stress level were most affected
by air temperature. All the four bio-signals influenced satisfaction and sensation, and it was found that skin
temperature was the most frequent factor.
Table 5.9 Summary of the Impact of IEQ on Bio-Signals and of Bio-Signals on Satisfaction/Sensation in the
Classroom
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Classroom
Bio-Signals
Lighting
(0.39%)
Temp
(1.84%)
EDA
Temp
(2.66%)
Lighting
(5.66%)
Acoustic
(7.18%)
CO2
(7.53%)
Skin Temp
Lighting
(0.75%)
Acoustic
(1.43%)
Heart Rate
17.9%
46.3%
28.4%
7.0%
0.5%
0
20
40
60
80
100
45-50 50-55 55-60 60-65 65-70
Frequency
Sound level (dBA)
Sound Level
41
Temp
(3.41%)
Stress Level
Classroom
Satisfaction/Sensation
EDA
(5.65%)
Skin temp
(9.81%)
HR
(11.49%)
Stress
(13.79%)
Thermal Satisfaction
Stress
(1.56%)
HR
(2.86%)
Thermal Sensation
Skin temp
(7.84%)
EDA
(12.27%)
HR
(19.33%)
Stress
(19.96%)
IAQ Satisfaction
HR
(4.43%)
EDA
(10.85%)
Stress
(11.99%)
Skin temp
(13.04%)
IAQ Sensation
Stress
(15.19%)
EDA
(16.59%)
HR
(17.43%)
Lighting Satisfaction
Skin temp
(21.28%)
EDA
(30.14%)
Stress
(31.04%)
HR
(32.42%)
Lighting Sensation
Skin temp
(23.63%)
EDA
(26.75%)
Stress
(27.39%)
Acoustic Satisfaction
Skin temp
(34.44%)
HR
(37.11%)
Stress
(38.63%)
EDA
(39.13%)
Acoustic Sensation
HR
(9.68%)
EDA
(19.21%)
Overall Satisfaction
5.3.1 Impact by Gender
Table 5.10 shows the summary of the impact of IEQ on bio-signals by different genders in the classroom. For the
impact on stress condition, both females and males were most significantly affected by lighting condition. However,
for the impact on skin temperature, females tended to be sensitive to CO2 followed by acoustic condition while males
were more sensitive to lighting condition and then CO2. For the impact on EDA, females were most affected by CO2
while males were mostly affected by lighting condition. From the overall table, the lighting condition mostly affected
all the four bio-signals followed by CO2. This is probably because the illuminance level in the basement classroom
was slightly darkened to see the projector screen, and by concentrating on the lecture and screen, their satisfaction
was easily affected by the lighting condition.
Table 5.10 Summary of the Impact of IEQ on Bio-Signals by Different Genders in the Classroom
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Female
CO2
(1.51%)
Lighting
(1.96%)
Acoustic
(2.22%)
EDA
Male
Lighting
(17.64%)
Temp
(18.39%)
Acoustic
(18.68%)
Female
Lighting
(7.47%)
CO2
(10.05%)
Temp
(15.17%)
Skin Temp
Male
CO2
(25.30%)
Acoustic
(25.57%)
Female
Lighting
(6.97%)
CO2
(9.13%)
Temp
(10.57%)
Heart Rate
Male
Temp
(1.02%)
Lighting
(1.65%)
Female
Lighting
(18.80%)
CO2
(19.71%)
Temp
(20.23%)
Stress Level
Male
Lighting
(9.90%)
Temp
(18.64%)
CO2
(21.14%)
Table 5.11 shows the summary of the impact of bio-signals on satisfaction and sensation by different genders in the
classroom. Except lighting sensation, the other collected sensation and satisfaction data were all most significantly
42
affected by skin temperature in the female group. Their lighting sensation was most affected by EDA. For the impact
on thermal sensation and acoustic satisfaction, males were easily affected by EDA while for the impact on IAQ
satisfaction, IAQ sensation, lighting sensation, and acoustic sensation, males were most affected by skin temperature.
From the overall table, in every set of the four-part survey datasets (thermal, IAQ, acoustic, lighting), skin temperature
was the top bio-signal affecting sensation and satisfaction in both female and male groups.
Table 5.11 Summary of the Impact of Bio-Signals on Satisfaction/Sensation by Different Genders in the Classroom
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Female
Skin temp
(11.36%)
HR
(12.49%)
Stress
(15.89%)
Thermal Sensation
Male
EDA
(10.87%)
HR
(13.74%)
Skin temp
(14.70%)
Female
Skin temp
(7.15%)
HR
(7.87%)
Thermal Satisfaction
Male
HR
(4.76%)
Skin temp
(8.97%)
EDA
(10.76%)
Stress
(12.35%)
Female
Skin temp
(6.37%)
EDA
(7.33%)
Stress
(8.51%)
HR
(8.96%)
IAQ Sensation
Male
Skin temp
(4.77%)
HR
(5.07%)
Stress
(6.68%)
Female
Skin temp
(4.15%)
HR
(5.57%)
Stress
(8.68%)
EDA
(9.03%)
IAQ Satisfaction
Male
Skin temp
(0.84%)
HR
(1.32%)
Stress
(4.10%)
Female
EDA
(4.25%)
Skin temp
(7.13%)
Stress
(9.39%)
Lighting Sensation
Male
Skin temp
(1.44%)
EDA
(2.85%)
HR
(3.71%)
Stress
(4.13%)
Female
Skin temp
(2.56%)
EDA
(4.54%)
Stress
(4.73%)
HR
(5.91%)
Lighting Satisfaction
Male
Stress
(0.71%)
Skin temp
(1.36%)
Female
EDA
(4.69%)
HR
(5.59%)
Stress
(8.46%)
Skin temp
(8.91%)
Acoustic Sensation
Male
Skin temp
(2.53%)
Stress
(3.90%)
HR
(7.70%)
Female
Skin temp
(1.67%)
Stress
(2.45%)
HR
(4.62%)
EDA
(5.12%)
Acoustic Satisfaction
Male
EDA
(2.92%)
HR
(4.38%)
Stress
(11.16%)
Skin temp
(13.06%)
Female
Stress
(3.21%)
HR
(4.65%)
EDA
(5.42%)
Skin temp
(6.01%)
Overall Satisfaction
Male
Stress
(5.60%)
HR
(10.80%)
5.3.2 Impact by Age
Table 5.12 illustrates the summary of the impact of IEQ on bio-signals by different age groups in the classroom. For
the impact on all the four bio-signals except skin temperature, the junior group was always most affected by lighting
condition. The skin temperature of the junior group was most affected by CO2. However, there were some differences
in the mid-age group. For the impact on EDA, the mid-age group was most affected by CO2 while for the impact on
skin temperature, the mid-age group was most influenced by acoustic condition.
Table 5.12 Summary of the Impact of IEQ on Bio-Signals by Different Age Groups in the Classroom
43
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Junior
Lighting
(3.04%)
Acoustic
(3.44%)
Temp
(3.59%)
CO2
(3.73%)
EDA
Mid-age
CO2
(1.26%)
Lighting
(3.11%)
Temp
(5.33%)
Junior
CO2
(7.44%)
Temp
(10.15%)
Lighting
(11.50%)
Skin Temp
Mid-age
Acoustic
(1.16%)
Junior
Lighting
(2.51%)
Temp
(3.28%)
CO2
(4.63%)
Heart Rate
Mid-age
Junior
Lighting
(11.46%)
Temp
(13.23%)
Stress Level
Mid-age
Table 5.13 illustrates the summary of the impact of bio-signals on satisfaction and sensation by different age groups
in the classroom. For the impact on thermal sensation/satisfaction and IAQ sensation, both junior and mid-age groups
were most affected by skin temperature. For the impact on lighting sensation and satisfaction, the mid-age group was
most affected by EDA followed by heart rate. Except lighting, for the impact on other sensations and satisfaction, the
mid-age group was mostly affected by skin temperature in this dataset. In the junior group, all the four bio-signals
affected their satisfaction and sensation. However, skin temperature was the most frequent factor affecting the junior
group’s satisfaction. From the overall table, in every set of the four-part survey datasets (thermal, IAQ, acoustic,
lighting), the mid-age group had almost the same bio-signal effect on both sensation and satisfaction while such effect
was different for males.
Table 5.13 Summary of the Impact of Bio-Signals on Satisfaction/Sensation by Different Age Groups in the
Classroom
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Junior
Skin temp
(5.91%)
HR
(8.20%)
EDA
(9.75%)
Stress
(10.27%)
Thermal Sensation
Mid-age
Skin temp
(32.21%)
Junior
Skin temp
(1.44%)
HR
(2.32%)
EDA
(2.97%)
Thermal Satisfaction
Mid-age
Skin temp
(79.87%)
Stress
(81.09%)
Junior
Skin temp
(1.90%)
HR
(2.88%)
Stress
(5.03%)
EDA
(5.64%)
IAQ Sensation
Mid-age
Skin temp
(76.71%)
HR
(80.88%)
Stress
(83.24%)
EDA
(83.68%)
Junior
HR
(1.54%)
Stress
(2.43%)
Skin temp
(2.98%)
EDA
(3.25%)
IAQ Satisfaction
Mid-age
Skin temp
(76.71%)
HR
(80.88%)
Stress
(83.24%)
EDA
(83.68%)
Junior
EDA
(0.23%)
HR
(0.43%)
Stress
(0.68%)
Lighting Sensation
Mid-age
EDA
(6.29%)
HR
(11.63%)
Stress
(17.85%)
Skin temp
(22.48%)
Junior
Skin temp
(0.89%)
HR
(1.39%)
EDA
(1.76%)
Lighting Satisfaction
Mid-age EDA HR Stress Skin temp
44
(6.29%) (11.63%) (17.85%) (22.48%)
Junior
Stress
(3.11%)
HR
(3.27%)
EDA
(3.44%)
Acoustic Sensation
Mid-age
Skin temp
(63.42%)
HR
(71.27%)
Stress
(72.86%)
Junior
Stress
(3.51%)
EDA
(4.45%)
Skin temp
(5.33%)
HR
(5.57%)
Acoustic Satisfaction
Mid-age
HR
(58.35%)
Skin temp
(65.30%)
Stress
(66.59%)
Junior
HR
(2.34%)
EDA
(2.74%)
Overall Satisfaction
Mid-age
Skin temp
(76.71%)
HR
(80.88%)
Stress
(83.24%)
EDA
(83.68%)
5.4 Location C: Office Zone C1
IEQ measurement, bio-signal measurement, and survey collection were conducted in location C, which is an open
office, for about eight hours per day for four days. Depending on the distribution of the locations of participants’ seats,
all the data collected in the office was divided into two parts: zone C1 and zone C2. This section discusses the analysis
of zone C1, and zone C2 is explained in the next section. The frequency distribution of indoor air temperature is shown
in Figure 5.9. According to the comfortable range for occupants based on the recommended guidelines, all measured
temperature values in office zone C1 were in the human comfortable range. In Figure 5.10, the CO2 condition was
also in the range that made occupants comfortable. Figure 5.11 shows the frequency distribution of lighting condition
in office zone C2. The lighting data was entirely outside the comfort zone according to the guidelines. The measured
illuminance values were lower than the minimum 200 lux. In Figure 5.12, the whole acoustic data measured in office
zone C1 was outside the comfort range, which is a maximum of 40 dBA. The data ranged between 50 to 70 dBA,
which is higher than the standard.
Figure 5.9 Frequency Distribution of Indoor Air Temperature in Office Zone C1
20.7%
48.1%
31.2%
0
200
400
600
800
1000
22-22.5 22.5-23 23-23.5
Frequency
Temperature (℃)
Air Temperature
45
Figure 5.10 Frequency Distribution of CO2 in Office Zone C1
Figure 5.11 Frequency Distribution of Lighting Conditions in Office Zone C1
Figure 5.12 Frequency Distribution of Acoustic Conditions in Office Zone C1
Table 5.14 shows the summary of IEQ data in office zone C1. All the CO2 level and air temperature data were within
the recommended ASHRAE standards. The mean value of indoor air temperature was 22.83℃, and the standard
0.1%
17.6%
55.1%
23.4%
3.8%
0
200
400
600
800
1000
1200
500-549 550-599 600-649 650-699 700-750
Frequency
CO2 (ppm)
CO2
5.6%
85.0%
9.2%
0.1% 0.1%
0
200
400
600
800
1000
1200
1400
1600
60-79 80-99 100-119 120-139 180-200
Frequency
Illuminance level (lux)
Lighting
61.6%
37.7%
0.5% 0.2%
0
200
400
600
800
1000
1200
50-55 55-60 60-65 65-70
Frequency
Sound level (dBA)
Sound Level
46
deviation was 0.32, which indicates that the air temperature remained stable in office zone C1. The mean value of CO2
level is 632.21 ppm, and the standard deviation is 33.90, which means that the CO2 data fluctuated in office zone C1.
The reason may be that occupants sometimes moved around in the office or went to the conference room for meetings
during measurement hours, so the CO2 level probably went high when more occupants stayed in zone C1. Zone C1
did not have a good lighting condition; none of the lighting data were in the range of the guidelines. The mean value
was 91.62 lux, which was lower than the minimum 200 lux, and the standard deviation is 13.37, which indicates that
the lighting data fluctuated in zone C1. This is because every seat in the office had an individual desk lamp to be more
energy efficient, and the lamps were controlled by the occupant. If they did not need more lighting, the lamp did not
need to be turned on. However, when the author did the measurements in this office, the lighting from the individual
lamps was not measured. This is why the measured lighting data looks lower than the comfort zone. The acoustic data
measured in office zone C1 was entirely outside the guidelines. The mean value was 54.80 dBA, which was higher
than the maximum 40 dBA, and the standard deviation is 1.37, which means the acoustic data was stable in zone C1.
According to the table, the distributions of acoustic and lighting conditions indicate that office zone C1 was a noisy
and slightly dark indoor environment.
Table 5.14 Summary of Measured IEQ Data in Office Zone C1
IEQ factor Mean StDev Within guideline
Air temperature (℃) 22.83 0.32 100%
CO2 level (ppm) 632.21 33.90 100%
Illuminance level (lux) 91.62 13.37 0%
Acoustic (dBA) 54.80 1.37 0%
Table 5.15 below illustrates the summary of the impact of IEQ on bio-signals and the impact of bio-signals on
satisfaction in office zone C1 according to the stepwise regression results. EDA was most affected by CO2 through
four days, skin temperature was most influenced by air temperature, and stress level was most affected by lighting
condition. In the four days, thermal satisfaction and acoustic satisfaction were most affected by stress condition while
IAQ satisfaction was most affected by heart rate, and the lighting satisfaction and overall satisfaction were mostly
affected by EDA.
Table 5.15 Summary of the Impact of IEQ on Bio-Signals and of Bio-Signals on Satisfaction in Office Zone C1
Group
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Office Zone C1
Bio-Signals
Day1
Lighting
(2.19%)
Temp
(2.64%)
EDA
Day2
Lighting
(0.47%)
Acoustic
(0.71%)
CO2
(0.87%)
Day3
CO2
(2.26%)
Day4
CO2
(2.39%)
Lighting
(3.30%)
Temp
(3.80%)
Day1
Lighting
(2.62%)
CO2
(2.98%)
Skin Temp
Day2
Temp
(5.55%)
Day3
Temp
(9.65%)
Day4
CO2
(0.88%)
Lighting
(1.11%)
Day1
Acoustic
(0.83%)
Lighting
(1.05%)
Heart Rate
Day2
CO2
(0.18%)
Day3
Lighting
(0.80%)
Day4 Temp
47
(0.25%)
Day1
Stress Level
Day2
Acoustic
(0.18%)
Day3
Lighting
(0.27%)
Day4
Lighting
(1.46%)
Office Zone C1
Satisfaction/Sensation
Day1
Skin temp
(7.99%)
EDA
(11.28%)
HR
(11.45%)
Stress
(11.80%)
Thermal
Satisfaction
Day2
EDA
(42.21%)
Skin temp
(45.62%)
HR
(46.23%)
Day3
Stress
(32.30%)
Skin temp
(47.56%)
EDA
(49.17%)
HR
(49.80%)
Day4
Stress
(9.90%)
EDA
(10.97%)
Skin temp
(12.24%)
HR
(12.65%)
Day1
Skin temp
(10.01%)
HR
(15.99%)
Stress
(24.11%)
EDA
(24.30%)
IAQ Satisfaction
Day2
EDA
(8.30%)
Stress
(22.17%)
HR
(45.27%)
Skin temp
(45.77%)
Day3
HR
(34.03%)
Skin temp
(41.21%)
EDA
(45.79%)
Stress
(46.06%)
Day4
HR
(15.26%)
EDA
(20.85%)
Skin temp
(24.46%)
Day1
Skin temp
(16.05%)
HR
(19.63%)
EDA
(19.94%)
Lighting
Satisfaction
Day2
EDA
(16.63%)
Stress
(20.47%)
HR
(26.54%)
Skin temp
(29.79%)
Day3
Skin temp
(10.71%)
HR
(14.75%)
Stress
(20.40%)
Day4
EDA
(5.08%)
HR
(8.29%)
Stress
(17.18%)
Day1
Skin temp
(0.25%)
Acoustic
Satisfaction
Day2
EDA
(9.06%)
Stress
(17.06%)
HR
(39.38%)
Skin temp
(39.92%)
Day3
Stress
(19.19%)
Skin temp
(25.61%)
EDA
(28.53%)
HR
(29.23%)
Day4
Stress
(8.55%)
EDA
(13.47%)
Skin temp
(15.52%)
HR
(15.83%)
Day1
EDA
(17.17%)
Skin temp
(25.42%)
HR
(34.45%)
Stress
(36.08%)
Overall Satisfaction
Day2
EDA
(27.57%)
Skin temp
(34.29%)
Stress
(42.06%)
HR
(58.20%)
Day3
Stress
(22.24%)
Skin temp
(34.73%)
HR
(38.48%)
EDA
(39.46%)
Day4
HR
(15.54%)
EDA
(20.29%)
Skin temp
(23.16%)
Stress
(24.18%)
5.4.1 Impact by Gender
Since the data measurement in the office was carried out for eight hours per day and a total of four days, this study
performed data analysis and stepwise regression for the dataset of each day separately. This is because the datasets of
48
location A and B are less than eight hours, and it is fair to compare different datasets based on whether each dataset
has a relatively equal amount of data.
Table 5.16 shows the summary of impact of IEQ on bio-signals by different genders in office zone C1. It can be seen
that for the impact on EDA through four days, males were most affected by lighting condition while females were
most affected by air temperature followed by CO2. For the impact on skin temperature through four days, both females
and males were most affected by air temperature. As for the second influencing factors, there was no difference
between females and males. For the impact on heart rate through four days, females were most affected by CO2 while
males were most affected by lighting condition. For the impact on stress level, females were most affected by lighting
followed by air temperature while males were most affected by lighting condition and then CO2.
Table 5.16 Summary of the Impact of IEQ on Bio-Signals by Different Genders in Office Zone C1
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Day1 female
Lighting
(1.92%)
Acoustic
(2.26%)
EDA
Day1 male
Lighting
(2.54%)
Temp
(3.45%)
Day2 female
Temp
(6.10%)
Day2 male
Lighting
(1.15%)
CO2
(1.87%)
Acoustic
(2.25%)
Day3 female
Temp
(11.62%)
CO2
(13.61%)
Lighting
(14.13%)
Day3 male
Temp
(1.74%)
CO2
(4.00%)
Lighting
(4.60%)
Day4 female
CO2
(1.96%)
Lighting
(2.79%)
Day4 male
CO2
(3.04%)
Lighting
(4.12%)
Temp
(5.70%)
Day1 female
Lighting
(0.57%)
Skin Temp
Day1 male
Lighting
(5.67%)
CO2
(7.69%)
Temp
(8.09%)
Day2 female
Temp
(8.50%)
CO2
(13.16%)
Lighting
(13.71%)
Acoustic
(14.14%)
Day2 male
Temp
(3.83%)
CO2
(5.56%)
Acoustic
(5.99%)
Day3 female
Temp
(27.22%)
CO2
(29.85%)
Day3 male
Temp
(5.68%)
CO2
(6.43%)
Day4 female
CO2
(2.21%)
Acoustic
(2.77%)
Lighting
(3.11%)
Temp
(3.68%)
Day4 male
Lighting
(2.28%)
CO2
(3.20%)
Temp
(4.45%)
Acoustic
(5.01%)
Day1 female
CO2
(0.56%)
Heart Rate
Day1 male
Lighting
(1.71%)
Acoustic
(2.87%)
Day2 female
CO2
(0.35%)
Day2 male
Lighting
(1.94%)
Acoustic
(2.38%)
49
Day3 female
Lighting
(3.05%)
CO2
(3.40%)
Temp
(3.72%)
Day3 male
CO2
(0.88%)
Day4 female
Temp
(0.71%)
Day4 male
Day1 female
Stress Level
Day1 male
Lighting
(0.53%)
Day2 female
Day2 male
Acoustic
(1.17%)
Day3 female
Lighting
(0.56%)
Day3 male
Day4 female
Temp
(1.58%)
Lighting
(2.16%)
CO2
(2.54%)
Day4 male
CO2
(7.07%)
Temp
(8.14%)
Lighting
(9.64%)
Table 5.17 shows the summary of the impact of bio-signals on satisfaction by different genders in office zone C1. It
can be seen that for the impact on thermal satisfaction through four days, females were most affected by stress level
followed by heart rate while males were most affected by EDA and then heart rate. For the impact on IAQ satisfaction
through four days, females tended to be most affected by skin temperature while males were most affected by EDA.
For the impact on lighting satisfaction, females were most affected by EDA, and males were most affected by heart
rate. For the impact on acoustic satisfaction, females were most likely affected by stress condition while males were
most affected by skin temperature. For the impact on overall satisfaction in all the four days, both females and males
were mostly affected by EDA.
Table 5.17 Summary of the Impact of Bio-Signals on Satisfaction by Different Genders in Office Zone C1
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Day1 female
Skin temp
(22.41%)
Thermal Satisfaction
Day1 male
EDA
(6.01%)
HR
(7.08%)
Day2 female
EDA
(15.79%)
HR
(20.53%)
Skin temp
(21.97%)
Stress
(22.56%)
Day2 male
EDA
(44.84%)
Skin temp
(48.46%)
Stress
(51.09%)
HR
(52.35%)
Day3 female
Stress
(49.21%)
HR
(52.00%)
Skin temp
(52.60%)
Day3 male
Skin temp
(50.78%)
HR
(53.81%)
EDA
(55.55%)
Day4 female
Stress
(20.70%)
Skin temp
(25.26%)
HR
(25.73%)
Day4 male
Stress
(2.85%)
HR
(5.41%)
EDA
(6.09%)
Day1 female
Skin temp
(20.11%)
HR
(26.08%)
Stress
(33.44%)
IAQ Satisfaction
Day1 male
Skin temp
(16.19%)
HR
(23.53%)
Stress
(24.67%)
50
Day2 female
Skin temp
(22.25%)
Stress
(38.12%)
HR
(46.88%)
EDA
(47.34%)
Day2 male
Day3 female
HR
(52.56%)
EDA
(58.22%)
Stress
(60.49%)
Skin temp
(61.97%)
Day3 male
HR
(25.20%)
Skin temp
(33.41%)
Stress
(35.21%)
EDA
(36.73%)
Day4 female
Stress
(14.98%)
EDA
(37.60%)
Skin temp
(38.97%)
Day4 male
HR
(26.00%)
Stress
(31.51%)
Skin temp
(34.73%)
EDA
(36.00%)
Day1 female
Skin temp
(18.77%)
Stress
(23.88%)
HR
(25.72%)
EDA
(26.69%)
Lighting Satisfaction
Day1 male
HR
(38.44%)
Stress
(49.25%)
Skin temp
(52.15%)
EDA
(53.55%)
Day2 female
Stress
(30.60%)
EDA
(38.04%)
Skin temp
(40.09%)
Day2 male
Skin temp
(40.20%)
HR
(44.54%)
EDA
(45.70%)
Day3 female
Skin temp
(3.80%)
EDA
(7.27%)
HR
(14.30%)
Stress
(27.62%)
Day3 male
HR
(26.67%)
Skin temp
(36.57%)
EDA
(39.58%)
Stress
(41.06%)
Day4 female
EDA
(43.23%)
Skin temp
(57.57%)
Stress
(58.66%)
HR
(58.93%)
Day4 male
HR
(41.12%)
Stress
(59.00%)
Skin temp
(62.70%)
Day1 female
Skin temp
(2.16%)
EDA
(3.96%)
Acoustic Satisfaction
Day1 male
HR
(15.40%)
Stress
(29.22%)
Skin temp
(31.42%)
Day2 female
EDA
(18.04%)
HR
(26.76%)
Stress
(40.21%)
Skin temp
(40.74%)
Day2 male
Skin temp
(23.61%)
EDA
(28.07%)
HR
(28.52%)
Day3 female
HR
(51.57%)
EDA
(54.98%)
Skin temp
(58.27%)
Stress
(59.15%)
Day3 male
Day4 female
Stress
(14.43%)
EDA
(38.51%)
HR
(39.06%)
Day4 male
Day1 female
Skin temp
(53.87%)
HR
(64.25%)
Stress
(69.45%)
EDA
(69.57%)
Overall Satisfaction
Day1 male
EDA
(67.48%)
HR
(72.99%)
Skin temp
(76.18%)
Stress
(77.33%)
Day2 female
Stress
(61.75%)
EDA
(69.67%)
HR
(70.00%)
Skin temp
(70.38%)
Day2 male
Skin temp
(40.04%)
HR
(44.04%)
EDA
(45.44%)
Day3 female
Stress
(43.25%)
HR
(81.12%)
EDA
(81.81%)
Skin temp
(81.94%)
Day3 male
Skin temp
(26.98%)
HR
(37.11%)
EDA
(40.25%)
Stress
(42.14%)
51
Day4 female
Stress
(21.99%)
EDA
(48.74%)
HR
(49.31%)
Skin temp
(49.72%)
Day4 male
HR
(43.72%)
Stress
(61.13%)
Skin temp
(64.50%)
5.4.2 Impact by Age
Table 5.18 illustrates the summary of the impact of IEQ on bio-signals by different age groups in office zone C1. This
table indicates that for the impact on EDA through four days, the mid-age group was most influenced by air
temperature followed by lighting while the junior group was most affected by lighting condition and then CO2. For
the impact on skin temperature through four days, both the junior and mid-age groups were most affected by air
temperature, followed by acoustics and CO2. For the impact on heart rate through four days, both the junior and mid-
age groups were most affected by air temperature and then lighting condition. Also, for the impact on stress level
through four days, both the junior and mid-age groups were most influenced by air temperature. From the overall table,
both the junior and mid-age groups were mostly affected by air temperature on the four measured bio-signals. The
reason is probably that bio-signals did not vary much during the measurement, and temperature is also the most stable
factor among measured IEQ factors.
Table 5.18 Summary of the Impact of IEQ on Bio-Signals by Different Age Groups in Office Zone C1
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Day1 junior
Lighting
(3.28%)
Temp
(3.81%)
EDA
Day1 mid-age
Lighting
(14.69%)
Temp
(20.46%)
Day2 junior
Lighting
(0.67%)
CO2
(1.20%)
Acoustic
(1.57%)
Day2 mid-age
Temp
(15.73%)
CO2
(17.86%)
Day3 junior
Temp
(10.34%)
CO2
(15.60%)
Lighting
(16.27%)
Day3 mid-age
Temp
(6.91%)
Lighting
(7.88%)
CO2
(9.72%)
Day4 junior
CO2
(1.96%)
Lighting
(2.79%)
Day4 mid-age
Temp
(2.62%)
Lighting
(4.78%)
CO2
(5.31%)
Day1 junior
Temp
(0.19%)
Lighting
(0.36%)
Acoustic
(0.48%)
Skin Temp
Day1 mid-age
Lighting
(13.30%)
CO2
(14.44%)
Day2 junior
Temp
(2.14%)
Acoustic
(2.64%)
Day2 mid-age
Temp
(18.62%)
CO2
(19.57%)
Lighting
(19.94%)
Day3 junior
Temp
(3.96%)
Day3 mid-age
Temp
(51.32%)
CO2
(52.74%)
Lighting
(52.97%)
Day4 junior
CO2
(2.21%)
Acoustic
(2.77%)
Lighting
(3.11%)
Temp
(3.68%)
Day4 mid-age
Temp
(3.67%)
Lighting
(5.59%)
Acoustic
(6.10%)
Day1 junior Temp Lighting
Heart Rate
52
(1.68%) (1.95%)
Day1 mid-age
Acoustic
(1.48%)
Lighting
(3.07%)
Day2 junior
Lighting
(0.59%)
Day2 mid-age
Temp
(1.01%)
CO2
(1.70%)
Day3 junior
Temp
(3.19%)
Day3 mid-age
Temp
(2.72%)
Lighting
(6.23%)
Day4 junior
Temp
(0.71%)
Day4 mid-age
CO2
(1.62%)
Temp
(2.82%)
Day1 junior
Stress Level
Day1 mid-age
Acoustic
(1.61%)
CO2
(3.24%)
Lighting
(3.67%)
Day2 junior
Lighting
(0.53%)
Acoustic
(0.98%)
Temp
(1.32%)
Day2 mid-age
Lighting
(2.73%)
CO2
(4.85%)
Temp
(6.38%)
Day3 junior
Temp
(2.49%)
CO2
(2.82%)
Day3 mid-age
Temp
(7.38%)
Lighting
(9.34%)
Day4 junior
Temp
(1.58%)
Lighting
(2.16%)
CO2
(2.54%)
Day4 mid-age
Temp
(6.26%)
Table 5.19 shows the summary of the impact of bio-signals on satisfaction by different age groups in office zone C1.
It can be seen from the table that for the impact on thermal satisfaction through four days, the junior and mid-age
groups were most affected by skin temperature. For the impact on IAQ satisfaction through four days, the junior and
mid-age groups were most influenced by skin temperature followed by EDA. For the impact on lighting satisfaction
through four days, the junior and mid-age groups were most affected by EDA, followed by skin temperature and stress
level. Also, for the impact on acoustic satisfaction through four days, the junior and mid-age groups were most affected
by skin temperature and then EDA. Thus, from the overall table, except lighting satisfaction, both junior and mid-age
groups were mostly affected by skin temperature and then EDA in all the satisfaction surveys.
Table 5.19 Summary of the Impact of Bio-Signals on Satisfaction by Different Age Groups in Office Zone C1
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Day1 junior
Skin temp
(11.57%)
EDA
(15.66%)
HR
(17.42%)
Thermal Satisfaction
Day1 mid-age
Skin temp
(2.46%)
EDA
(4.30%)
Day2 junior
EDA
(61.45%)
HR
(62.23%)
Stress
(62.35%)
Day2 mid-age
Day3 junior
Skin temp
(34.90%)
Stress
(44.59%)
HR
(44.83%)
Day3 mid-age HR EDA Skin temp Stress
53
(45.60%) (57.95%) (61.90%) (62.47%)
Day4 junior
Stress
(20.70%)
Skin temp
(25.26%)
HR
(25.73%)
Day4 mid-age
Skin temp
(6.36%)
EDA
(8.13%)
Day1 junior
Skin temp
(15.55%)
EDA
(17.93%)
Stress
(20.57%)
HR
(30.18%)
IAQ Satisfaction
Day1 mid-age
Skin temp
(29.52%)
HR
(41.72%)
EDA
(42.41%)
Day2 junior
EDA
(12.87%)
Stress
(30.52%)
HR
(52.71%)
Day2 mid-age
Day3 junior
Skin temp
(27.65%)
EDA
(48.25%)
HR
(54.06%)
Stress
(55.90%)
Day3 mid-age
Stress
(65.43%)
Skin temp
(66.91%)
HR
(68.59%)
EDA
(69.13%)
Day4 junior
Stress
(14.98%)
EDA
(37.60%)
Skin temp
(38.97%)
Day4 mid-age
HR
(33.72%)
Skin temp
(42.19%)
EDA
(45.84%)
Day1 junior
Stress
(28.03%)
Skin temp
(31.78%)
HR
(32.72%)
EDA
(33.10%)
Lighting Satisfaction
Day1 mid-age
EDA
(16.83%)
HR
(25.13%)
Skin temp
(26.58%)
Day2 junior
EDA
(12.10%)
Stress
(22.69%)
HR
(46.38%)
Skin temp
(48.41%)
Day2 mid-age
EDA
(41.37%)
Stress
(49.09%)
Skin temp
(50.84%)
Day3 junior
HR
(20.23%)
Skin temp
(25.71%)
EDA
(29.30%)
Stress
(30.10%)
Day3 mid-age
Stress
(27.20%)
Skin temp
(34.57%)
EDA
(35.83%)
HR
(36.83%)
Day4 junior
EDA
(43.23%)
Skin temp
(57.57%)
Stress
(58.66%)
HR
(58.93%)
Day4 mid-age
HR
(28.76%)
Stress
(35.77%)
Skin temp
(36.87%)
Day1 junior
HR
(6.59%)
EDA
(8.37%)
Skin temp
(9.18%)
Stress
(9.99%)
Acoustic Satisfaction
Day1 mid-age
Skin temp
(3.68%)
EDA
(4.68%)
Day2 junior
EDA
(12.00%)
Stress
(25.46%)
HR
(49.19%)
Skin temp
(50.07%)
Day2 mid-age
Skin temp
(15.14%)
Stress
(27.22%)
Day3 junior
Skin temp
(31.77%)
EDA
(51.56%)
Stress
(55.72%)
HR
(56.15%)
Day3 mid-age
Skin temp
(2.54%)
EDA
(6.58%)
Day4 junior
Stress
(14.43%)
EDA
(38.51%)
HR
(39.06%)
Day4 mid-age
Skin temp
(3.32%)
Day1 junior EDA Skin temp HR Stress Overall Satisfaction
54
(20.66%) (29.87%) (41.04%) (45.45%)
Day1 mid-age
Day2 junior
EDA
(23.94%)
Stress
(32.56%)
HR
(55.68%)
Skin temp
(62.42%)
Day2 mid-age
EDA
(69.22%)
Stress
(78.65%)
Day3 junior
Stress
(38.66%)
HR
(54.90%)
Skin temp
(60.39%)
EDA
(64.95%)
Day3 mid-age
Stress
(69.21%)
EDA
(70.68%)
Skin temp
(71.88%)
HR
(73.22%)
Day4 junior
Stress
(21.99%)
EDA
(48.74%)
HR
(49.31%)
Skin temp
(49.72%)
Day4 mid-age
HR
(77.83%)
Stress
(84.25%)
Skin temp
(86.34%)
5.5 Location C: Office Zone C2
This section discusses the analysis of zone C2. The frequency distribution of indoor air temperature is shown in Figure
5.13. According to the comfortable range for occupants based on the recommended guidelines, all measured
temperature values in office zone C2 were in the human comfortable range. In Figure 5.14, less than half of the CO2
condition was in the range that made occupants comfortable. Most of them were higher than the maximum standard
1000 ppm. Figure 5.15 shows the frequency distribution of the lighting condition in office zone C2. More than half of
the lighting data was outside the comfort zone according to the guidelines. Most of the measured illuminance data
were higher than maximum standard 500 lux. In Figure 5.16, the whole acoustic data measured in office zone C2 was
outside the comfort range, which is a maximum of 40 dBA. The data ranged from 45 to 70 dBA, which is higher than
the standard.
Figure 5.13 Frequency Distribution of Indoor Air Temperature in Office Zone C2
34.7%
37.9%
17.6%
6.5%
3.3%
0
100
200
300
400
500
600
700
800
21-22 22-23 23-24 24-25 25-26
Frequency
Temperature (℃)
Air Temperature
55
Figure 5.14 Frequency Distribution of CO2 in Office Zone C2
Figure 5.15 Frequency Distribution of Lighting Condition in Office Zone C2
Figure 5.16 Frequency Distribution of Acoustic Condition in Office Zone C2
Table 5.20 shows the summary of the measured IEQ data in office zone C2. All the air temperature data were within
the recommended ASHRAE standard. The mean value of indoor air temperature was 22.53℃, and the standard
15.2%
51.2%
30.8%
2.8%
0
100
200
300
400
500
600
700
800
900
1000
900-999 1000-1099 1100-1199 1200-1300
Frequency
CO2 (ppm)
CO2
46.3%
31.6%
11.1%
3.6%
1.9%
5.6%
0
200
400
600
800
1000
0-99 100-199 200-299 300-399 400-500 >500
Frequency
Illuminance level (lux)
Lighting
0.1%
77.6%
18.4%
3.6%
0.3%
0
200
400
600
800
1000
1200
1400
1600
45-50 50-55 55-60 60-65 65-70
Frequency
Sound level (dBA)
Sound Level
56
deviation was 1.01, which indicates that the air temperature remained stable in office zone C2. The CO2 data had only
15% in the comfort zone. The mean value of the CO2 level was 1071.13 ppm, and the standard deviation was 66.82,
which means that the CO2 data fluctuated in office zone C2. The reason is probably that zone C2 is close to a big
window wall, and the CO2 from the outdoor air, which was not in the comfort zone, influenced indoor air quality.
Zone C2 did not have good lighting condition; only 17% of the lighting data was within the range of the guidelines.
The mean value was 172.20 lux, which was lower than the minimum of 200 lux, and the standard deviation was 233.59,
which indicates that the lighting data fluctuated significantly in zone C2. The values were lower than the standard
because the lighting from individual lamps was not measured, and values higher than the standard were due to the
high illuminance level at noon since it was near the big window. The acoustic data measured in office zone C2 was
entirely outside the guidelines. The mean value was 54.02 dBA, which was higher than the maximum 40 dBA, and
the standard deviation was 2.43 which means the acoustic data was stable in zone C2. According to the table, the
distributions of acoustics, CO2 level, and lighting condition indicate that occupants in zone C2 were in a noisy, low-
air-quality, and slightly dark indoor environment.
Table 5.20 Summary of Measured IEQ Data in Office Zone C2
IEQ factor Mean StDev Within guideline
Air temperature (℃) 22.53 1.01 100%
CO2 level (ppm) 1071.13 66.82 15%
Illuminance level (lux) 172.20 233.59 17%
Acoustic (dBA) 54.02 2.43 0%
Table 5.21 below summarizes the impact of IEQ on bio-signals and the impact of bio-signals on satisfaction in office
zone C2. Through the four days, EDA was most affected by CO2, skin temperature was mostly affected by air
temperature, and heart rate and stress level were mostly influenced by lighting condition, which is similar to office
zone C1. Thermal satisfaction, acoustic satisfaction, and overall satisfaction were most affected by skin temperature.
Lighting and IAQ satisfaction were most affected by EDA.
Table 5.21 Summary of the Impact of IEQ on Bio-Signals and of Bio-Signals on Satisfaction in Office Zone C2
Group
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Office Zone C2
Bio-Signals
Day1
Acoustic
(1.43%)
Temp
(2.30%)
Lighting
(2.60%)
EDA
Day2
Temp
(0.52%)
Lighting
(0.92%)
Day3
CO2
(0.86%)
Day4
CO2
(2.85%)
Day1
Lighting
(2.35%)
CO2
(3.24%)
Skin Temp
Day2
Temp
(6.46%)
Lighting
(8.30%)
CO2
(8.45%)
Day3
Temp
(10.59%)
Acoustic
(10.80%)
Lighting
(10.94%)
CO2
(11.10%)
Day4
CO2
(5.83%)
Temp
(6.18%)
Acoustic
(6.37%)
Day1
Lighting
(4.84%)
CO2
(5.87%)
Temp
(7.11%)
Heart Rate
Day2
Lighting
(1.87%)
CO2
(2.42%)
Day3
Lighting
(1.56%)
Temp
(1.79%)
Acoustic
(2.00%)
Day4
CO2
(3.08%)
Temp
(3.44%)
Day1 CO2 Temp Acoustic Stress Level
57
(4.63%) (5.59%) (5.85%)
Day2
Lighting
(1.11%)
Temp
(1.65%)
Day3
Lighting
(1.71%)
Acoustic
(2.19%)
Temp
(2.62%)
Day4
Lighting
(1.26%)
Temp
(1.76%)
Office Zone C2
Satisfaction/Sensation
Day1
Skin temp
(13.98%)
Stress
(14.64%)
Thermal Satisfaction
Day2
Skin temp
(0.99%)
EDA
(1.40%)
Day3
EDA
(14.19%)
Stress
(18.32%)
Skin temp
(23.35%)
Day4
HR
(3.50%)
Skin temp
(4.94%)
Stress
(5.97%)
EDA
(6.82%)
Day1
Stress
(44.18%)
HR
(46.59%)
EDA
(46.85%)
IAQ Satisfaction
Day2
Skin temp
(29.43%)
HR
(38.45%)
Stress
(43.26%)
Day3
EDA
(58.90%)
Skin temp
(60.34%)
Day4
EDA
(20.18%)
HR
(23.61%)
Skin temp
(24.64%)
Stress
(25.33%)
Day1
Stress
(12.29%)
HR
(12.80%)
Skin temp
(13.72%)
Lighting Satisfaction
Day2
Skin temp
(43.61%)
HR
(47.14%)
Stress
(53.91%)
Day3
EDA
(25.00%)
Skin temp
(27.69%)
HR
(28.85%)
Stress
(32.67%)
Day4
EDA
(18.51%)
HR
(19.71%)
Skin temp
(20.60%)
Stress
(21.19%)
Day1
Skin temp
(1.97%)
Stress
(26.17%)
HR
(28.14%)
Acoustic Satisfaction
Day2
Skin temp
(48.10%)
HR
(52.52%)
Stress
(60.03%)
Day3
EDA
(64.31%)
Skin temp
(71.68%)
Stress
(72.35%)
HR
(72.48%)
Day4
EDA
(17.81%)
HR
(20.82%)
Skin temp
(21.79%)
Stress
(22.37%)
Day1
Skin temp
(13.75%)
Stress
(30.28%)
Overall Satisfaction
Day2
Skin temp
(21.47%)
Stress
(26.18%)
Day3
Day4
EDA
(18.86%)
Stress
(20.80%)
5.5.1 Impact by Gender
Table 5.22 shows the summary of the impact of IEQ on bio-signals by different genders in office zone C2. For the
impact on EDA through four days, both females and males were most affected by air temperature. For the impact on
skin temperature through four days, females and males were most influenced by CO2. For the impact on heart rate
through four days, both females and males were most affected by air temperature, then CO2, then lighting condition.
58
For the impact on stress level through four days, females were most influenced by lighting condition while males were
most affected by CO2 level. From the overall table, both females and males were mostly affected by CO2 on the four
bio-signals, but air temperature and lighting condition also had an effect on them.
Table 5.22 Summary of the Impact of IEQ on Bio-Signals by Different Genders in Office Zone C2
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Day1 female
Acoustic
(2.30%)
Temp
(3.61%)
Lighting
(4.00%)
EDA
Day1 male
Temp
(1.07%)
Day2 female
Temp
(1.25%)
Lighting
(1.72%)
Acoustic
(2.20%)
Day2 male
Temp
(5.89%)
Lighting
(6.77%)
Day3 female
Temp
(0.49%)
CO2
(0.87%)
Day3 male
CO2
(1.66%)
Day4 female
Lighting
(3.57%)
CO2
(4.85%)
Temp
(5.86%)
Day4 male
CO2
(3.46%)
Lighting
(6.83%)
Temp
(8.55%)
Day1 female
Lighting
(4.50%)
Acoustic
(4.66%)
Skin Temp
Day1 male
CO2
(6.72%)
Lighting
(10.71%)
Day2 female
Temp
(8.36%)
Lighting
(10.07%)
CO2
(10.33%)
Day2 male
Temp
(8.92%)
Lighting
(15.71%)
Day3 female
CO2
(4.38%)
Temp
(7.64%)
Acoustic
(7.98%)
Day3 male
Temp
(51.94%)
CO2
(63.88%)
Lighting
(64.71%)
Day4 female
CO2
(1.42%)
Day4 male
CO2
(24.39%)
Temp
(30.32%)
Lighting
(30.74%)
Day1 female
Lighting
(6.03%)
Temp
(6.76%)
CO2
(7.35%)
Heart Rate
Day1 male
CO2
(4.68%)
Temp
(8.06%)
Acoustic
(9.05%)
Day2 female
Lighting
(3.02%)
Temp
(3.39%)
Day2 male
CO2
(1.45%)
Day3 female
CO2
(0.90%)
Temp
(1.59%)
Day3 male
Lighting
(3.48%)
CO2
(5.10%)
Acoustic
(5.75%)
Day4 female
Temp
(1.40%)
Day4 male CO2 Lighting
59
(8.04%) (8.34%)
Day1 female
Lighting
(9.54%)
Stress Level
Day1 male
CO2
(5.73%)
Temp
(8.40%)
Acoustic
(10.13%)
Day2 female
Lighting
(1.05%)
Acoustic
(1.97%)
Day2 male
Lighting
(3.33%)
Temp
(7.42%)
Day3 female
CO2
(1.04%)
Temp
(1.87%)
Day3 male
Lighting
(2.74%)
CO2
(6.87%)
Acoustic
(8.70%)
Day4 female
Lighting
(1.33%)
Day4 male
Temp
(4.39%)
CO2
(5.14%)
Table 5.23 shows the summary of the impact of bio-signals on satisfaction by different genders in office zone C2.
According to the table, for the impact on thermal satisfaction through four days, females were most affected by skin
temperature while males were most affected by heart rate. As for the impact on IAQ satisfaction through four days,
males were most affected by heart rate, followed by skin temperature, followed by stress condition while females had
no correlation between their IAQ satisfaction with bio-signals in the dataset of zone C2. For the impact on lighting
satisfaction through four days, both females and males were most affected by EDA. For the impact on acoustic
satisfaction through four days, females were most affected by skin temperature while males were most affected by
EDA and then heart rate. For the impact on overall satisfaction through four days, both females and males were most
affected by skin temperature.
Table 5.23 Summary of the Impact of Bio-Signals on Satisfaction by Different Genders in Office Zone C2
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Day1 female
Skin temp
(3.61%)
EDA
(4.95%)
HR
(6.08%)
Stress
(7.44%)
Thermal Satisfaction
Day1 male
HR
(10.71%)
EDA
(18.78%)
Stress
(19.72%)
Skin temp
(20.40%)
Day2 female
Skin temp
(5.69%)
EDA
(8.20%)
HR
(9.49%)
Stress
(14.16%)
Day2 male
Skin temp
(5.10%)
Day3 female
Stress
(7.59%)
EDA
(10.77%)
Day3 male
Stress
(30.55%)
HR
(37.98%)
Skin temp
(39.04%)
Day4 female
Stress
(2.64%)
HR
(4.91%)
EDA
(6.20%)
Day4 male
HR
(27.97%)
Skin temp
(35.82%)
Stress
(36.97%)
EDA
(38.07%)
Day1 female
IAQ Satisfaction
Day1 male
Stress
(48.12%)
Skin temp
(53.70%)
HR
(55.31%)
Day2 female
Day2 male
HR
(30.75%)
Skin temp
(33.68%)
Stress
(34.87%)
EDA
(35.90%)
60
Day3 female
Day3 male
EDA
(69.74%)
Skin temp
(77.16%)
Stress
(78.37%)
HR
(81.46%)
Day4 female
Day4 male
HR
(44.54%)
Skin temp
(48.84%)
EDA
(53.78%)
Stress
(56.62%)
Day1 female
Skin temp
(8.58%)
Lighting Satisfaction
Day1 male
EDA
(22.11%)
Skin temp
(35.71%)
Day2 female
Day2 male
EDA
(22.05%)
HR
(24.49%)
Day3 female
Stress
(2.35%)
HR
(8.94%)
EDA
(11.54%)
Day3 male
EDA
(55.22%)
HR
(57.23%)
Stress
(61.58%)
Day4 female
EDA
(7.92%)
Skin temp
(14.52%)
HR
(14.82%)
Stress
(15.76%)
Day4 male
HR
(30.96%)
EDA
(37.22%)
Skin temp
(42.15%)
Stress
(44.87%)
Day1 female
Skin temp
(1.09%)
Acoustic Satisfaction
Day1 male
EDA
(12.11%)
HR
(12.86%)
Day2 female
Day2 male
EDA
(28.17%)
HR
(35.01%)
Skin temp
(37.70%)
Day3 female
Day3 male
EDA
(77.90%)
Stress
(80.98%)
HR
(85.48%)
Skin temp
(86.40%)
Day4 female
Day4 male
HR
(37.18%)
Skin temp
(40.97%)
EDA
(44.80%)
Stress
(47.59%)
Day1 female
HR
(14.46%)
Skin temp
(24.58%)
Stress
(25.31%)
Overall Satisfaction
Day1 male
Day2 female
Skin temp
(24.68%)
Stress
(41.87%)
HR
(43.33%)
EDA
(43.65%)
Day2 male
Skin temp
(18.08%)
HR
(29.92%)
EDA
(30.79%)
Day3 female
Day3 male
Day4 female
Skin temp
(6.67%)
EDA
(8.92%)
Stress
(9.97%)
HR
(15.64%)
Day4 male
HR
(44.61%)
Skin temp
(50.74%)
EDA
(55.61%)
Stress
(58.06%)
5.5.2 Impact by Age
61
Table 5.24 illustrates the summary of the impact of IEQ on bio-signals by different age groups in office zone C2. The
table indicates that for the impact on EDA through four days, both the junior and mid-age groups were most affected
by CO2 level. For the impact on skin temperature through four days, both the junior and mid-age groups were most
affected by air temperature. Also, for the impact on heart rate through four days, the junior group was most influenced
by lighting condition while the mid-age group was mostly affected by CO2. For the impact on stress level through four
days, both the junior and mid-age groups were most affected by lighting condition. From the overall table, it can be
seen that both the junior and mid-age groups were most affected by CO2 and lighting condition while air temperature
also had an effect on them.
Table 5.24 Summary of the Impact of IEQ on Bio-Signals by Different Age Groups in Office Zone C2
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Day1 junior
Acoustic
(1.71%)
Temp
(2.74%)
EDA
Day1 mid-age
Acoustic
(2.72%)
Temp
(5.76%)
CO2
(7.36%)
Lighting
(9.65%)
Day2 junior
Temp
(1.13%)
Lighting
(1.75%)
Day2 mid-age
Temp
(49.52%)
Lighting
(52.40%)
Day3 junior
CO2
(1.88%)
Temp
(2.11%)
Lighting
(2.36%)
Day3 mid-age
CO2
(26.90%)
Temp
(28.51%)
Day4 junior
CO2
(3.84%)
Day4 mid-age
CO2
(3.45%)
Lighting
(4.43%)
Temp
(6.31%)
Day1 junior
Lighting
(4.96%)
CO2
(5.89%)
Acoustic
(6.29%)
Skin Temp
Day1 mid-age
CO2
(4.37%)
Acoustic
(7.50%)
Day2 junior
Temp
(4.30%)
Lighting
(5.90%)
Day2 mid-age
Temp
(72.73%)
Lighting
(81.81%)
Acoustic
(82.11%)
Day3 junior
Temp
(9.91%)
CO2
(11.36%)
Acoustic
(11.59%)
Lighting
(11.79%)
Day3 mid-age
Temp
(30.06%)
CO2
(60.30%)
Lighting
(61.27%)
Day4 junior
CO2
(2.79%)
Day4 mid-age
Temp
(43.92%)
CO2
(49.98%)
Lighting
(52.36%)
Day1 junior
Temp
(6.34%)
CO2
(8.88%)
Heart Rate
Day1 mid-age
Lighting
(4.69%)
Temp
(8.19%)
Day2 junior
Lighting
(3.23%)
CO2
(3.50%)
Acoustic
(3.71%)
Day2 mid-age
Temp
(7.38%)
Lighting
(8.08%)
Day3 junior
Lighting
(1.93%)
Temp
(2.43%)
CO2
(2.80%)
Acoustic
(3.08%)
62
Day3 mid-age
CO2
(2.66%)
Lighting
(4.14%)
Day4 junior
CO2
(2.34%)
Temp
(2.82%)
Day4 mid-age
CO2
(7.14%)
Acoustic
(8.09%)
Day1 junior
Temp
(5.12%)
CO2
(7.77%)
Acoustic
(8.20%)
Stress Level
Day1 mid-age
Lighting
(4.70%)
Temp
(10.78%)
Day2 junior
Lighting
(1.25%)
Acoustic
(1.54%)
Day2 mid-age
Temp
(13.69%)
Day3 junior
Lighting
(2.60%)
Acoustic
(3.57%)
Temp
(4.29%)
Day3 mid-age
CO2
(2.91%)
Lighting
(7.88%)
Day4 junior
Temp
(1.06%)
Day4 mid-age
Lighting
(7.16%)
Table 5.25 shows the summary of the impact of bio-signals on satisfaction by different age groups in office zone C2.
The table indicates that for the impact on thermal satisfaction through four days, the junior group was most affected
by EDA and skin temperature while the mid-age group was mostly influenced by EDA and stress condition. For the
impact on IAQ satisfaction through four days, both the junior and mid-age groups were most affected by EDA
followed by heart rate. As for the impact on lighting satisfaction and acoustic satisfaction through four days, both the
junior and mid-age groups were most affected by EDA followed by skin temperature. For the impact on overall IEQ
satisfaction, there are too many blanks in the table, which means Minitab did not find any correlation between the
selected factors. From the overall table, both the junior and mid-age groups tended to be affected by EDA mostly with
regard to thermal conditions, IAQ, lighting, and acoustic satisfaction.
Table 5.25 Summary of the Impact of Bio-Signals on Satisfaction by Different Age Groups in Office Zone C2
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Day1 junior
Skin temp
(18.01%)
Stress
(19.52%)
EDA
(19.87%)
Thermal Satisfaction
Day1 mid-age
Stress
(6.09%)
EDA
(8.05%)
Skin temp
(9.54%)
Day2 junior
EDA
(0.91%)
Day2 mid-age
EDA
(12.71%)
Skin temp
(35.81%)
Day3 junior
EDA
(15.91%)
Stress
(23.58%)
Skin temp
(26.42%)
HR
(26.92%)
Day3 mid-age
Stress
(12.03%)
EDA
(20.83%)
HR
(22.02%)
Day4 junior
Skin temp
(2.48%)
EDA
(3.52%)
HR
(3.81%)
Stress
(4.55%)
Day4 mid-age
EDA
(5.50%)
Day1 junior
Stress
(51.16%)
HR
(54.46%)
EDA
(54.80%)
IAQ Satisfaction
63
Day1 mid-age
EDA
(11.16%)
Day2 junior
Skin temp
(34.08%)
HR
(42.84%)
Stress
(48.97%)
Day2 mid-age
Day3 junior
EDA
(58.78%)
Skin temp
(61.38%)
Stress
(61.50%)
Day3 mid-age
HR
(13.15%)
EDA
(18.52%)
Skin temp
(19.97%)
Day4 junior
EDA
(18.07%)
Stress
(21.39%)
Skin temp
(23.29%)
Day4 mid-age
Day1 junior
Stress
(18.01%)
HR
(18.81%)
Skin temp
(19.33%)
Lighting Satisfaction
Day1 mid-age
Stress
(4.70%)
EDA
(8.95%)
Day2 junior
Skin temp
(48.55%)
HR
(51.97%)
Stress
(59.44%)
Day2 mid-age
EDA
(10.17%)
Skin temp
(32.99%)
Day3 junior
EDA
(26.40%)
Skin temp
(32.63%)
HR
(33.58%)
Stress
(38.83%)
Day3 mid-age
EDA
(36.36%)
HR
(38.90%)
Skin temp
(41.53%)
Day4 junior
EDA
(16.10%)
Skin temp
(16.95%)
HR
(17.44%)
Day4 mid-age
Day1 junior
Stress
(23.00%)
Skin temp
(40.09%)
HR
(42.91%)
Acoustic Satisfaction
Day1 mid-age
Stress
(9.59%)
EDA
(17.09%)
Skin temp
(20.86%)
Day2 junior
Skin temp
(51.27%)
HR
(56.03%)
Stress
(63.64%)
EDA
(63.83%)
Day2 mid-age
EDA
(24.63%)
Skin temp
(31.48%)
Day3 junior
EDA
(62.72%)
Skin temp
(77.11%)
Day3 mid-age
Day4 junior
EDA
(15.78%)
Stress
(18.59%)
Skin temp
(20.35%)
Day4 mid-age
Day1 junior
Stress
(28.55%)
Skin temp
(32.79%)
EDA
(33.30%)
Overall Satisfaction
Day1 mid-age
Day2 junior
Skin temp
(32.93%)
Stress
(43.75%)
Day2 mid-age
Day3 junior
Day3 mid-age
Day4 junior
EDA
(18.98%)
Stress
(21.73%)
64
Day4 mid-age
5.6 Comparison of Different Locations
Table 5.26 below shows the summary of measured IEQ data in all locations. The mean values of air temperature in
office zones C1 and C2 were lower than the temperature setting in the studio and the classroom. Although the
temperature data was different among these locations, the standard deviations of each dataset were small, which means
the temperature in each location remained stable, and all the temperature data in four locations were in the range of
comfort zone. The mean values of CO2 in the classroom and in office zone C2 were higher than those in the studio
and in office zone C1, and the standard deviation in office zone C2 was the highest. Except for office zone C2, the
CO2 data in other three locations were in the comfort zone guidelines. This is because the studio and office were open
environments, and the occupants moved around instead of always staying in their seats like the occupants in the
classroom. The classroom was also at a basement without a window, so the CO2 concentration would be high. While
office zone C2 was very close to a big window, the indoor air quality would be influenced by outdoor air. The mean
value of illuminance levels in all four locations were below minimum 200 lux according to the standard. This is
because both the studio and office provided individual lamps for each seat, which were not measured by the lighting
sensor, and classroom needed a slightly dark environment for the projector to work better. The standard deviations of
lighting data in the studio and office zone C2 were much higher than those in the classroom and in office zone C1
because the measured locations in the studio and in office zone C2 were near windows, and the illuminance level went
high at noon. The mean values and standard deviations of acoustic data in all four locations have no significant
difference, and they were all outside the comfort zone. This is because the four locations were not quiet, and talking
was allowed.
Table 5.26 Summary of Measured IEQ Data in All Locations
IEQ factor Mean StDev Within guideline
Studio
Air temperature (℃) 23.58 0.22 100%
CO2 level (ppm) 505.88 30.70 100%
Illuminance level (lux) 139.93 97.84 44%
Acoustic (dBA) 55.44 2.66 0%
Classroom
Air temperature (℃) 24.32 0.40 100%
CO2 level (ppm) 903.18 22.03 100%
Illuminance level (lux) 127.60 3.80 0%
Acoustic (dBA) 53.92 3.91 0%
Office
Zone C1
Air temperature (℃) 22.83 0.32 100%
CO2 level (ppm) 632.21 33.90 100%
Illuminance level (lux) 91.62 13.37 0%
Acoustic (dBA) 54.80 1.37 0%
Office
Zone C2
Air temperature (℃) 22.53 1.01 100%
CO2 level (ppm) 1071.13 66.82 15%
Illuminance level (lux) 172.20 233.59 17%
Acoustic (dBA) 54.02 2.43 0%
Table 5.27 shows the summary of the impact of IEQ on bio-signals by different locations. With the datasets of office
zones C1 and C2, EDA was most affected by lighting and CO2 condition in office zone C1 while it was most affected
by CO2 in office zone C2. Skin temperature was most affected by air temperature in office zones C1 and C2, heart
rate was most influenced by lighting condition in office zone C2, and stress level was most affected by lighting
condition in both office zones C1 and C2. The effects obtained from the zone C1 dataset and the zone C2 dataset were
not significantly different.
Figures 5.17 to 5.20 illustrate the frequency distribution of the first- and second-ranking IEQ factors affecting each
bio-signal within the datasets from all the locations. Among all the locations, EDA was most affected by lighting and
CO2 condition followed by air temperature, skin temperature was most affected by air temperature and then lighting
condition, the heart rate of occupants was most affected by lighting condition followed by CO2 level, and stress level
was most influenced by lighting condition and then air temperature. Overall, except skin temperature, the other three
65
bio-signals were all mostly affected by lighting condition while skin temperature was most influenced by air
temperature.
Table 5.27 Summary of the Impact of IEQ on Bio-Signals by Different Locations
Location
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Studio
Lighting
(3.71%)
Acoustic
(4.03%)
Temp
(4.14%)
CO2
(4.31%)
EDA
Classroom
Lighting
(0.39%)
Temp
(1.84%)
Office Zone C1 day1
Lighting
(2.19%)
Temp
(2.64%)
Office Zone C1 day2
Lighting
(0.47%)
Acoustic
(0.71%)
CO2
(0.87%)
Office Zone C1 day3
CO2
(2.26%)
Office Zone C1 day4
CO2
(2.39%)
Lighting
(3.30%)
Temp
(3.80%)
Office Zone C2 day1
Acoustic
(1.43%)
Temp
(2.30%)
Lighting
(2.60%)
Office Zone C2 day2
Temp
(0.52%)
Lighting
(0.92%)
Office Zone C2 day3
CO2
(0.86%)
Office Zone C2 day4
CO2
(2.85%)
Studio
CO2
(6.41%)
Temp
(8.92%)
Lighting
(10.33%)
Skin Temp
Classroom
Temp
(2.66%)
Lighting
(5.66%)
Acoustic
(7.18%)
CO2
(7.53%)
Office Zone C1 day1
Lighting
(2.62%)
CO2
(2.98%)
Office Zone C1 day2
Temp
(5.55%)
Office Zone C1 day3
Temp
(9.65%)
Office Zone C1 day4
CO2
(0.88%)
Lighting
(1.11%)
Office Zone C2 day1
Lighting
(2.35%)
CO2
(3.24%)
Office Zone C2 day2
Temp
(6.46%)
Lighting
(8.30%)
CO2
(8.45%)
Office Zone C2 day3
Temp
(10.59%)
Acoustic
(10.80%)
Lighting
(10.94%)
CO2
(11.10%)
Office Zone C2 day4
CO2
(5.83%)
Temp
(6.18%)
Acoustic
(6.37%)
Studio
Lighting
(4.07%)
CO2
(4.96%)
Temp
(6.20%)
Heart Rate
Classroom
Lighting
(0.75%)
Acoustic
(1.43%)
Office Zone C1 day1
Acoustic
(0.83%)
Lighting
(1.05%)
Office Zone C1 day2
CO2
(0.18%)
66
Office Zone C1 day3
Lighting
(0.80%)
Office Zone C1 day4
Temp
(0.25%)
Office Zone C2 day1
Lighting
(4.84%)
CO2
(5.87%)
Temp
(7.11%)
Office Zone C2 day2
Lighting
(1.87%)
CO2
(2.42%)
Office Zone C2 day3
Lighting
(1.56%)
Temp
(1.79%)
Acoustic
(2.00%)
Office Zone C2 day4
CO2
(3.08%)
Temp
(3.44%)
Studio
Lighting
(15.16%)
Temp
(16.34%)
Stress Level
Classroom
Temp
(3.41%)
Office Zone C1 day1
Office Zone C1 day2
Acoustic
(0.18%)
Office Zone C1 day3
Lighting
(0.27%)
Office Zone C1 day4
Lighting
(1.46%)
Office Zone C2 day1
CO2
(4.63%)
Temp
(5.59%)
Acoustic
(5.85%)
Office Zone C2 day2
Lighting
(1.11%)
Temp
(1.65%)
Office Zone C2 day3
Lighting
(1.71%)
Acoustic
(2.19%)
Temp
(2.62%)
Office Zone C2 day4
Lighting
(1.26%)
Temp
(1.76%)
Figure 5.17 Frequency Distribution of the First and Second
IEQ Factors Affecting EDA
10%
40% 40%
10%
0
1
2
3
4
5
Acoustic CO2 Lighting Temp
Frequency
(10 cases)
1st Affect EDA
29% 29%
43%
0
1
2
3
4
Acoustic Lighting Temp
Frequency
(7 cases)
2nd Affect EDA
67
Figure 5.18 Frequency Distribution of the First and Second
IEQ Factors Affecting Skin Temperature
Figure 5.19 Frequency Distribution of the First and Second
IEQ Factors Affecting Heart Rate
Figure 5.20 Frequency Distribution of the First and Second
IEQ Factors Affecting Stress Level
Table 5.28 shows the summary of the impact of bio-signals on satisfaction and sensation in all locations. With the
datasets of office zones C1 and C2, thermal satisfaction was most affected by EDA and then skin temperature in zone
C1, and it was most affected by skin temperature and then stress level in zone C2. IAQ satisfaction was most affected
by heart rate in office zone C1, and it was most affected by EDA and then heart rate in zone C2. Lighting satisfaction
was most affected by EDA followed by heart rate in both office zones C1 and C2. Acoustic satisfaction was most
influenced by stress level in office zone C1, and it was mostly influenced by skin temperature and EDA in zone C2.
Overall satisfaction was most affected by EDA, followed by skin temperature in office zone C1, while it was most
affected by skin temperature and then stress in zone C2. From the overall observations, the results for zones C1 and
C2 were similar only with regard to the impact on lighting satisfaction; the results differed on other satisfaction data.
Figures 5.21 to 5.25 illustrate the frequency distribution of the first- and second-ranking physiological factors affecting
each satisfaction within the datasets in all the locations. In all the locations, EDA affected thermal satisfaction the
most followed by skin temperature. IAQ satisfaction was most affected by EDA and skin temperature, followed by
30%
20%
50%
0
2
4
6
CO2 Lighting Temp
Frequency
(10 cases)
1st Affect Skin Temp
13%
25%
38%
25%
0
1
2
3
4
Acoustic CO2 Lighting Temp
Frequency
(8 cases)
2nd Affect Skin Temp
10%
20%
60%
10%
0
2
4
6
8
Acoustic CO2 Lighting Temp
Frequency
(10 cases)
1st Affect Heart Rate
14%
43%
14%
29%
0
1
2
3
4
Acoustic CO2 Lighting Temp
Frequency
(7 cases)
2nd Affect Heart Rate
13% 13%
63%
13%
0
2
4
6
Acoustic CO2 Lighting Temp
Frequency
(8 cases)
1st Affect Stress Level
25%
75%
0
1
2
3
4
Acoustic Temp
Frequency
(4 cases)
2nd Affect Stress Level
68
heart rate. Lighting satisfaction was most affected by skin temperate and EDA, followed by heart rate. Acoustic
satisfaction was most influenced by skin temperature and then heart rate. Overall satisfaction was most affected by
EDA and heart rate followed by skin temperature. From these observations, it can be seen that EDA was the biometric
factor that affected most satisfaction data (acoustic satisfaction being the exception). Acoustic satisfaction was mostly
influenced by skin temperature.
Table 5.28 Summary of the Impact of Bio-Signals on Satisfaction/Sensation in All Locations
Location
1st affect
(r-sq)
2nd affect
(r-sq)
3rd affect
(r-sq)
4th affect
(r-sq)
Studio
EDA
(1.19%)
Skin temp
(1.63%)
Thermal Satisfaction
Classroom
EDA
(5.65%)
Skin temp
(9.81%)
HR
(11.49%)
Stress
(13.79%)
Office Zone C1 day1
Skin temp
(7.99%)
EDA
(11.28%)
HR
(11.45%)
Stress
(11.80%)
Office Zone C1 day2
EDA
(42.21%)
Skin temp
(45.62%)
HR
(46.23%)
Office Zone C1 day3
Stress
(32.30%)
Skin temp
(47.56%)
EDA
(49.17%)
HR
(49.80%)
Office Zone C1 day4
EDA
(9.90%)
Skin temp
(10.97%)
Stress
(12.24%)
HR
(12.65%)
Office Zone C2 day1
Skin temp
(13.98%)
Stress
(14.64%)
Office Zone C2 day2
Skin temp
(0.99%)
EDA
(1.40%)
Office Zone C2 day3
EDA
(14.19%)
Stress
(18.32%)
Skin temp
(23.35%)
Office Zone C2 day4
HR
(3.50%)
Skin temp
(4.94%)
Stress
(5.97%)
EDA
(6.82%)
Studio
Skin temp
(5.27%)
EDA
(6.29%)
HR
(6.50%)
Stress
(7.97%)
Thermal Sensation
Classroom
Stress
(1.56%)
HR
(2.86%)
Office Zone C1 day1
Office Zone C1 day2
Office Zone C1 day3
Office Zone C1 day4
Office Zone C2 day1
Office Zone C2 day2
Office Zone C2 day3
Office Zone C2 day4
Studio
HR
(0.42%)
EDA
(0.79%)
IAQ Satisfaction
Classroom
Skin temp
(7.84%)
EDA
(12.27%)
HR
(19.33%)
Stress
(19.96%)
Office Zone C1 day1
Skin temp
(10.01%)
HR
(15.99%)
Stress
(24.11%)
EDA
(24.30%)
Office Zone C1 day2
EDA
(8.30%)
Stress
(22.17%)
HR
(45.27%)
Skin temp
(45.77%)
Office Zone C1 day3
HR
(34.03%)
Skin temp
(41.21%)
EDA
(45.79%)
Stress
(46.06%)
69
Office Zone C1 day4
HR
(15.26%)
EDA
(20.85%)
Skin temp
(24.46%)
Office Zone C2 day1
Stress
(44.18%)
HR
(46.59%)
EDA
(46.85%)
Office Zone C2 day2
Skin temp
(29.43%)
HR
(38.45%)
Stress
(43.26%)
Office Zone C2 day3
EDA
(58.90%)
Skin temp
(60.34%)
Office Zone C2 day4
EDA
(20.18%)
HR
(23.61%)
Skin temp
(24.64%)
Stress
(25.33%)
Studio
EDA
(0.58%)
Stress
(1.00%)
IAQ Sensation
Classroom
HR
(4.43%)
EDA
(10.85%)
Stress
(11.99%)
Skin temp
(13.04%)
Office Zone C1 day1
Office Zone C1 day2
Office Zone C1 day3
Office Zone C1 day4
Office Zone C2 day1
Office Zone C2 day2
Office Zone C2 day3
Office Zone C2 day4
Studio
Skin temp
(1.00%)
EDA
(1.25%)
Lighting Satisfaction
Classroom
Stress
(15.19%)
EDA
(16.59%)
HR
(17.43%)
Office Zone C1 day1
Skin temp
(16.05%)
HR
(19.63%)
EDA
(19.94%)
Office Zone C1 day2
EDA
(16.63%)
Stress
(20.47%)
HR
(26.54%)
Skin temp
(29.79%)
Office Zone C1 day3
Skin temp
(10.71%)
HR
(14.75%)
Stress
(20.40%)
Office Zone C1 day4
EDA
(5.08%)
HR
(8.29%)
Stress
(17.18%)
Office Zone C2 day1
Stress
(12.29%)
HR
(12.80%)
Skin temp
(13.72%)
Office Zone C2 day2
Skin temp
(43.61%)
HR
(47.14%)
Stress
(53.91%)
Office Zone C2 day3
EDA
(25.00%)
Skin temp
(27.69%)
HR
(28.85%)
Stress
(32.67%)
Office Zone C2 day4
EDA
(18.51%)
HR
(19.71%)
Skin temp
(20.60%)
Stress
(21.19%)
Studio
Stress
(0.66%)
HR
(1.06%)
EDA
(1.41%)
Lighting Sensation
Classroom
Skin temp
(21.28%)
EDA
(30.14%)
Stress
(31.04%)
HR
(32.42%)
Office Zone C1 day1
Office Zone C1 day2
Office Zone C1 day3
Office Zone C1 day4
70
Office Zone C2 day1
Office Zone C2 day2
Office Zone C2 day3
Office Zone C2 day4
Studio
Stress
(1.36%)
HR
(3.84%)
EDA
(4.58%)
Acoustic Satisfaction
Classroom
Skin temp
(23.63%)
EDA
(26.75%)
Stress
(27.39%)
Office Zone C1 day1
Skin temp
(0.25%)
Office Zone C1 day2
EDA
(9.06%)
Stress
(17.06%)
HR
(39.38%)
Skin temp
(39.92%)
Office Zone C1 day3
Stress
(19.19%)
Skin temp
(25.61%)
EDA
(28.53%)
HR
(29.23%)
Office Zone C1 day4
Stress
(8.55%)
EDA
(13.47%)
Skin temp
(15.52%)
HR
(15.83%)
Office Zone C2 day1
Skin temp
(1.97%)
Stress
(26.17%)
HR
(28.14%)
Office Zone C2 day2
Skin temp
(48.10%)
HR
(52.52%)
Stress
(60.03%)
Office Zone C2 day3
EDA
(64.30%)
Skin temp
(71.68%)
Stress
(72.35%)
HR
(72.48%)
Office Zone C2 day4
EDA
(17.81%)
HR
(20.82%)
Skin temp
(21.79%)
Stress
(22.37%)
Studio
Stress
(1.32%)
HR
(2.64%)
Skin temp
(3.10%)
Acoustic Sensation
Classroom
Skin temp
(34.33%)
HR
(37.11%)
Stress
(38.63%)
EDA
(39.13%)
Office Zone C1 day1
Office Zone C1 day2
Office Zone C1 day3
Office Zone C1 day4
Office Zone C2 day1
Office Zone C2 day2
Office Zone C2 day3
Office Zone C2 day4
Studio
HR
(1.54%)
EDA
(2.12%)
Skin temp
(2.39%)
Stress
(2.60%)
Overall Satisfaction
Classroom
HR
(9.68%)
EDA
(19.21%)
Office Zone C1 day1
EDA
(17.17%)
Skin temp
(25.42%)
HR
(34.45%)
Stress
(36.08%)
Office Zone C1 day2
EDA
(27.57%)
Skin temp
(34.29%)
Stress
(42.06%)
HR
(58.20%)
Office Zone C1 day3
Stress
(22.24%)
Skin temp
(34.73%)
HR
(38.48%)
EDA
(39.46%)
Office Zone C1 day4
HR
(15.54%)
EDA
(20.29%)
Skin temp
(23.16%)
Stress
(24.18%)
Office Zone C2 day1
Skin temp
(13.75%)
Stress
(30.28%)
71
Office Zone C2 day2
Skin temp
(21.47%)
Stress
(26.18%)
Office Zone C2 day3
Office Zone C2 day4
EDA
(18.86%)
Stress
(20.80%)
Figure 5.21 Frequency Distribution of the First and Second
Bio-Signals Affecting Thermal Satisfaction
Figure 5.22 Frequency Distribution of the First and Second
Bio-Signals Affecting IAQ Satisfaction
Figure 5.23 Frequency Distribution of the First and Second
Bio-Signals Affecting Lighting Satisfaction
50%
10%
30%
10%
0
2
4
6
EDA HR Skin temp Stress
Frequency
(10 cases)
1st Affect Thermal Satisfaction
20%
60%
20%
0
2
4
6
8
EDA Skin temp Stress
Frequency
(10 cases)
2nd Affect Thermal Satisfaction
30% 30% 30%
10%
0
1
2
3
4
EDA HR Skin temp Stress
Frequency
(10 cases)
1st Affect IAQ Satisfaction
30%
40%
20%
10%
0
1
2
3
4
5
EDA HR Skin temp Stress
Frequency
(10 cases)
2nd Affect IAQ Satisfaction
40% 40%
20%
0
1
2
3
4
5
EDA Skin temp Stress
Frequency
(10 cases)
1st Affect Lighting Satisfaction
20%
60%
10% 10%
0
2
4
6
8
EDA HR Skin temp Stress
Frequency
(10 cases)
2nd Affect Lighting Satisfaction
72
Figure 5.24 Frequency Distribution of the First and Second
Bio-Signals Affecting Acoustic Satisfaction
Figure 5.25 Frequency Distribution of the First and Second
Bio-Signals Affecting Overall IEQ Satisfaction
30%
40%
30%
0
1
2
3
4
5
EDA Skin temp Stress
Frequency
(10 cases)
1st Affect Acoustic Satisfaction
22%
33%
22% 22%
0
1
2
3
4
EDA HR Skin temp Stress
Frequency
(9 cases)
2nd Affect Acoustic Satisfaction
33% 33%
22%
11%
0
1
2
3
4
EDA HR Skin temp Stress
Frequency
(9 cases)
1st Affect Overall Satisfaction
33% 33% 33%
0
1
2
3
4
EDA Skin temp Stress
Frequency
(9 cases)
2nd Affect Overall Satisfaction
73
6. CONCLUSIONS AND FUTURE WORK
6.1 Conclusions
The research aimed to investigate the relationships among biometric signals, IEQ, and survey satisfaction. The author
collected data from three different locations: studio, classroom, and office. Two rounds of data were collected in the
studio, and each collection lasted four hours. One round of data collection was carried out in the classroom, and it
lasted about four hours as well. Four rounds of data collection were conducted in the office, which were day1, day2,
day3, and day4, and each round (day) collection lasted eight hours. Since the occupants in the office did not sit close
enough, the author decided to divide the office into zone C1 and zone C2 based on seating location. Two wearable
sensors were used to measure human bio-signals, including EDA, skin temperature, heart rate, and stress level. Three
types of IEQ sensors were used to complete the on-site measurement of IEQ data, which included indoor air
temperature, CO2 level, illuminance level, and acoustic condition. Survey data was collected by requiring the
participants to fill out a questionnaire once per hour. After finishing the collection part, data analysis was performed
by using Minitab’s stepwise regression analysis to identify the relationships among the three parts of data. The author
performed analysis based on the dataset of each location. The data from two collection rounds in the studio was
analyzed together, and the data from the office was divided into two parts, that is, office zone C1 and office zone C2.
After finishing the analysis of four datasets (studio, classroom, office zone C1, office zone C2), all the datasets were
put together and compared to determine the consistent findings from different locations. The obtained conclusions
were as follows:
• Lighting condition most significantly affected all measured bio-signals, including EDA, skin temperature,
heart rate, and stress level.
• Besides the first-ranking IEQ parameter (lighting condition), EDA, skin temperature, and stress condition
were secondly affected by air temperature; heart rate was secondly affected by CO2 level.
• Acoustic satisfaction was most affected by EDA followed by heart rate.
• Except for acoustic satisfaction, the other surveyed satisfactions (thermal satisfaction, IAQ satisfaction,
lighting satisfaction, overall satisfaction) were most significantly affected by EDA followed by skin
temperature.
• Females were more easily affected by CO2 while males were more affected by lighting condition.
• Males’ IEQ satisfaction was more likely to be affected by skin temperature while females’ IEQ satisfaction
tended to be most affected by EDA.
• The junior group was more likely to be affected by lighting condition while the mid-age group was more
affected by air temperature.
• The mid-age group’s IEQ satisfaction tend to be most influenced by skin temperature while the junior group’s
satisfaction was more likely to be influenced by skin temperature and EDA, which means both affected the
mid-age group’s satisfaction.
6.2 Limitations
In this study, as to data collection and analysis, there were some limitations which may have influenced the preciseness
of the results. First, during data collection, there were four biometric parameters, which may be influenced by some
other factors. The author collected skin temperature as one of the bio-signals to analyze the relation between IEQ and
survey data. However, hand temperature may have a slight influence on wrist temperature. For instance, if a participant
holds a hot coffee, then his skin temperature may increase slightly. Additionally, people’s responses to caffeine are
different. For example, some people are sensitive to it, so it may influence their heart rate slightly if they take too
much. As the collection of this experiment is done in real working conditions, it is difficult to eliminate these
limitations and reduce the influence of the data collection process and sensors of the experiment on the normal work
of occupants.
Additionally, the data from the questionnaires were not absolutely precise. The reason is that the standards of every
person may be different when they were filling out the satisfaction survey. For example, some people may think that
the indoor air temperature is correctly at the comfort zone for them, which means that they feel neither cold nor hot at
this temperature, and then the answer should be “neutral” for the thermal sensation question. Some people may be
comfortable with the current air temperature; thus, the answer should be “satisfied” for the thermal sensation question.
74
Thus, when the survey data from different participants was compared and analyzed with the same standards, it is
difficult to avoid these errors completely.
Moreover, the number of participants in the junior and mid-age groups are uneven, with the number of participants in
the mid-age group being greatly smaller than that in the junior group. When the author was performing data analysis,
such percentage would influence the analysis results in the distribution of age. It is difficult in Minitab to perform
stepwise regression and find consistent findings from insufficient data.
Also, in the data cleaning steps adopted in the data preprocessing part, the author selected an expected range of
measured biometric data and removed the data points outside the range. These unusual data were regarded as
measurement errors. The range was chosen based on participants’ activities in the indoor environment, which was the
concentrating or resting condition. However, the data points removed were also possibly reasonable when the
occupants are under significant stress. Thus, this is still a debate and may be different from similar studies by other
scholars.
All in all, there are still some points for possible improvements in this study. In future research, these limitations will
be improved so that more accurate and realistic results can be obtained.
6.3 Future Work
As to the further research on this topic, it can establish a model for prediction based on an understanding of the
relations among bio-signals, IEQ, and satisfaction. With the collected bio-signal data, according to its relationships
with satisfaction, it could predict that if an occupant is in their comfort zone, they are satisfied with the current indoor
environment. Of course, the bio-signal and survey data of the occupants in the building need to be collected first; in
other words, the database must be ready. After the establishment of the database, a more accurate prediction model
could be built by adopting machine learning. For instance, artificial neural networks (ANN) could be regarded as a
good method to establish a model and predict the variables. Similar to the neutral network in the brain, ANN can learn
input variables and optimize the model constantly in accordance with these variables to increase the accuracy of the
prediction results. In addition to ANN, the other machine learning and statistical analysis methods may also be used
to arrive at the same prediction. After several analyses using different algorithms, the study could compare these
algorithms or methods and finally select the most suitable one with the highest accuracy.
After establishing the prediction model, this study may even go further, that is, by calculating the specific indoor
environment parameters with the most comfortable feeling of occupants based on the relationships among occupants’
satisfaction, IEQ, and bio-signals, for example, air temperature, CO2 concentration, and illumination. Then these
comfort IEQ values will be inputted to update the HVAC system. When such a process could be done in real time, the
occupants in the building could always feel comfortable. Besides, this can also save energy and reduce unnecessary
energy consumption by the HVAC system.
Thus, when the two models are combined, the whole process will be that, first, judgment about whether the occupant
is comfortable will be made based on the measured bio-signals, and then if the judgment is “comfortable,” the current
system setting will be maintained; if the judgment is “uncomfortable,” the specific values of the IEQ comfort
parameter will be calculated and inputted into the HVAC control system. Finally, when the process is successful, as
long as the occupants wear the sensors that measure their bio-signals, they will enjoy a comfortable built environment.
6.4 Summary
This chapter concluded the significant findings from the analysis discussed in the previous chapter and also pointed
out the limitations of this study and future work. The most important findings in this research are the relationships
between IEQ, human comfort, and biometric signals; that is, air temperature was the most significant IEQ factor
affecting bio-signals, and human IEQ satisfaction was most clearly affected by EDA and skin temperature. The
conclusions also prove the hypothesis of this study. Each biometric factor could be found to have a relationship with
IEQ factors, and the bio-signals collected from a crowd in a multioccupancy condition could somehow reflect the
satisfaction or dissatisfaction of occupants. However, this study still has some possible improvements in the future,
such as the distribution of the participants’ age and the data-cleaning approach.
75
References
Acharya, Usha, Michael Beth Edwards, Ramon A. Jorquera, Hugo Silva, Kunio Nagashima, Pedro Labarca, and
Jairaj K. Acharya. 2006. “Drosophila Melanogaster Scramblases Modulate Synaptic Transmission.” The
Journal of Cell Biology 173 (1): 69–82. https://doi.org/10.1083/jcb.200506159.
“ANSI/ASHRAE Standard 55-2017 Thermal Environmental Conditions for Human Occupancy.” 2017. Atlanta.
https://doi.org/10.1111/ina.12364.
“ANSI/ASHRAE Standard 62.1-2013 Ventilation for Acceptable Indoor Air Quality.” 2013. Atlanta.
www.ashrae.org.
“ASHRAE Standard 90.1 Energy Standard for Buildings Except Low-Rise Residential Buildings.” 2010. Atlanta.
www.ashrae.org/technology.
Berg, Ruben Geert van den. n.d. “SPSS Correlation Analyis - Simple Tutorial & Example.” SPSS. Accessed
February 26, 2019. https://www.spss-tutorials.com/spss-correlation-analysis/.
BERNTSON, GARY G., J. THOMAS BIGGER, DWAIN L. ECKBERG, PAUL GROSSMAN, PETER G.
KAUFMANN, MAREK MALIK, HAIKADY N. NAGARAJA, et al. 1997. “Heart Rate Variability: Origins,
Methods, and Interpretive Caveats.” Psychophysiology 34 (6): 623–48. https://doi.org/10.1111/j.1469-
8986.1997.tb02140.x.
Blasco, Jorge, Thomas M Chen, Juan Tapiador, and Pedro Peris-Lopez. 2016. “A Survey of Wearable Biometric
Recognition Systems.” ACM Comput. Surv 49 (43). https://doi.org/10.1145/2968215.
Bornoiu, Ionuţ-Vlad, and Ovidiu Grigore. 2014. “EVALUATION OF MULTIPLE LEVELS OF
PSYCHOLOGICAL STRESS USING ONLY ELECTRODERMAL ACTIVITY DATA.” Rev. Roum. Sci.
Techn.-Électrotechn. et Énerg. Vol. 59.
https://pdfs.semanticscholar.org/9a52/ad4ffe2c1530660a603eb154a2dc5dd8b89d.pdf.
Boucsein, Wolfram. 2012. Electrodermal Activity. 2nd ed. New York: Springer Science+Business Media.
https://doi.org/10.1007/978-1-4614-1126-0.
Choi, Joon-Ho, and Vivian Loftness. 2012. “Investigation of Human Body Skin Temperatures as a Bio-Signal to
Indicate Overall Thermal Sensations.” Building and Environment 58: 258–69.
https://doi.org/10.1016/j.buildenv.2012.07.003.
Chrisinger, Benjamin W., and Abby C. King. 2018. “Stress Experiences in Neighborhood and Social Environments
(SENSE): A Pilot Study to Integrate the Quantified Self with Citizen Science to Improve the Built
Environment and Health.” International Journal of Health Geographics 17 (1): 17.
https://doi.org/10.1186/s12942-018-0140-1.
“Class 2 Data-Logging Sound Level Meter PCE-SDL 1 | PCE Instruments.” n.d. PCE Instruments. Accessed
February 22, 2019. https://www.pce-instruments.com/us/measuring-instruments/test-meters/sound-level-
meter-noise-level-meter-pce-instruments-class-2-data-logging-sound-level-meter-pce-sdl-1-det_2144973.htm.
“Embrace Seizure Monitoring | Smarter Epilepsy Management | Embrace Watch | Empatica.” n.d. Empatica Inc.
Accessed February 25, 2019. https://www.empatica.com/en-int/embrace2/.
Èppe, Peter Ho. 2002. “Different Aspects of Assessing Indoor and Outdoor Thermal Comfort.” Energy and
Buildings 34: 661–65. https://pdf.sciencedirectassets.com/271089/1-s2.0-S0378778800X0031X/1-s2.0-
S0378778802000178/main.pdf?x-amz-security-
token=AgoJb3JpZ2luX2VjEJP%252F%252F%252F%252F%252F%252F%252F%252F%252F%252FwEaC
XVzLWVhc3QtMSJHMEUCIFAGwrgrAdPdEGJuoDbEYWZJD1Yx8MiO57t7iOQ0na1rAiEAgucspK%252
FU%252.
Göçer, Özgür, Kenan Göçer, Ebru Ergöz, Karahan & Işıl, İlhan Oygür, Ebru Ergöz Karahan, and Işıl İlhan Oygür.
2018. “Exploring Mobility & Workplace Choice in a Flexible Office through Post-Occupancy
Evaluation.” Ergonomics 61 (2): 226–42. https://doi.org/10.1080/00140139.2017.1349937.
“Health, Wellbeing and Productivity in Offices: The Next Chapter for Green Building.” 2014. London.
https://www.worldgbc.org/news-media/health-wellbeing-and-productivity-offices-next-chapter-green-
building.
“Heart Rate Variability and Stress Level.” n.d. Garmin Corporation . Accessed April 22, 2019.
https://www8.garmin.com/manuals/webhelp/vivosmart3/EN-US/GUID-9282196F-D969-404D-B678-
F48A13D8D0CB.html.
“HOBO MX1102 Carbon Dioxide | Onset HOBO Data Logger.” n.d. Onset Computer Corporation. Accessed
February 22, 2019. https://www.onsetcomp.com/products/data-loggers/mx1102.
“HOBO MX2202 Temperature/Light | Onset HOBO Data Loggers.” n.d. Onset Computer Corporation. Accessed
76
February 22, 2019. https://www.onsetcomp.com/products/data-loggers/mx2202.
Hwang, Taeyon, and Jeong Tai Kim. 2013. “Assessment of Indoor Environmental Quality in Open-Plan Offices” 22
(1): 139–56. https://doi.org/10.1177/1420326X12470280.
“India’s Aadhaar Database to Offer Face Recognition.” 2018. https://doi.org/10.1016/S0969-4765(18)30025-0.
Karim, Nasim, Jahan Ara Hasan, and Syed Sanowar Ali. 2011. “HEART RATE VARIABILITY-A REVIEW.”
Journal of Basic and Applied Sciences 7 (1): 71–77. https://search-proquest-
com.libproxy1.usc.edu/docview/902689558/fulltextPDF/D5B486693D924C3EPQ/1?accountid=14749.
Kim, Joyce, Yuxun Zhou, Stefano Schiavon, Paul Raftery, and Gail Brager. 2018. “Personal Comfort Models:
Predicting Individuals’ Thermal Preference Using Occupant Heating and Cooling Behavior and Machine
Learning.” Building and Environment 129: 96–106. https://doi.org/10.1016/j.buildenv.2017.12.011.
Kim, Jungsoo, Richard De Dear, Christhina Cândido, Hui Zhang, and Edward Arens. 2013. “Gender Differences in
Office Occupant Perception of Indoor Environmental Quality (IEQ).” Building and Environment 70: 245–56.
https://doi.org/10.1016/j.buildenv.2013.08.022.
Learning from Our Buildings. 2002. Washington, D.C.: National Academies Press. https://doi.org/10.17226/10288.
Liu, Weiwei, Zhiwei Lian, Qihong Deng, and Yuanmou Liu. 2011. “Evaluation of Calculation Methods of Mean
Skin Temperature for Use in Thermal Comfort Study.” Building and Environment 46: 478–88.
https://doi.org/10.1016/j.buildenv.2010.08.011.
Mallawaarachchi, Harshini, Lalith De Silva, and Raufdeen Rameezdeen. 2016. “Indoor Environmental Quality and
Occupants’ Productivity” 6: 462–77. https://doi.org/10.1108/BEPAM-09-2015-0046.
“Minitab Statistical Software - Minitab.” n.d. Minitab, LLC. Accessed February 26, 2019.
http://www.minitab.com/en-us/products/minitab/.
Nayak, Tapsya, Tinghe Zhang, Zijing Mao, Xiaojing Xu, Lin Zhang, Daniel Pack, Bing Dong, and Yufei Huang.
2018. “Prediction of Human Performance Using Electroencephalography under Different Indoor Room
Temperatures.” Brain Sciences 8 (4): 74. https://doi.org/10.3390/brainsci8040074.
Ramos, Nuno M.M., João M.P.Q. Delgado, Ricardo M.S.F. Almeida, Maria L Simões, and Sofia Manuel. 2016.
Application of Data Mining Techniques in the Analysis of Indoor Hygrothermal Conditions. 1st ed. 20. Cham,
Switzerland : Springer. https://doi.org/https://doi-org.libproxy2.usc.edu/10.1007/978-3-319-22294-3.
Schaffner, Chris. 2009. “CO2 MONITORING Advances Air Quality and Energy Efficiency.” Vol. 103.
https://search-proquest-
com.libproxy1.usc.edu/docview/210229365/fulltextPDF/7620F55F1BBC45A5PQ/1?accountid=14749.
Shan, Xin, Anastasia Nissa Melina, and En-Hua Yang. 2018. “Impact of Indoor Environmental Quality on Students’
Wellbeing and Performance in Educational Building through Life Cycle Costing Perspective.”
https://doi.org/10.1016/j.jclepro.2018.09.002.
Syima Mahbob, Naziatul, Syahrul Nizam Kamaruzzaman, Naziah Salleh, and Raha Sulaiman. 2011. “A Correlation
Studies of Indoor Environmental Quality(IEQ) Towards Productive Workplace.” In 2nd International
Conference on Environmental Science and Technology , 434–38. Singapore.
https://pdfs.semanticscholar.org/c2f9/f0dfae4a982e0961b9f6e848104d2907fd18.pdf.
“Vívosmart® 3 | Fitness Activity Tracker | GARMIN.” n.d. Garmin Ltd. Accessed February 25, 2019.
https://buy.garmin.com/en-US/US/p/567813.
“What Is the Empatica Research Portal? – Empatica Support.” 2019. 2019. https://support.empatica.com/hc/en-
us/articles/115004319903-What-is-the-Empatica-Research-Portal-.
77
Appendix A
78
Appendix B
79
Appendix C
80
Appendix D
Stepwise regression with the dataset: studio
For each analysis: α to enter = 0.15, α to remove = 0.15
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 2.9432 0.831 0.823
Stress Level -0.005445 0.000 -0.01493 0.000 -0.01471 0.000
HR 0.03181 0.000 0.03233 0.000
EDA -0.1029 0.000
S 1.20533 1.19032 1.18601
R-sq 1.36% 3.84% 4.58%
R-sq(adj) 1.32% 3.76% 4.46%
R-sq(pred) 1.19% 3.60% 4.28%
Mallows’ Cp 80.68 19.66 3.02
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 2.4459 1.099 2.432
Stress Level -0.004686 0.000 -0.01073 0.000 -0.01101 0.000
HR 0.02029 0.000 0.02180 0.000
Skin Temp -0.0467 0.001
S 1.05383 1.04697 1.04474
R-sq 1.32% 2.64% 3.10%
R-sq(adj) 1.28% 2.56% 2.98%
R-sq(pred) 1.15% 2.41% 2.81%
Mallows’ Cp 44.55 13.51 4.14
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2-----
Coef P Coef P
Constant 1.545 1.398
Skin Temp 0.0631 0.000 0.0685 0.000
EDA -0.0479 0.015
S 0.968108 0.967123
R-sq 1.00% 1.25%
R-sq(adj) 0.96% 1.16%
R-sq(pred) 0.85% 1.02%
Mallows’ Cp 7.71 3.76
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 2.5179 3.010 3.013
Stress Level 0.002183 0.000 0.004392 0.000 0.004303 0.000
HR -0.00741 0.002 -0.00761 0.001
EDA 0.0414 0.003
81
S 0.699203 0.697928 0.696816
R-sq 0.66% 1.06% 1.41%
R-sq(adj) 0.61% 0.98% 1.29%
R-sq(pred) 0.50% 0.84% 1.13%
Mallows’ Cp 17.94 10.04 3.30
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------
Coef P Coef P
Constant 3.809 3.822
HR -0.00410 0.001 -0.00445 0.001
EDA 0.0370 0.003
S 0.618020 0.617012
R-sq 0.42% 0.79%
R-sq(adj) 0.38% 0.70%
R-sq(pred) 0.27% 0.59%
Mallows’ Cp 10.51 3.58
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 3.16894 3.2132 2.956
EDA 0.03198 0.000 0.03455 0.000 0.03242 0.000
Stress Level -0.001079 0.001 -0.001104 0.001
Skin Temp 0.00839 0.140
S 0.426348 0.425518 0.425415
R-sq 0.58% 1.00% 1.09%
R-sq(adj) 0.54% 0.92% 0.97%
R-sq(pred) 0.32% 0.67% 0.68%
Mallows’ Cp 12.84 4.36 4.18
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 3.3359 1.703
EDA 0.1321 0.000 0.1182 0.000
Skin Temp 0.0531 0.001
S 1.21952 1.21708
R-sq 1.19% 1.63%
R-sq(adj) 1.15% 1.55%
R-sq(pred) 1.09% 1.45%
Mallows’ Cp 11.10 2.38
Thermal sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2---- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -2.197 -1.864 -2.143 -2.986
Skin Temp 0.1601 0.000 0.1478 0.000 0.1441 0.000 0.1383 0.000
EDA 0.1088 0.000 0.1053 0.000 0.1111 0.000
82
HR 0.00500 0.022 0.02229 0.000
Stress Level -0.00822 0.000
S 1.04869 1.04325 1.04233 1.03432
R-sq 5.27% 6.29% 6.50% 7.97%
R-sq(adj) 5.23% 6.22% 6.38% 7.81%
R-sq(pred) 5.12% 6.11% 6.24% 7.63%
Mallows’ Cp 69.87 45.02 41.68 5.00
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 3.872 3.896 3.002 3.263
HR -0.01125 0 -0.01189 0 -0.0125 0 -0.01789 0
EDA 0.067 0 0.0596 0.001 0.0577 0.001
Skin Temp 0.0307 0.01 0.0326 0.007
Stress Level 0.00255 0.023
S 0.882508 0.880051 0.87903 0.878268
R-sq 1.54% 2.12% 2.39% 2.60%
R-sq(adj) 1.49% 2.04% 2.27% 2.44%
R-sq(pred) 1.39% 1.95% 2.15% 2.30%
Mallows’ Cp 25.18 12.73 8.15 5
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 0.1332 -1.276 3.09 5.92
Lighting 0.00219 0 0.00188 0 0.00205 0 0.001908 0
Acoustic 0.02616 0 0.02198 0.004 0.02344 0.002
Temp -0.1762 0.044 -0.2597 0.005
CO2 -0.00182 0.009
S 1.10445 1.10274 1.10229 1.10141
R-sq 3.71% 4.03% 4.14% 4.31%
R-sq(adj) 3.68% 3.98% 4.06% 4.21%
R-sq(pred) 3.62% 3.91% 3.96% 4.05%
Mallows’ Cp 22.42 11.93 9.87 5
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 77.402 99.62 285.2
Lighting 0.02758 0.000 0.02300 0.000 0.02635 0.000
CO2 -0.04248 0.000 -0.06253 0.000
Temp -7.46 0.000
S 12.9838 12.9258 12.8430
R-sq 4.07% 4.96% 6.20%
R-sq(adj) 4.05% 4.90% 6.12%
R-sq(pred) 3.97% 4.78% 5.97%
Mallows’ Cp 76.83 46.50 3.02
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- -----Step 2----- ------Step 3-----
83
Coef P Coef P Coef P
Constant 22.637 -13.59 -9.58
CO2 0.015523 0.000 0.02006 0.000 0.02180 0.000
Temp 1.439 0.000 1.217 0.000
Lighting 0.002402 0.000
S 1.79010 1.76615 1.75257
R-sq 6.41% 8.92% 10.33%
R-sq(adj) 6.38% 8.87% 10.27%
R-sq(pred) 6.31% 8.77% 10.14%
Mallows’ Cp 179.62 66.79 3.93
Stress vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 26.895 345.7 402.2
Lighting 0.10267 0.000 0.11279 0.000 0.11070 0.000
Temp -13.59 0.000 -15.28 0.000
CO2 -0.0317 0.076
S 24.2187 24.0534 24.0445
R-sq 15.16% 16.34% 16.43%
R-sq(adj) 15.13% 16.28% 16.34%
R-sq(pred) 15.04% 16.17% 16.19%
Mallows’ Cp 43.95 4.50 3.35
Stepwise regression with the dataset: studio (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.4416 3.546 3.744 2.302
Stress Level 0.00543 0 0.00939 0 0.00946 0 0.00948 0
HR -0.0159 0 -0.0179 0 -0.01824 0
EDA -0.2177 0.001 -0.2356 0
Skin Temp 0.0478 0.002
S 0.776125 0.77057 0.767742 0.765569
R-sq 3.21% 4.65% 5.42% 6.01%
R-sq(adj) 3.14% 4.52% 5.22% 5.76%
R-sq(pred) 2.98% 4.32% 1.36% 0.58%
Mallows’ Cp 43.07 22.37 12.4 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant -0.581 -0.443 -2.124 -2.075
Skin Temp 0.103 0 0.1037 0 0.1015 0 0.1067 0
Stress Level -0.0035 0.001 -0.0097 0 -0.00966 0
HR 0.02514 0 0.023 0
EDA -0.2293 0.005
S 1.01214 1.00845 0.997555 0.995266
R-sq 1.67% 2.45% 4.62% 5.12%
R-sq(adj) 1.61% 2.32% 4.42% 4.86%
84
R-sq(pred) 1.40% 2.05% 4.07% 3.97%
Mallows’ Cp 52.55 42.39 10.8 5
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.1603 1.379 0.3 -1.062
EDA -0.5708 0 -0.5111 0 -0.5033 0 -0.5203 0
HR 0.00949 0 0.02749 0 0.02715 0
Stress Level -0.0083 0 -0.00822 0
Skin Temp 0.0452 0.007
S 0.8378 0.8341 0.8216 0.819894
R-sq 4.69% 5.59% 8.46% 8.91%
R-sq(adj) 4.62% 5.46% 8.27% 8.66%
R-sq(pred) 0.00% 0.00% 0.00% 0.00%
Mallows’ Cp 67.36 54.79 10.3 5
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 0.348 0.531 0.483 1.516
Skin Temp 0.1066 0 0.0986 0 0.0979 0 0.1004 0
EDA 0.3706 0 0.3893 0 0.3382 0
Stress Level 0.00144 0.088 0.00527 0
HR -0.01576 0
S 0.8431 0.8348 0.8343 0.8294
R-sq 2.56% 4.54% 4.73% 5.91%
R-sq(adj) 2.49% 4.41% 4.53% 5.65%
R-sq(pred) 2.21% 0.00% 0.00% 0.00%
Mallows’ Cp 51.45 22.4 21.45 5
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 2.5982 0.003 -0.121
EDA 0.4022 0.000 0.3739 0.000 0.4220 0.000
Skin Temp 0.0844 0.000 0.0826 0.000
Stress Level 0.003712 0.000
S 0.621435 0.612217 0.604944
R-sq 4.25% 7.13% 9.39%
R-sq(adj) 4.18% 7.01% 9.20%
R-sq(pred) 0.00% 0.00% 0.00%
Mallows’ Cp 82.68 37.76 3.05
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 5.78 6.286 6.906 6.926
Skin Temp -0.0776 0 -0.0767 0 -0.0762 0 -0.07408 0
HR -0.0066 0 -0.0172 0 -0.01805 0
Stress Level 0.00488 0 0.004908 0
85
EDA -0.0923 0.017
S 0.4778 0.4744 0.4667 0.466
R-sq 4.15% 5.57% 8.68% 9.03%
R-sq(adj) 4.09% 5.44% 8.49% 8.78%
R-sq(pred) 3.89% 5.19% 8.17% 5.66%
Mallows’ Cp 78.07 57.08 8.76 5
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ -------Step 3------ -------Step 4------
Coef P Coef P Coef P Coef P
Constant 5.105 5.056 5.103 4.858
Skin Temp -0.0644 0 -0.0623 0 -0.0616 0 -0.06216 0
EDA -0.0989 0 -0.1168 0 -0.1047 0
Stress Level -0.0014 0 -0.00229 0
HR 0.00373 0.007
S 0.3165 0.315 0.3131 0.3124
R-sq 6.37% 7.33% 8.51% 8.96%
R-sq(adj) 6.31% 7.21% 8.33% 8.71%
R-sq(pred) 6.11% 4.95% 4.53% 5.27%
Mallows’ Cp 40.95 27.36 10.27 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -4.969 -4.016 -3.985
Skin Temp 0.2697 0.000 0.2714 0.000 0.2755 0.000
HR -0.01241 0.001 -0.01395 0.000
EDA -0.178 0.082
S 1.24532 1.24092 1.24007
R-sq 7.15% 7.87% 8.06%
R-sq(adj) 7.09% 7.75% 7.87%
R-sq(pred) 6.88% 7.47% 5.43%
Mallows’ Cp 13.61 4.11 3.08
Thermal sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -6.284 -7.327 -8.823
Skin Temp 0.2959 0.000 0.2940 0.000 0.2929 0.000
HR 0.01358 0.000 0.03916 0.000
Stress Level -0.01177 0.000
S 1.05969 1.05327 1.03296
R-sq 11.36% 12.49% 15.89%
R-sq(adj) 11.30% 12.37% 15.72%
R-sq(pred) 11.10% 12.13% 15.44%
Mallows’ Cp 79.36 61.52 3.88
Stress vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
86
Constant 31.949 77.8 323.8
Lighting 1.1850 0.000 1.0880 0.000 1.1357 0.000
CO2 -0.0877 0.000 -0.1182 0.000
Temp -9.80 0.001
S 22.8741 22.7518 22.6843
R-sq 18.80% 19.71% 20.23%
R-sq(adj) 18.75% 19.62% 20.09%
R-sq(pred) 18.61% 19.44% 19.86%
Mallows’ Cp 30.87 12.83 3.38
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 78.937 111.06 299.4
Lighting 0.3698 0.000 0.3055 0.000 0.3464 0.000
CO2 -0.06184 0.000 -0.08261 0.000
Temp -7.56 0.000
S 12.0243 11.8867 11.7948
R-sq 6.97% 9.13% 10.57%
R-sq(adj) 6.92% 9.04% 10.44%
R-sq(pred) 6.81% 8.86% 10.20%
Mallows’ Cp 83.17 34.80 3.18
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 29.9693 25.643 -16.62
Lighting 0.04612 0.000 0.05402 0.000 0.04506 0.000
CO2 0.008358 0.000 0.01288 0.000
Temp 1.700 0.000
S 1.45889 1.43865 1.39738
R-sq 7.47% 10.05% 15.17%
R-sq(adj) 7.43% 9.98% 15.07%
R-sq(pred) 7.33% 9.82% 14.85%
Mallows’ Cp 234.45 157.53 3.16
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------ -------Step 3------ -------Step 4------
Coef P Coef P Coef P Coef P
Constant 1.221 1.447 0.965 -1.44
CO2 -0.002 0 -0.0023 0 -0.0025 0 -0.00223 0
Lighting -0.0037 0.001 -0.0051 0 -0.00576 0
Acoustic 0.01032 0.013 0.01209 0.004
Temp 0.0932 0.066
S 0.4787 0.4778 0.4772 0.477
R-sq 1.51% 1.96% 2.22% 2.36%
R-sq(adj) 1.47% 1.87% 2.09% 2.19%
R-sq(pred) 1.23% 1.62% 1.80% 1.90%
Mallows’ Cp 19.22 10.62 6.39 5
Stepwise regression with the dataset: studio (male)
87
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------
Coef P Coef P
Constant 3.7374 1.488
Stress Level -0.00812 0.000 -0.02080 0.000
HR 0.03571 0.000
S 0.818654 0.796199
R-sq 5.60% 10.80%
R-sq(adj) 5.50% 10.61%
R-sq(pred) 5.19% 10.27%
Mallows’ Cp 52.77 1.57
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 3.0361 1.687 -1.719 1.108
EDA -0.1644 0.000 -0.2097 0.000 -0.1883 0.000 -0.1681 0.000
HR 0.01838 0.000 0.07701 0.000 0.08629 0.000
Stress Level -0.02943 0.000 -0.03186 0.000
Skin Temp -0.1118 0.000
S 1.42366 1.41363 1.36335 1.34940
R-sq 2.92% 4.38% 11.16% 13.06%
R-sq(adj) 2.82% 4.18% 10.88% 12.69%
R-sq(pred) 2.62% 3.91% 10.53% 12.29%
Mallows’ Cp 109.00 95.11 23.62 5.00
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 5.837 5.681 3.772 3.537
Skin Temp -0.1065 0.000 -0.0946 0.000 -0.1298 0.000 -0.1246 0.000
Stress Level -0.00613 0.000 -0.02263 0.000 -0.02211 0.000
HR 0.04747 0.000 0.04872 0.000
EDA -0.0479 0.093
S 1.24618 1.23805 1.21393 1.21276
R-sq 2.53% 3.90% 7.70% 7.98%
R-sq(adj) 2.43% 3.70% 7.41% 7.59%
R-sq(pred) 2.18% 3.37% 7.03% 7.19%
Mallows’ Cp 54.85 42.82 5.82 5.00
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.4028 1.948 1.549
Stress Level -0.00380 0.009 -0.00436 0.003 -0.00781 0.005
Skin Temp 0.0477 0.013 0.0403 0.041
HR 0.00993 0.147
S 1.09523 1.09220 1.09156
88
R-sq 0.71% 1.36% 1.58%
R-sq(adj) 0.61% 1.15% 1.27%
R-sq(pred) 0.33% 0.88% 0.93%
Mallows’ Cp 7.38 3.13 3.02
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.064 4.341 4.663 5.005
Skin Temp -0.05 0 -0.0606 0 -0.0523 0 -0.0479 0.001
EDA 0.0636 0 0.0808 0 0.0772 0
HR -0.0079 0.004 -0.01602 0.001
Stress Level 0.00401 0.044
S 0.7805 0.7753 0.7723 0.771
R-sq 1.44% 2.85% 3.71% 4.13%
R-sq(adj) 1.33% 2.64% 3.41% 3.72%
R-sq(pred) 1.10% 2.35% 3.00% 3.24%
Mallows’ Cp 25.48 13.61 7.07 5
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.496 2.329 1.425 1.905
Skin Temp 0.0369 0.005 0.0293 0.03 0.0193 0.151
HR 0.00533 0.032 0.02606 0 0.02778 0
Stress Level -0.01 0 -0.01036 0
S 0.7547 0.7533 0.743 0.7434
R-sq 0.84% 1.32% 4.10% 3.89%
R-sq(adj) 0.74% 1.11% 3.80% 3.69%
R-sq(pred) 0.49% 0.75% 3.31% 3.24%
Mallows’ Cp 31.54 28.82 3.43 3.5
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 1.357 1.450 1.936
Skin Temp 0.06209 0.000 0.06632 0.000 0.07170 0.000
HR -0.00298 0.084 -0.01413 0.000
Stress Level 0.00536 0.000
S 0.522879 0.522331 0.518153
R-sq 4.77% 5.07% 6.68%
R-sq(adj) 4.67% 4.87% 6.38%
R-sq(pred) 4.42% 4.37% 5.81%
Mallows’ Cp 18.75 17.72 3.43
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 1.679 5.103 5.735 4.787
HR 0.02381 0.000 0.02989 0.000 0.02467 0.000 0.04727 0.000
Skin Temp -0.1255 0.000 -0.1360 0.000 -0.1481 0.000
89
EDA 0.1074 0.000 0.1175 0.000
Stress Level -0.01111 0.000
S 1.08635 1.06265 1.05269 1.04384
R-sq 4.76% 8.97% 10.76% 12.35%
R-sq(adj) 4.66% 8.78% 10.48% 11.98%
R-sq(pred) 4.34% 8.44% 10.12% 11.53%
Mallows’ Cp 80.63 37.38 20.07 5.00
Thermal sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 2.4324 3.718 2.238
EDA 0.2157 0.000 0.2589 0.000 0.2502 0.000
HR -0.01752 0.000 -0.01963 0.000
Skin Temp 0.0532 0.001
S 0.927723 0.913119 0.908531
R-sq 10.87% 13.74% 14.70%
R-sq(adj) 10.77% 13.56% 14.43%
R-sq(pred) 10.61% 13.22% 14.02%
Mallows’ Cp 41.58 11.78 3.22
Stress vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- -----Step 3----
Coef P Coef P Coef P
Constant 21.04 801.7 536.8
Lighting 0.8356 0.000 1.0797 0.000 1.2050 0.000
Temp -33.27 0.000 -25.52 0.000
CO2 0.1573 0.000
S 23.1327 21.9908 21.6599
R-sq 9.90% 18.64% 21.14%
R-sq(adj) 9.82% 18.50% 20.94%
R-sq(pred) 9.58% 18.21% 20.60%
Mallows’ Cp 165.71 38.07 3.04
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2----
Coef P Coef P
Constant 219.4 252.4
Temp -6.05 0.000 -7.51 0.000
Lighting 0.1218 0.003
S 13.1673 13.1305
R-sq 1.02% 1.65%
R-sq(adj) 0.95% 1.50%
R-sq(pred) 0.76% 1.23%
Mallows’ Cp 9.42 2.86
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
90
Constant 10.168 12.63
CO2 0.03934 0.000 0.03932 0.000
Acoustic -0.0443 0.018
S 1.98415 1.98112
R-sq 25.30% 25.57%
R-sq(adj) 25.25% 25.47%
R-sq(pred) 25.10% 25.29%
Mallows’ Cp 4.93 1.32
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -0.0804 15.92 11.14 14.99
Lighting 0.07693 0.000 0.08122 0.000 0.07551 0.000 0.07310 0.000
Temp -0.682 0.000 -0.564 0.005 -0.675 0.001
Acoustic 0.0376 0.029 0.0396 0.021
CO2 -0.00257 0.119
S 1.52407 1.51762 1.51551 1.51471
R-sq 17.64% 18.39% 18.68% 18.83%
R-sq(adj) 17.58% 18.27% 18.50% 18.59%
R-sq(pred) 17.41% 18.06% 18.27% 18.17%
Mallows’ Cp 18.88 8.24 5.43 5.00
Stepwise regression with the dataset: studio (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------
Coef P Coef P
Constant 4.202 4.213
HR -0.01481 0.000 -0.01524 0.000
EDA 0.0542 0.003
S 0.893707 0.892076
R-sq 2.34% 2.74%
R-sq(adj) 2.30% 2.65%
R-sq(pred) 2.18% 2.55%
Mallows’ Cp 11.10 3.99
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 3.1565 3.1913 0.942 0.528
Stress Level -0.0089 0 -0.0086 0 -0.0088 0 -0.01194 0
EDA -0.1095 0 -0.1279 0 -0.1274 0
Skin Temp 0.0734 0 0.064 0
HR 0.0106 0.016
S 1.1666 1.1612 1.1561 1.1549
R-sq 3.50% 4.43% 5.31% 5.56%
R-sq(adj) 3.45% 4.35% 5.18% 5.39%
R-sq(pred) 3.31% 4.19% 5.00% 5.16%
Mallows’ Cp 47.93 27.72 8.8 5
91
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 2.6197 2.115 2.111
Stress Level -0.007504 0.000 -0.00977 0.000 -0.00974 0.000
HR 0.00760 0.050 0.00785 0.042
EDA -0.0416 0.048
S 1.04625 1.04559 1.04491
R-sq 3.10% 3.27% 3.44%
R-sq(adj) 3.06% 3.18% 3.31%
R-sq(pred) 2.91% 3.01% 3.12%
Mallows’ Cp 6.92 5.06 3.14
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 1.655 1.996 1.796
Skin Temp 0.0608 0.000 0.0696 0.000 0.0762 0.000
HR -0.00766 0.001 -0.00740 0.001
EDA -0.0590 0.003
S 0.990677 0.988385 0.986718
R-sq 0.89% 1.39% 1.77%
R-sq(adj) 0.84% 1.30% 1.64%
R-sq(pred) 0.72% 1.15% 1.46%
Mallows’ Cp 19.50 10.03 3.43
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 2.6387 2.918 3.201
EDA 0.0316 0.025 0.0338 0.016 0.0334 0.018
HR -0.00351 0.026 -0.00834 0.001
Stress Level 0.002288 0.018
S 0.699433 0.698819 0.698112
R-sq 0.22% 0.44% 0.69%
R-sq(adj) 0.18% 0.36% 0.56%
R-sq(pred) 0.07% 0.21% 0.39%
Mallows’ Cp 11.26 8.29 4.73
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.155 4.637 3.817 3.918
HR -0.0084 0 -0.0166 0 -0.0186 0 -0.01856 0
Stress Level 0.0039 0 0.0044 0 0.004313 0
Skin Temp 0.03107 0 0.02735 0.002
EDA 0.0318 0.012
S 0.623 0.6203 0.6187 0.6179
R-sq 1.55% 2.44% 2.99% 3.26%
92
R-sq(adj) 1.51% 2.36% 2.86% 3.09%
R-sq(pred) 1.39% 2.21% 2.69% 2.91%
Mallows’ Cp 38.68 20.05 9.34 5
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- -------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.022 2.223 2.54 2.639
Skin Temp 0.03689 0 0.04205 0 0.04856 0 0.04491 0
HR -0.0045 0 -0.0133 0 -0.01325 0
Stress Level 0.00406 0 0.003973 0
EDA 0.03126 0
S 0.41 0.4079 0.4035 0.4023
R-sq 1.89% 2.90% 5.05% 5.66%
R-sq(adj) 1.85% 2.81% 4.93% 5.49%
R-sq(pred) 1.73% 2.62% 4.71% 5.13%
Mallows’ Cp 88.78 66.78 17.48 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 0.484 1.046 1.373
Skin Temp 0.0962 0.000 0.1107 0.000 0.0998 0.000
HR -0.01265 0.000 -0.01308 0.000
EDA 0.0965 0.000
S 1.22914 1.22389 1.22009
R-sq 1.44% 2.32% 2.97%
R-sq(adj) 1.39% 2.23% 2.84%
R-sq(pred) 1.26% 2.06% 2.69%
Mallows’ Cp 35.04 16.57 3.57
Thermal sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -2.560 -3.343 -2.905 -3.299
Skin Temp 0.1691 0.000 0.1490 0.000 0.1345 0.000 0.1259 0.000
HR 0.01761 0.000 0.01704 0.000 0.02820 0.000
EDA 0.1292 0.000 0.1321 0.000
Stress Level -0.00517 0.000
S 1.04117 1.02870 1.02023 1.01751
R-sq 5.91% 8.19% 9.74% 10.26%
R-sq(adj) 5.87% 8.11% 9.62% 10.10%
R-sq(pred) 5.74% 7.95% 9.47% 9.90%
Mallows’ Cp 107.76 52.69 15.99 5.00
Stress vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 30.818 407.3
Lighting 0.9383 0.000 1.0703 0.000
93
Temp -16.04 0.000
S 24.0202 23.7828
R-sq 11.46% 13.23%
R-sq(adj) 11.42% 13.17%
R-sq(pred) 11.33% 13.04%
Mallows’ Cp 55.08 1.07
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- ------Step 3-----
Coef P Coef P Coef P
Constant 78.938 204.7 298.9
Lighting 0.2312 0.000 0.2747 0.000 0.2367 0.000
Temp -5.36 0.000 -8.15 0.000
CO2 -0.05484 0.000
S 12.8798 12.8306 12.7431
R-sq 2.51% 3.28% 4.63%
R-sq(adj) 2.48% 3.22% 4.54%
R-sq(pred) 2.39% 3.10% 4.37%
Mallows’ Cp 70.08 46.15 3.06
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 22.093 -15.10 -10.83
CO2 0.016674 0.000 0.02131 0.000 0.02278 0.000
Temp 1.478 0.000 1.250 0.000
Lighting 0.02548 0.000
S 1.78278 1.75672 1.74375
R-sq 7.44% 10.15% 11.50%
R-sq(adj) 7.42% 10.11% 11.43%
R-sq(pred) 7.34% 9.99% 11.30%
Mallows’ Cp 172.48 58.68 3.20
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 0.1640 -1.469 4.02 6.58
Lighting 0.02242 0.000 0.01869 0.000 0.02101 0.000 0.01968 0.000
Acoustic 0.03031 0.000 0.02496 0.003 0.02643 0.002
Temp -0.2217 0.019 -0.297 0.003
CO2 -0.001655 0.027
S 1.15292 1.15076 1.14999 1.14933
R-sq 3.04% 3.44% 3.59% 3.73%
R-sq(adj) 3.02% 3.38% 3.51% 3.62%
R-sq(pred) 2.95% 3.30% 3.41% 3.45%
Mallows’ Cp 23.16 11.41 7.89 5.00
Stepwise regression with the dataset: studio (mid-age)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
94
-----Step 1----- ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant -5.96 -2.481 -2.099 -2.957
Skin Temp 0.2779 0 0.203 0 0.1927 0 0.2217 0
HR -0.0174 0 -0.0195 0 -0.01857 0
Stress Level 0.01015 0 0.01043 0
EDA -1.223 0.032
S 0.2417 0.2196 0.2062 0.2041
R-sq 76.71% 80.88% 83.24% 83.68%
R-sq(adj) 76.57% 80.66% 82.95% 83.30%
R-sq(pred) 73.58% 77.28% 76.01% 74.37%
Mallows’ Cp 72.87 30.62 7.68 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant -5.962 8.19 7.40
HR 0.11904 0.000 0.0716 0.000 0.0759 0.000
Skin Temp -0.3566 0.000 -0.3353 0.000
Stress Level -0.02101 0.010
S 0.902583 0.826256 0.813053
R-sq 58.35% 65.30% 66.59%
R-sq(adj) 58.12% 64.90% 66.01%
R-sq(pred) 57.31% 62.07% 60.71%
Mallows’ Cp 43.33 9.10 4.35
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 16.231 7.34 6.75
Skin Temp -0.4712 0.000 -0.2798 0.000 -0.2640 0.000
HR 0.04452 0.000 0.04767 0.000
Stress Level -0.01558 0.002
S 0.564849 0.502027 0.489303
R-sq 63.42% 71.27% 72.86%
R-sq(adj) 63.21% 70.94% 72.39%
R-sq(pred) 60.38% 67.98% 66.61%
Mallows’ Cp 61.91 13.30 5.02
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.2646 3.283 3.346 7.78
EDA -2.628 0.001 -4.052 0 -4.476 0 -2.12 0.056
HR -0.0134 0.001 -0.0154 0 -0.02648 0
Stress Level 0.01474 0 0.01551 0
Skin Temp -0.1259 0.002
S 0.4329 0.4215 0.4076 0.3971
R-sq 6.29% 11.63% 17.85% 22.48%
R-sq(adj) 5.76% 10.62% 16.44% 20.69%
95
R-sq(pred) 2.30% 3.61% 2.85% 2.59%
Mallows’ Cp 35.14 25.21 13.33 5
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.2646 3.283 3.346 7.78
EDA -2.628 0.001 -4.052 0 -4.476 0 -2.12 0.056
HR -0.0134 0.001 -0.0154 0 -0.02648 0
Stress Level 0.01474 0 0.01551 0
Skin Temp -0.1259 0.002
S 0.4329 0.4215 0.4076 0.3971
R-sq 6.29% 11.63% 17.85% 22.48%
R-sq(adj) 5.76% 10.62% 16.44% 20.69%
R-sq(pred) 2.30% 3.61% 2.85% 2.59%
Mallows’ Cp 35.14 25.21 13.33 5
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 11.96 8.481 8.099 8.957
Skin Temp -0.2779 0 -0.203 0 -0.1927 0 -0.2217 0
HR 0.01741 0 0.01946 0 0.01857 0
Stress Level -0.0102 0 -0.01043 0
EDA 1.223 0.032
S 0.2417 0.2196 0.2062 0.2041
R-sq 76.71% 80.88% 83.24% 83.68%
R-sq(adj) 76.57% 80.66% 82.95% 83.30%
R-sq(pred) 73.58% 77.28% 76.01% 74.37%
Mallows’ Cp 72.87 30.62 7.68 5
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 11.96 8.481 8.099 8.957
Skin Temp -0.2779 0 -0.203 0 -0.1927 0 -0.2217 0
HR 0.01741 0 0.01946 0 0.01857 0
Stress Level -0.0102 0 -0.01043 0
EDA 1.223 0.032
S 0.2417 0.2196 0.2062 0.2041
R-sq 76.71% 80.88% 83.24% 83.68%
R-sq(adj) 76.57% 80.66% 82.95% 83.30%
R-sq(pred) 73.58% 77.28% 76.01% 74.37%
Mallows’ Cp 72.87 30.62 7.68 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------
Coef P Coef P
Constant 14.218 14.244
Skin Temp -0.3803 0.000 -0.3788 0.000
Stress Level -0.00970 0.001
S 0.301367 0.292890
96
R-sq 79.87% 81.09%
R-sq(adj) 79.75% 80.88%
R-sq(pred) 78.10% 77.62%
Mallows’ Cp 12.10 2.78
Thermal sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1-----
Coef P
Constant -1.104
Skin Temp 0.1583 0.000
S 0.362469
R-sq 32.21%
R-sq(adj) 31.83%
R-sq(pred) 30.70%
Mallows’ Cp 1.59
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------- -------Step 3-------
Coef P Coef P Coef P
Constant -0.0572 -0.1257 0.883
CO2 0.000254 0.047 0.000395 0.005 0.000424 0.002
Lighting -0.001300 0.015 -0.002120 0.001
Temp -0.0435 0.007
S 0.0408907 0.0405702 0.0401658
R-sq 1.26% 3.11% 5.33%
R-sq(adj) 0.94% 2.49% 4.42%
R-sq(pred) 0.05% 1.52% 3.28%
Mallows’ Cp 13.39 9.30 3.99
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 34.08 29.31
Acoustic -0.0680 0.054 -0.0779 0.029
CO2 0.01008 0.071
S 1.80823 1.80180
R-sq 1.16% 2.17%
R-sq(adj) 0.85% 1.55%
R-sq(pred) 0.00% 0.25%
Mallows’ Cp 3.11 1.85
Stepwise regression with the dataset: classroom
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2-----
Coef P Coef P
Constant 1.580 0.930
HR 0.02654 0.000 0.03367 0.000
EDA 0.08665 0.000
S 0.702794 0.665035
97
R-sq 9.68% 19.21%
R-sq(adj) 9.57% 19.03%
R-sq(pred) 9.32% 18.81%
Mallows’ Cp 105.28 2.73
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant -9.872 -10.243 -9.761
Skin Temp 0.4170 0.000 0.4268 0.000 0.4156 0.000
EDA 0.04864 0.000 0.03807 0.000
Stress Level -0.00293 0.005
S 0.656433 0.643278 0.640820
R-sq 23.63% 26.75% 27.39%
R-sq(adj) 23.55% 26.58% 27.14%
R-sq(pred) 23.28% 26.29% 26.81%
Mallows’ Cp 44.73 8.83 3.03
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant -8.392 -7.044 -6.36 -6.213
Skin Temp 0.3632 0 0.3457 0 0.3432 0 0.3382 0
HR -0.0104 0 -0.0208 0 -0.01985 0
Stress Level 0.00489 0 0.00379 0.001
EDA -0.01573 0.007
S 0.4389 0.4301 0.4251 0.4236
R-sq 34.44% 37.11% 38.63% 39.13%
R-sq(adj) 34.37% 36.97% 38.43% 38.86%
R-sq(pred) 34.09% 36.61% 38.01% 38.44%
Mallows’ Cp 67.16 30.36 10.26 5
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 4.0545 3.9171 2.890
Stress Level -0.01538 0.000 -0.01321 0.000 -0.01835 0.000
EDA 0.0433 0.000 0.0386 0.001
HR 0.01617 0.003
S 0.833648 0.827180 0.823491
R-sq 15.19% 16.59% 17.43%
R-sq(adj) 15.09% 16.41% 17.15%
R-sq(pred) 14.84% 16.18% 16.83%
Mallows’ Cp 23.71 10.66 3.71
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- -------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant -6.488 -6.989 -6.533 -7.712
Skin Temp 0.3165 0 0.3297 0 0.3191 0 0.3283 0
98
EDA 0.06566 0 0.05564 0 0.05195 0
Stress Level -0.0028 0.001 -0.00713 0
HR 0.0139 0
S 0.5331 0.5025 0.4995 0.4948
R-sq 21.28% 30.14% 31.04% 32.42%
R-sq(adj) 21.19% 29.99% 30.81% 32.12%
R-sq(pred) 21.01% 29.84% 30.63% 31.87%
Mallows’ Cp 144.79 30.84 21.09 5
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 8.1 7.795 5.513 5.848
Skin Temp -0.165 0 -0.157 0 -0.1257 0 -0.1252 0
EDA 0.03987 0 0.05438 0 0.06056 0
HR 0.01671 0 0.01035 0.001
Stress Level 0.00324 0.008
S 0.4955 0.4838 0.4642 0.4626
R-sq 7.84% 12.27% 19.33% 19.96%
R-sq(adj) 7.73% 12.07% 19.05% 19.60%
R-sq(pred) 7.50% 11.76% 18.66% 19.06%
Mallows’ Cp 132.93 86 10.04 5
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 2.091 1.76 2.178 4.006
HR 0.01114 0 0.01477 0 0.00721 0.011 0.00617 0.03
EDA 0.04411 0 0.05143 0 0.04928 0
Stress Level 0.00384 0.001 0.0038 0.001
Skin Temp -0.055 0.001
S 0.4485 0.4334 0.4309 0.4286
R-sq 4.43% 10.85% 11.99% 13.04%
R-sq(adj) 4.32% 10.65% 11.69% 12.65%
R-sq(pred) 4.03% 10.28% 11.17% 12.13%
Mallows’ Cp 86.57 23.29 13.74 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 3.7067 -3.23 -5.25 -6.41
EDA 0.0815 0.000 0.0861 0.000 0.0989 0.000 0.0776 0.000
Skin Temp 0.2186 0.000 0.2463 0.000 0.2445 0.000
HR 0.01479 0.000 0.03673 0.000
Stress Level -0.01116 0.000
S 0.909767 0.889976 0.882149 0.871095
R-sq 5.65% 9.81% 11.49% 13.79%
R-sq(adj) 5.54% 9.61% 11.19% 13.40%
R-sq(pred) 5.28% 9.09% 10.59% 12.71%
Mallows’ Cp 82.50 41.82 26.60 5.00
Thermal sensation vs bio
99
Candidate terms: EDA, Skin Temp, HR, Stress Level
-------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 3.1976 2.236 0.546
Stress Level -0.003687 0.000 -0.00821 0.000 -0.00831 0.000
HR 0.01491 0.001 0.01601 0.000
Skin Temp 0.0508 0.051
S 0.671495 0.667432 0.666367
R-sq 1.56% 2.86% 3.28%
R-sq(adj) 1.45% 2.64% 2.95%
R-sq(pred) 1.19% 2.29% 2.51%
Mallows’ Cp 16.57 6.69 4.85
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 7.67 -12.17
Lighting -0.0498 0.027 -0.1364 0.000
Temp 1.270 0.000
S 2.82200 2.80250
R-sq 0.39% 1.84%
R-sq(adj) 0.31% 1.69%
R-sq(pred) 0.00% 1.27%
Mallows’ Cp 18.60 2.36
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 110.42 100.8
Lighting -0.2443 0.001 -0.2614 0.000
Acoustic 0.2183 0.002
S 10.3584 10.3267
R-sq 0.75% 1.43%
R-sq(adj) 0.68% 1.29%
R-sq(pred) 0.49% 1.04%
Mallows’ Cp 8.82 1.17
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- -----Step 3----- -----Step 4-----
Coef P Coef P Coef P Coef P
Constant 19.41 16.56 14.30 13.44
Temp 0.4989 0.000 0.998 0.000 1.040 0.000 0.965 0.000
Lighting -0.0728 0.000 -0.0782 0.000 -0.0813 0.000
Acoustic 0.03578 0.000 0.03393 0.000
CO2 0.00354 0.018
S 1.11787 1.10090 1.09233 1.09067
R-sq 2.66% 5.66% 7.18% 7.53%
R-sq(adj) 2.60% 5.53% 7.00% 7.28%
R-sq(pred) 2.09% 4.75% 6.18% 6.41%
100
Mallows’ Cp 78.81 31.66 8.64 5.00
Stress vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 386.1 360.5
Temp -14.07 0.000 -13.74 0.000
Acoustic 0.327 0.059
S 23.8894 23.8646
R-sq 3.41% 3.68%
R-sq(adj) 3.33% 3.53%
R-sq(pred) 3.08% 3.20%
Mallows’ Cp 2.94 1.37
Stepwise regression with the dataset: classroom (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.4416 3.546 3.744 2.302
Stress Level 0.00543 0 0.00939 0 0.00946 0 0.00948 0
HR -0.0159 0 -0.0179 0 -0.01824 0
EDA -0.2177 0.001 -0.2356 0
Skin Temp 0.0478 0.002
S 0.7761 0.7706 0.7677 0.7656
R-sq 3.21% 4.65% 5.42% 6.01%
R-sq(adj) 3.14% 4.52% 5.22% 5.76%
R-sq(pred) 2.98% 4.32% 1.36% 0.58%
Mallows’ Cp 43.07 22.37 12.4 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant -0.581 -0.443 -2.124 -2.075
Skin Temp 0.103 0 0.1037 0 0.1015 0 0.1067 0
Stress Level -0.0035 0.001 -0.0097 0 -0.00966 0
HR 0.02514 0 0.023 0
EDA -0.2293 0.005
S 1.0121 1.0085 0.9976 0.9953
R-sq 1.67% 2.45% 4.62% 5.12%
R-sq(adj) 1.61% 2.32% 4.42% 4.86%
R-sq(pred) 1.40% 2.05% 4.07% 3.97%
Mallows’ Cp 52.55 42.39 10.8 5
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.1603 1.379 0.3 -1.062
EDA -0.5708 0 -0.5111 0 -0.5033 0 -0.5203 0
HR 0.00949 0 0.02749 0 0.02715 0
Stress Level -0.0083 0 -0.00822 0
Skin Temp 0.0452 0.007
101
S 0.8378 0.8341 0.8216 0.8199
R-sq 4.69% 5.59% 8.46% 8.91%
R-sq(adj) 4.62% 5.46% 8.27% 8.66%
R-sq(pred) 0.00% 0.00% 0.00% 0.00%
Mallows’ Cp 67.36 54.79 10.3 5
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 0.348 0.531 0.483 1.516
Skin Temp 0.1066 0 0.0986 0 0.0979 0 0.1004 0
EDA 0.3706 0 0.3893 0 0.3382 0
Stress Level 0.00144 0.088 0.00527 0
HR -0.01576 0
S 0.8431 0.8348 0.8343 0.8294
R-sq 2.56% 4.54% 4.73% 5.91%
R-sq(adj) 2.49% 4.41% 4.53% 5.65%
R-sq(pred) 2.21% 0.00% 0.00% 0.00%
Mallows’ Cp 51.45 22.4 21.45 5
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 2.5982 0.003 -0.121
EDA 0.4022 0.000 0.3739 0.000 0.4220 0.000
Skin Temp 0.0844 0.000 0.0826 0.000
Stress Level 0.003712 0.000
S 0.621435 0.612217 0.604944
R-sq 4.25% 7.13% 9.39%
R-sq(adj) 4.18% 7.01% 9.20%
R-sq(pred) 0.00% 0.00% 0.00%
Mallows’ Cp 82.68 37.76 3.05
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 5.78 6.286 6.906 6.926
Skin Temp -0.0776 0 -0.0767 0 -0.0762 0 -0.07408 0
HR -0.0066 0 -0.0172 0 -0.01805 0
Stress Level 0.00488 0 0.004908 0
EDA -0.0923 0.017
S 0.4778 0.4744 0.4667 0.466
R-sq 4.15% 5.57% 8.68% 9.03%
R-sq(adj) 4.09% 5.44% 8.49% 8.78%
R-sq(pred) 3.89% 5.19% 8.17% 5.66%
Mallows’ Cp 78.07 57.08 8.76 5
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ -------Step 3------ -------Step 4------
Coef P Coef P Coef P Coef P
Constant 5.105 5.056 5.103 4.858
102
Skin Temp -0.0644 0 -0.0623 0 -0.0616 0 -0.06216 0
EDA -0.0989 0 -0.1168 0 -0.1047 0
Stress Level -0.0014 0 -0.00229 0
HR 0.00373 0.007
S 0.3165 0.315 0.3131 0.3124
R-sq 6.37% 7.33% 8.51% 8.96%
R-sq(adj) 6.31% 7.21% 8.33% 8.71%
R-sq(pred) 6.11% 4.95% 4.53% 5.27%
Mallows’ Cp 40.95 27.36 10.27 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -4.969 -4.016 -3.985
Skin Temp 0.2697 0.000 0.2714 0.000 0.2755 0.000
HR -0.01241 0.001 -0.01395 0.000
EDA -0.178 0.082
S 1.24532 1.24092 1.24007
R-sq 7.15% 7.87% 8.06%
R-sq(adj) 7.09% 7.75% 7.87%
R-sq(pred) 6.88% 7.47% 5.43%
Mallows’ Cp 13.61 4.11 3.08
Thermal sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -6.284 -7.327 -8.823
Skin Temp 0.2959 0.000 0.2940 0.000 0.2929 0.000
HR 0.01358 0.000 0.03916 0.000
Stress Level -0.01177 0.000
S 1.05969 1.05327 1.03296
R-sq 11.36% 12.49% 15.89%
R-sq(adj) 11.30% 12.37% 15.72%
R-sq(pred) 11.10% 12.13% 15.44%
Mallows’ Cp 79.36 61.52 3.88
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------ -------Step 3------ -------Step 4------
Coef P Coef P Coef P Coef P
Constant 1.221 1.447 0.965 -1.44
CO2 -0.002 0 -0.0023 0 -0.0025 0 -0.00223 0
Lighting -0.0037 0.001 -0.0051 0 -0.00576 0
Acoustic 0.01032 0.013 0.01209 0.004
Temp 0.0932 0.066
S 0.4787 0.4778 0.4772 0.477
R-sq 1.51% 1.96% 2.22% 2.36%
R-sq(adj) 1.47% 1.87% 2.09% 2.19%
R-sq(pred) 1.23% 1.62% 1.80% 1.90%
Mallows’ Cp 19.22 10.62 6.39 5
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
103
-----Step 1----- ------Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 29.9693 25.643 -16.62
Lighting 0.04612 0.000 0.05402 0.000 0.04506 0.000
CO2 0.008358 0.000 0.01288 0.000
Temp 1.700 0.000
S 1.45889 1.43865 1.39738
R-sq 7.47% 10.05% 15.17%
R-sq(adj) 7.43% 9.98% 15.07%
R-sq(pred) 7.33% 9.82% 14.85%
Mallows’ Cp 234.45 157.53 3.16
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 78.937 111.06 299.4
Lighting 0.3698 0.000 0.3055 0.000 0.3464 0.000
CO2 -0.06184 0.000 -0.08261 0.000
Temp -7.56 0.000
S 12.0243 11.8867 11.7948
R-sq 6.97% 9.13% 10.57%
R-sq(adj) 6.92% 9.04% 10.44%
R-sq(pred) 6.81% 8.86% 10.20%
Mallows’ Cp 83.17 34.80 3.18
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 31.949 77.8 323.8
Lighting 1.1850 0.000 1.0880 0.000 1.1357 0.000
CO2 -0.0877 0.000 -0.1182 0.000
Temp -9.80 0.001
S 22.8741 22.7518 22.6843
R-sq 18.80% 19.71% 20.23%
R-sq(adj) 18.75% 19.62% 20.09%
R-sq(pred) 18.61% 19.44% 19.86%
Mallows’ Cp 30.87 12.83 3.38
Stepwise regression with the dataset: classroom (male)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------
Coef P Coef P
Constant 3.7374 1.488
Stress Level -0.00812 0.000 -0.02080 0.000
HR 0.03571 0.000
S 0.818654 0.796199
R-sq 5.60% 10.80%
R-sq(adj) 5.50% 10.61%
104
R-sq(pred) 5.19% 10.27%
Mallows’ Cp 52.77 1.57
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 3.0361 1.687 -1.719 1.108
EDA -0.1644 0.000 -0.2097 0.000 -0.1883 0.000 -0.1681 0.000
HR 0.01838 0.000 0.07701 0.000 0.08629 0.000
Stress Level -0.02943 0.000 -0.03186 0.000
Skin Temp -0.1118 0.000
S 1.42366 1.41363 1.36335 1.34940
R-sq 2.92% 4.38% 11.16% 13.06%
R-sq(adj) 2.82% 4.18% 10.88% 12.69%
R-sq(pred) 2.62% 3.91% 10.53% 12.29%
Mallows’ Cp 109.00 95.11 23.62 5.00
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 5.837 5.681 3.772 3.537
Skin Temp -0.1065 0.000 -0.0946 0.000 -0.1298 0.000 -0.1246 0.000
Stress Level -0.00613 0.000 -0.02263 0.000 -0.02211 0.000
HR 0.04747 0.000 0.04872 0.000
EDA -0.0479 0.093
S 1.24618 1.23805 1.21393 1.21276
R-sq 2.53% 3.90% 7.70% 7.98%
R-sq(adj) 2.43% 3.70% 7.41% 7.59%
R-sq(pred) 2.18% 3.37% 7.03% 7.19%
Mallows’ Cp 54.85 42.82 5.82 5.00
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.4028 1.948 1.549
Stress Level -0.00380 0.009 -0.00436 0.003 -0.00781 0.005
Skin Temp 0.0477 0.013 0.0403 0.041
HR 0.00993 0.147
S 1.09523 1.09220 1.09156
R-sq 0.71% 1.36% 1.58%
R-sq(adj) 0.61% 1.15% 1.27%
R-sq(pred) 0.33% 0.88% 0.93%
Mallows’ Cp 7.38 3.13 3.02
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.064 4.341 4.663 5.005
Skin Temp -0.05 0 -0.0606 0 -0.0523 0 -0.0479 0.001
EDA 0.0636 0 0.0808 0 0.0772 0
105
HR -0.0079 0.004 -0.01602 0.001
Stress Level 0.00401 0.044
S 0.7805 0.7753 0.7723 0.771
R-sq 1.44% 2.85% 3.71% 4.13%
R-sq(adj) 1.33% 2.64% 3.41% 3.72%
R-sq(pred) 1.10% 2.35% 3.00% 3.24%
Mallows’ Cp 25.48 13.61 7.07 5
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.496 2.329 1.425 1.905
Skin Temp 0.0369 0.005 0.0293 0.03 0.0193 0.151
HR 0.00533 0.032 0.02606 0 0.02778 0
Stress Level -0.01 0 -0.01036 0
S 0.7547 0.7533 0.743 0.7434
R-sq 0.84% 1.32% 4.10% 3.89%
R-sq(adj) 0.74% 1.11% 3.80% 3.69%
R-sq(pred) 0.49% 0.75% 3.31% 3.24%
Mallows’ Cp 31.54 28.82 3.43 3.5
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 1.357 1.450 1.936
Skin Temp 0.06209 0.000 0.06632 0.000 0.07170 0.000
HR -0.00298 0.084 -0.01413 0.000
Stress Level 0.00536 0.000
S 0.522879 0.522331 0.518153
R-sq 4.77% 5.07% 6.68%
R-sq(adj) 4.67% 4.87% 6.38%
R-sq(pred) 4.42% 4.37% 5.81%
Mallows’ Cp 18.75 17.72 3.43
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 1.679 5.103 5.735 4.787
HR 0.02381 0.000 0.02989 0.000 0.02467 0.000 0.04727 0.000
Skin Temp -0.1255 0.000 -0.1360 0.000 -0.1481 0.000
EDA 0.1074 0.000 0.1175 0.000
Stress Level -0.01111 0.000
S 1.08635 1.06265 1.05269 1.04384
R-sq 4.76% 8.97% 10.76% 12.35%
R-sq(adj) 4.66% 8.78% 10.48% 11.98%
R-sq(pred) 4.34% 8.44% 10.12% 11.53%
Mallows’ Cp 80.63 37.38 20.07 5.00
Thermal sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
106
Coef P Coef P Coef P
Constant 2.4324 3.718 2.238
EDA 0.2157 0.000 0.2589 0.000 0.2502 0.000
HR -0.01752 0.000 -0.01963 0.000
Skin Temp 0.0532 0.001
S 0.927723 0.913119 0.908531
R-sq 10.87% 13.74% 14.70%
R-sq(adj) 10.77% 13.56% 14.43%
R-sq(pred) 10.61% 13.22% 14.02%
Mallows’ Cp 41.58 11.78 3.22
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -0.0804 15.92 11.14 14.99
Lighting 0.07693 0.000 0.08122 0.000 0.07551 0.000 0.07310 0.000
Temp -0.682 0.000 -0.564 0.005 -0.675 0.001
Acoustic 0.0376 0.029 0.0396 0.021
CO2 -0.00257 0.119
S 1.52407 1.51762 1.51551 1.51471
R-sq 17.64% 18.39% 18.68% 18.83%
R-sq(adj) 17.58% 18.27% 18.50% 18.59%
R-sq(pred) 17.41% 18.06% 18.27% 18.17%
Mallows’ Cp 18.88 8.24 5.43 5.00
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 10.168 12.63
CO2 0.03934 0.000 0.03932 0.000
Acoustic -0.0443 0.018
S 1.98415 1.98112
R-sq 25.30% 25.57%
R-sq(adj) 25.25% 25.47%
R-sq(pred) 25.10% 25.29%
Mallows’ Cp 4.93 1.32
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2----
Coef P Coef P
Constant 219.4 252.4
Temp -6.05 0.000 -7.51 0.000
Lighting 0.1218 0.003
S 13.1673 13.1305
R-sq 1.02% 1.65%
R-sq(adj) 0.95% 1.50%
R-sq(pred) 0.76% 1.23%
Mallows’ Cp 9.42 2.86
Stress level vs IEQ
107
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- -----Step 3----
Coef P Coef P Coef P
Constant 21.04 801.7 536.8
Lighting 0.8356 0.000 1.0797 0.000 1.2050 0.000
Temp -33.27 0.000 -25.52 0.000
CO2 0.1573 0.000
S 23.1327 21.9908 21.6599
R-sq 9.90% 18.64% 21.14%
R-sq(adj) 9.82% 18.50% 20.94%
R-sq(pred) 9.58% 18.21% 20.60%
Mallows’ Cp 165.71 38.07 3.04
Stepwise regression with the dataset: classroom (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------
Coef P Coef P
Constant 4.202 4.213
HR -0.01481 0.000 -0.01524 0.000
EDA 0.0542 0.003
S 0.893707 0.892076
R-sq 2.34% 2.74%
R-sq(adj) 2.30% 2.65%
R-sq(pred) 2.18% 2.55%
Mallows’ Cp 11.10 3.99
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 3.158 3.1928 0.945 0.534
Stress Level -0.0089 0 -0.0086 0 -0.0088 0 -0.01193 0
EDA -0.1096 0 -0.1279 0 -0.1275 0
Skin Temp 0.0734 0 0.064 0
HR 0.01051 0.017
S 1.1667 1.1613 1.1562 1.155
R-sq 3.51% 4.45% 5.33% 5.57%
R-sq(adj) 3.47% 4.36% 5.20% 5.40%
R-sq(pred) 3.33% 4.20% 5.02% 5.17%
Mallows’ Cp 47.83 27.6 8.7 5
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 2.6205 2.119 2.115
Stress Level -0.007516 0.000 -0.00977 0.000 -0.00973 0.000
HR 0.00755 0.051 0.00781 0.044
EDA -0.0416 0.048
S 1.04643 1.04577 1.04509
R-sq 3.11% 3.27% 3.44%
108
R-sq(adj) 3.07% 3.19% 3.31%
R-sq(pred) 2.92% 3.01% 3.12%
Mallows’ Cp 6.87 5.06 3.14
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 1.654 1.994 1.794
Skin Temp 0.0609 0.000 0.0696 0.000 0.0762 0.000
HR -0.00763 0.001 -0.00737 0.001
EDA -0.0590 0.003
S 0.990844 0.988574 0.986909
R-sq 0.89% 1.39% 1.76%
R-sq(adj) 0.85% 1.30% 1.63%
R-sq(pred) 0.72% 1.15% 1.46%
Mallows’ Cp 19.38 10.02 3.43
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 2.6380 2.909 3.191
EDA 0.0317 0.024 0.0339 0.016 0.0335 0.017
HR -0.00341 0.031 -0.00822 0.001
Stress Level 0.002281 0.019
S 0.699005 0.698436 0.697732
R-sq 0.23% 0.43% 0.68%
R-sq(adj) 0.18% 0.34% 0.55%
R-sq(pred) 0.08% 0.20% 0.38%
Mallows’ Cp 10.93 8.25 4.72
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.152 4.635 3.814 3.916
HR -0.0083 0 -0.0166 0 -0.0186 0 -0.01853 0
Stress Level 0.0039 0 0.0044 0 0.004311 0
Skin Temp 0.03105 0 0.02733 0.002
EDA 0.0318 0.012
S 0.6231 0.6204 0.6188 0.618
R-sq 1.54% 2.43% 2.98% 3.25%
R-sq(adj) 1.50% 2.34% 2.85% 3.08%
R-sq(pred) 1.38% 2.20% 2.67% 2.90%
Mallows’ Cp 38.64 20.03 9.35 5
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- -------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.02 2.218 2.535 2.634
Skin Temp 0.03693 0 0.04202 0 0.04853 0 0.04487 0
HR -0.0045 0 -0.0132 0 -0.01317 0
109
Stress Level 0.00405 0 0.003969 0
EDA 0.0313 0
S 0.4097 0.4077 0.4032 0.402
R-sq 1.90% 2.88% 5.03% 5.64%
R-sq(adj) 1.85% 2.79% 4.91% 5.48%
R-sq(pred) 1.73% 2.60% 4.69% 5.11%
Mallows’ Cp 88.12 66.78 17.53 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 0.483 1.044 1.371
Skin Temp 0.0963 0.000 0.1107 0.000 0.0998 0.000
HR -0.01262 0.000 -0.01304 0.000
EDA 0.0965 0.000
S 1.22936 1.22414 1.22033
R-sq 1.44% 2.32% 2.97%
R-sq(adj) 1.39% 2.23% 2.84%
R-sq(pred) 1.26% 2.06% 2.68%
Mallows’ Cp 34.91 16.58 3.57
Thermal sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -2.561 -3.347 -2.909 -3.303
Skin Temp 0.1692 0.000 0.1490 0.000 0.1345 0.000 0.1258 0.000
HR 0.01767 0.000 0.01709 0.000 0.02826 0.000
EDA 0.1292 0.000 0.1321 0.000
Stress Level -0.00517 0.000
S 1.04137 1.02884 1.02035 1.01764
R-sq 5.91% 8.20% 9.75% 10.27%
R-sq(adj) 5.87% 8.12% 9.63% 10.11%
R-sq(pred) 5.74% 7.96% 9.48% 9.92%
Mallows’ Cp 108.05 52.72 16.01 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 0.1640 -1.469 4.02 6.58
Lighting 0.02242 0.000 0.01869 0.000 0.02101 0.000 0.01968 0.000
Acoustic 0.03031 0.000 0.02496 0.003 0.02643 0.002
Temp -0.2217 0.019 -0.297 0.003
CO2 -0.001655 0.027
S 1.15292 1.15076 1.14999 1.14933
R-sq 3.04% 3.44% 3.59% 3.73%
R-sq(adj) 3.02% 3.38% 3.51% 3.62%
R-sq(pred) 2.95% 3.30% 3.41% 3.45%
Mallows’ Cp 23.16 11.41 7.89 5.00
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
110
------Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 22.093 -15.10 -10.83
CO2 0.016674 0.000 0.02131 0.000 0.02278 0.000
Temp 1.478 0.000 1.250 0.000
Lighting 0.02548 0.000
S 1.78278 1.75672 1.74375
R-sq 7.44% 10.15% 11.50%
R-sq(adj) 7.42% 10.11% 11.43%
R-sq(pred) 7.34% 9.99% 11.30%
Mallows’ Cp 172.48 58.68 3.20
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- ------Step 3-----
Coef P Coef P Coef P
Constant 78.938 204.7 298.9
Lighting 0.2312 0.000 0.2747 0.000 0.2367 0.000
Temp -5.36 0.000 -8.15 0.000
CO2 -0.05484 0.000
S 12.8798 12.8306 12.7431
R-sq 2.51% 3.28% 4.63%
R-sq(adj) 2.48% 3.22% 4.54%
R-sq(pred) 2.39% 3.10% 4.37%
Mallows’ Cp 70.08 46.15 3.06
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 30.818 407.3
Lighting 0.9383 0.000 1.0703 0.000
Temp -16.04 0.000
S 24.0202 23.7828
R-sq 11.46% 13.23%
R-sq(adj) 11.42% 13.17%
R-sq(pred) 11.33% 13.04%
Mallows’ Cp 55.08 1.07
Stepwise regression with the dataset: classroom (mid-age)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant -5.96 -2.481 -2.099 -2.957
Skin Temp 0.2779 0 0.203 0 0.1927 0 0.2217 0
HR -0.0174 0 -0.0195 0 -0.01857 0
Stress Level 0.01015 0 0.01043 0
EDA -1.223 0.032
S 0.241698 0.219592 0.206222 0.204078
R-sq 76.71% 80.88% 83.24% 83.68%
R-sq(adj) 76.57% 80.66% 82.95% 83.30%
111
R-sq(pred) 73.58% 77.28% 76.01% 74.37%
Mallows’ Cp 72.87 30.62 7.68 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant -5.962 8.19 7.40
HR 0.11904 0.000 0.0716 0.000 0.0759 0.000
Skin Temp -0.3566 0.000 -0.3353 0.000
Stress Level -0.02101 0.010
S 0.902583 0.826256 0.813053
R-sq 58.35% 65.30% 66.59%
R-sq(adj) 58.12% 64.90% 66.01%
R-sq(pred) 57.31% 62.07% 60.71%
Mallows’ Cp 43.33 9.10 4.35
Acoustic sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 16.231 7.34 6.75
Skin Temp -0.4712 0.000 -0.2798 0.000 -0.2640 0.000
HR 0.04452 0.000 0.04767 0.000
Stress Level -0.01558 0.002
S 0.564849 0.502027 0.489303
R-sq 63.42% 71.27% 72.86%
R-sq(adj) 63.21% 70.94% 72.39%
R-sq(pred) 60.38% 67.98% 66.61%
Mallows’ Cp 61.91 13.30 5.02
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.8236 2.144 2.103 -0.851
EDA 1.752 0.001 2.701 0 2.984 0 1.414 0.056
HR 0.00891 0.001 0.01024 0 0.01765 0
Stress Level -0.0098 0 -0.01034 0
Skin Temp 0.084 0.002
S 0.28858 0.281028 0.271736 0.264731
R-sq 6.29% 11.63% 17.85% 22.48%
R-sq(adj) 5.76% 10.62% 16.44% 20.69%
R-sq(pred) 2.30% 3.61% 2.85% 2.59%
Mallows’ Cp 35.14 25.21 13.33 5
Lighting sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.2646 3.283 3.346 7.78
EDA -2.628 0.001 -4.052 0 -4.476 0 -2.12 0.056
HR -0.0134 0.001 -0.0154 0 -0.02648 0
Stress Level 0.01474 0 0.01551 0
Skin Temp -0.1259 0.002
112
S 0.432869 0.421542 0.407604 0.397096
R-sq 6.29% 11.63% 17.85% 22.48%
R-sq(adj) 5.76% 10.62% 16.44% 20.69%
R-sq(pred) 2.30% 3.61% 2.85% 2.59%
Mallows’ Cp 35.14 25.21 13.33 5
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 11.96 8.481 8.099 8.957
Skin Temp -0.2779 0 -0.203 0 -0.1927 0 -0.2217 0
HR 0.01741 0 0.01946 0 0.01857 0
Stress Level -0.0102 0 -0.01043 0
EDA 1.223 0.032
S 0.241698 0.219592 0.206222 0.204078
R-sq 76.71% 80.88% 83.24% 83.68%
R-sq(adj) 76.57% 80.66% 82.95% 83.30%
R-sq(pred) 73.58% 77.28% 76.01% 74.37%
Mallows’ Cp 72.87 30.62 7.68 5
IAQ sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 11.96 8.481 8.099 8.957
Skin Temp -0.2779 0 -0.203 0 -0.1927 0 -0.2217 0
HR 0.01741 0 0.01946 0 0.01857 0
Stress Level -0.0102 0 -0.01043 0
EDA 1.223 0.032
S 0.241698 0.219592 0.206222 0.204078
R-sq 76.71% 80.88% 83.24% 83.68%
R-sq(adj) 76.57% 80.66% 82.95% 83.30%
R-sq(pred) 73.58% 77.28% 76.01% 74.37%
Mallows’ Cp 72.87 30.62 7.68 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------
Coef P Coef P
Constant 14.218 14.244
Skin Temp -0.3803 0.000 -0.3788 0.000
Stress Level -0.00970 0.001
S 0.301367 0.292890
R-sq 79.87% 81.09%
R-sq(adj) 79.75% 80.88%
R-sq(pred) 78.10% 77.62%
Mallows’ Cp 12.10 2.78
Thermal sensation vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1-----
Coef P
Constant -1.104
Skin Temp 0.1583 0.000
113
S 0.362469
R-sq 32.21%
R-sq(adj) 31.83%
R-sq(pred) 30.70%
Mallows’ Cp 1.59
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------- -------Step 3-------
Coef P Coef P Coef P
Constant -0.0572 -0.1257 0.883
CO2 0.000254 0.047 0.000395 0.005 0.000424 0.002
Lighting -0.001300 0.015 -0.002120 0.001
Temp -0.0435 0.007
S 0.0408907 0.0405702 0.0401658
R-sq 1.26% 3.11% 5.33%
R-sq(adj) 0.94% 2.49% 4.42%
R-sq(pred) 0.05% 1.52% 3.28%
Mallows’ Cp 13.39 9.30 3.99
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 34.08 29.31
Acoustic -0.0680 0.054 -0.0779 0.029
CO2 0.01008 0.071
S 1.80823 1.80180
R-sq 1.16% 2.17%
R-sq(adj) 0.85% 1.55%
R-sq(pred) 0.00% 0.25%
Mallows’ Cp 3.11 1.85
Stepwise regression with the dataset: office zone C1 – day1
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3------ -------Step 4------
Coef P Coef P Coef P Coef P
Constant 3.8183 8.09 7.693 6.368
EDA 0.689 0 0.6976 0 0.6572 0 0.6273 0
Skin Temp -0.1424 0 -0.1679 0 -0.138 0
HR 0.01548 0 0.0237 0
Stress Level -0.00414 0
S 0.5012 0.4757 0.4461 0.4406
R-sq 17.17% 25.42% 34.45% 36.08%
R-sq(adj) 17.12% 25.33% 34.34% 35.93%
R-sq(pred) 9.47% 17.79% 27.62% 29.52%
Mallows’ Cp 537.39 304.46 49.31 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1-----
Coef P
Constant 5.613
114
Skin Temp -0.0518 0.036
S 1.17094
R-sq 0.25%
R-sq(adj) 0.19%
R-sq(pred) 0.02%
Mallows’ Cp 1.21
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- -----Step 4-----
Coef P Coef P Coef P Coef P
Constant -10.623 -11.204 -11.210 -10.359
Skin Temp 0.4642 0.000 0.4259 0.000 0.4255 0.000 0.4064 0.000
HR 0.02298 0.000 0.02271 0.000 0.01742 0.000
EDA 0.2244 0.010 0.2464 0.005
Stress Level 0.00267 0.098
S 1.19113 1.16576 1.16389 1.16333
R-sq 16.05% 19.63% 19.94% 20.06%
R-sq(adj) 16.00% 19.54% 19.80% 19.88%
R-sq(pred) 15.88% 19.37% 19.15% 19.10%
Mallows’ Cp 88.15 10.45 5.74 5.00
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 11.906 11.408 6.872 6.975
Skin Temp -0.243 0 -0.2758 0 -0.174 0 -0.1766 0
HR 0.01968 0 0.04768 0 0.04689 0
Stress Level -0.0142 0 -0.01389 0
EDA 0.1185 0.036
S 0.8175 0.7901 0.7512 0.7504
R-sq 10.01% 15.99% 24.11% 24.30%
R-sq(adj) 9.96% 15.89% 23.99% 24.13%
R-sq(pred) 9.79% 15.72% 23.79% 23.00%
Mallows’ Cp 334.69 196.27 7.4 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 11.634 11.600 11.497 10.301
Skin Temp -0.2707 0.000 -0.2730 0.000 -0.2797 0.000 -0.2528 0.000
EDA 0.6114 0.000 0.6059 0.000 0.5749 0.000
HR 0.00406 0.070 0.01149 0.001
Stress Level -0.00376 0.007
S 1.03049 1.01219 1.01154 1.00979
R-sq 7.99% 11.28% 11.45% 11.80%
R-sq(adj) 7.94% 11.18% 11.30% 11.60%
R-sq(pred) 7.77% 9.93% 10.00% 10.35%
Mallows’ Cp 75.77 11.48 10.18 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
115
-------Step 1------ -------Step 2------
Coef P Coef P
Constant 0.6299 -3.07
Lighting -0.004730 0.000 -0.003786 0.000
Temp 0.1554 0.000
S 0.316303 0.315649
R-sq 2.19% 2.64%
R-sq(adj) 2.16% 2.56%
R-sq(pred) 2.07% 2.46%
Mallows’ Cp 13.31 3.08
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2-----
Coef P Coef P
Constant 27.042 30.42
Lighting 0.02672 0.000 0.02667 0.000
CO2 -0.00551 0.001
S 1.63002 1.62728
R-sq 2.62% 2.98%
R-sq(adj) 2.58% 2.91%
R-sq(pred) 2.40% 2.71%
Mallows’ Cp 10.22 2.04
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- ----Step 2---- -----Step 3-----
Coef P Coef P Coef P
Constant -62.4 -77.6 -31.1
Temp 5.96 0.001 5.29 0.003 3.43 0.080
Acoustic 0.560 0.006 0.607 0.003
Lighting -0.0637 0.026
S 12.6809 12.6626 12.6514
R-sq 0.50% 0.83% 1.05%
R-sq(adj) 0.46% 0.74% 0.92%
R-sq(pred) 0.32% 0.57% 0.70%
Mallows’ Cp 12.59 7.00 4.01
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----
Coef P
Constant 9.2
CO2 0.0555 0.125
S 29.2868
R-sq 0.12%
R-sq(adj) 0.07%
R-sq(pred) 0.00%
Mallows’ Cp 2.11
Stepwise regression with the dataset: office zone C1 – day1 (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
116
------Step 1------ ------Step 2------ -------Step 3------ -------Step 4------
Coef P Coef P Coef P Coef P
Constant 11.946 11.386 7.963 7.976
Skin Temp -0.2744 0 -0.2891 0 -0.2055 0 -0.20559 0
HR 0.01314 0 0.02963 0 0.02971 0
Stress Level -0.0076 0 -0.00782 0
EDA -0.0577 0.033
S 0.307 0.2704 0.2501 0.2497
R-sq 53.87% 64.25% 69.45% 69.57%
R-sq(adj) 53.83% 64.19% 69.37% 69.47%
R-sq(pred) 53.75% 64.09% 69.23% 69.28%
Mallows’ Cp 598.06 204.16 7.57 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 8.225 9.099
Skin Temp -0.1508 0.000 -0.1771 0.000
EDA -0.588 0.000
S 1.22567 1.21491
R-sq 2.16% 3.96%
R-sq(adj) 2.08% 3.79%
R-sq(pred) 1.80% 0.00%
Mallows’ Cp 43.63 23.52
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -9.891 -6.396 -1.40 -1.30
Skin Temp 0.4274 0.000 0.2963 0.000 0.1900 0.000 0.1893 0.000
Stress Level 0.00980 0.000 0.02024 0.000 0.01879 0.000
HR -0.02975 0.000 -0.02914 0.000
EDA -0.433 0.000
S 1.07509 1.04117 1.02891 1.02265
R-sq 18.77% 23.88% 25.72% 26.69%
R-sq(adj) 18.70% 23.75% 25.53% 26.44%
R-sq(pred) 18.55% 23.53% 25.24% 23.45%
Mallows’ Cp 124.44 45.52 18.29 5.00
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 15.759 14.807 5.70
Skin Temp -0.3752 0.000 -0.4001 0.000 -0.1777 0.000
HR 0.02232 0.000 0.06619 0.000
Stress Level -0.02029 0.000
S 0.904312 0.870211 0.826133
R-sq 20.11% 26.08% 33.44%
R-sq(adj) 20.04% 25.96% 33.27%
117
R-sq(pred) 19.86% 25.76% 32.96%
Mallows’ Cp 233.40 131.04 4.56
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1-----
Coef P
Constant 15.700
Skin Temp -0.4100 0.000
S 0.922375
R-sq 22.41%
R-sq(adj) 22.34%
R-sq(pred) 22.15%
Mallows’ Cp 1.24
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------
Coef P Coef P
Constant 0.5402 -0.156
Lighting -0.003885 0.000 -0.003921 0.000
Acoustic 0.01274 0.019
S 0.280627 0.280231
R-sq 1.92% 2.26%
R-sq(adj) 1.86% 2.14%
R-sq(pred) 1.72% 1.87%
Mallows’ Cp 6.54 3.02
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 29.025 36.65
Lighting 0.00872 0.002 0.00666 0.036
Temp -0.320 0.128
S 1.16382 1.16334
R-sq 0.57% 0.71%
R-sq(adj) 0.51% 0.59%
R-sq(pred) 0.24% 0.27%
Mallows’ Cp 2.19 1.87
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 108.1 13.0
CO2 -0.0528 0.005 -0.0408 0.044
Temp 3.78 0.123
S 12.8844 12.8781
R-sq 0.56% 0.73%
R-sq(adj) 0.49% 0.59%
R-sq(pred) 0.28% 0.30%
Mallows’ Cp 3.34 2.96
Stepwise regression with the dataset: office zone C1 – day1 (male)
118
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.0965 3.1542 0.231 -0.103
EDA 1.0052 0 0.7993 0 0.7305 0 0.6874 0
HR 0.01333 0 0.01213 0 0.00749 0
Skin Temp 0.1008 0 0.1193 0
Stress Level 0.003029 0
S 0.2711 0.2473 0.2324 0.2269
R-sq 67.48% 72.99% 76.18% 77.33%
R-sq(adj) 67.43% 72.91% 76.07% 77.19%
R-sq(pred) 66.56% 72.16% 75.38% 76.52%
Mallows’ Cp 283.03 126.07 35.93 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 2.962 1.956 -0.918
HR 0.02393 0.000 0.04495 0.000 0.03964 0.000
Stress Level -0.01261 0.000 -0.01132 0.000
Skin Temp 0.1069 0.000
S 0.569546 0.521371 0.513615
R-sq 15.40% 29.22% 31.42%
R-sq(adj) 15.26% 28.99% 31.08%
R-sq(pred) 14.75% 28.38% 30.32%
Mallows’ Cp 142.89 21.40 3.72
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant -1.348 -3.057 -9.41 -7.35
HR 0.0727 0 0.10843 0 0.09671 0 0.0923 0
Stress Level -0.0214 0 -0.0186 0 -0.02068 0
Skin Temp 0.2361 0 0.1781 0
EDA 0.491 0
S 0.9341 0.8488 0.8249 0.8134
R-sq 38.44% 49.25% 52.15% 53.55%
R-sq(adj) 38.34% 49.08% 51.92% 53.25%
R-sq(pred) 38.04% 48.70% 51.46% 52.75%
Mallows’ Cp 198.08 57.72 21.44 5
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- -------Step 3------
Coef P Coef P Coef P
Constant -0.125 0.204 0.348
Skin Temp 0.1647 0.000 0.1272 0.000 0.1154 0.000
HR 0.01080 0.000 0.01501 0.000
Stress Level -0.002318 0.002
S 0.350969 0.335525 0.333281
119
R-sq 16.19% 23.53% 24.67%
R-sq(adj) 16.05% 23.28% 24.30%
R-sq(pred) 15.61% 22.73% 23.66%
Mallows’ Cp 68.09 10.45 3.17
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2-----
Coef P Coef P
Constant 3.5559 2.615
EDA 0.733 0.000 0.534 0.000
HR 0.01330 0.008
S 1.06421 1.05900
R-sq 6.01% 7.08%
R-sq(adj) 5.86% 6.78%
R-sq(pred) 4.93% 5.71%
Mallows’ Cp 8.09 3.01
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ ------Step 2------
Coef P Coef P
Constant 0.755 -5.43
Lighting -0.00592 0.000 -0.00449 0.000
Temp 0.2603 0.001
S 0.361345 0.359825
R-sq 2.54% 3.45%
R-sq(adj) 2.45% 3.27%
R-sq(pred) 2.18% 2.98%
Mallows’ Cp 9.95 1.60
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 24.347 34.66 6.0
Lighting 0.05113 0.000 0.04999 0.000 0.05635 0.000
CO2 -0.01668 0.000 -0.01309 0.001
Temp 1.114 0.028
S 2.05845 2.03715 2.03365
R-sq 5.67% 7.69% 8.09%
R-sq(adj) 5.58% 7.52% 7.84%
R-sq(pred) 5.07% 7.02% 7.33%
Mallows’ Cp 28.10 5.82 3.01
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 91.22 36.8
Lighting -0.1580 0.000 -0.1627 0.000
Acoustic 0.999 0.001
S 12.2714 12.2056
120
R-sq 1.71% 2.87%
R-sq(adj) 1.59% 2.65%
R-sq(pred) 1.29% 2.25%
Mallows’ Cp 10.60 2.36
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 26.53 -42.5
Lighting 0.1793 0.048 0.1795 0.047
Acoustic 1.256 0.072
S 24.8590 24.8213
R-sq 0.53% 0.96%
R-sq(adj) 0.39% 0.69%
R-sq(pred) 0.02% 0.22%
Mallows’ Cp 4.18 2.93
Stepwise regression with the dataset: office zone C1 – day1 (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3----- -------Step 4------
Coef P Coef P Coef P Coef P
Constant 3.7319 8.597 7.586 4.395
EDA 0.7894 0 0.7937 0 0.604 0 0.5876 0
Skin Temp -0.162 0 -0.1948 0 -0.1199 0
HR 0.0279 0 0.04487 0
Stress Level -0.00823 0
S 0.5728 0.5387 0.4942 0.4755
R-sq 20.66% 29.87% 41.04% 45.45%
R-sq(adj) 20.60% 29.77% 40.90% 45.29%
R-sq(pred) 9.80% 19.28% 33.07% 37.94%
Mallows’ Cp 602.56 380.23 110.4 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- -----Step 4-----
Coef P Coef P Coef P Coef P
Constant 1.521 1.251 3.716 6.18
HR 0.03754 0.000 0.04262 0.000 0.04513 0.000 0.03211 0.000
EDA -0.4524 0.000 -0.4660 0.000 -0.4482 0.000
Skin Temp -0.0880 0.001 -0.1462 0.000
Stress Level 0.00638 0.001
S 1.13582 1.12535 1.12081 1.11625
R-sq 6.59% 8.37% 9.18% 9.99%
R-sq(adj) 6.52% 8.23% 8.97% 9.71%
R-sq(pred) 6.34% 7.06% 7.60% 8.41%
Mallows’ Cp 47.81 24.20 14.59 5.00
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- -----Step 4-----
Coef P Coef P Coef P Coef P
Constant 2.5878 -4.421 -6.88 -7.04
121
Stress Level 0.02543 0.000 0.01975 0.000 0.01451 0.000 0.01423 0.000
Skin Temp 0.2400 0.000 0.2745 0.000 0.2742 0.000
HR 0.02198 0.000 0.02519 0.000
EDA -0.2293 0.007
S 1.09561 1.06711 1.06010 1.05753
R-sq 28.03% 31.78% 32.72% 33.10%
R-sq(adj) 27.97% 31.67% 32.57% 32.89%
R-sq(pred) 27.81% 31.45% 32.29% 32.60%
Mallows’ Cp 96.92 26.55 10.29 5.00
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 13.955 13.87 11.708 5.829
Skin Temp -0.314 0 -0.3143 0 -0.2346 0 -0.1541 0
EDA 0.418 0 0.4784 0 0.2821 0
Stress Level -0.0069 0 -0.01944 0
HR 0.05414 0
S 0.8933 0.881 0.867 0.8132
R-sq 15.55% 17.93% 20.57% 30.18%
R-sq(adj) 15.48% 17.80% 20.38% 29.96%
R-sq(pred) 15.32% 14.20% 16.04% 27.33%
Mallows’ Cp 269.64 227.65 180.82 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 11.954 11.84 11.2 10.073
Skin Temp -0.2794 0 -0.2797 0 -0.2999 0 -0.2733 0
EDA 0.5654 0 0.4674 0 0.4592 0
HR 0.0174 0 0.02334 0
Stress Level -0.00292 0.057
S 0.942668 0.920972 0.911655 0.910727
R-sq 11.57% 15.66% 17.42% 17.66%
R-sq(adj) 11.50% 15.53% 17.23% 17.40%
R-sq(pred) 11.26% 13.64% 15.71% 15.83%
Mallows’ Cp 94.43 32.28 6.64 5
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------
Coef P Coef P
Constant 0.8897 -3.76
Lighting -0.006745 0.000 -0.005613 0.000
Temp 0.1957 0.001
S 0.355805 0.354915
R-sq 3.28% 3.81%
R-sq(adj) 3.23% 3.71%
R-sq(pred) 3.08% 3.54%
Mallows’ Cp 11.83 3.33
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
122
----Step 1---- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 21.83 18.27 17.72
Temp 0.348 0.054 0.480 0.014 0.432 0.029
Lighting 0.00518 0.079 0.00482 0.103
Acoustic 0.0311 0.132
S 1.15035 1.14972 1.14934
R-sq 0.19% 0.36% 0.48%
R-sq(adj) 0.14% 0.25% 0.32%
R-sq(pred) 0.00% 0.01% 0.01%
Mallows’ Cp 4.51 3.41 3.14
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant -140.6 -99.0 -103.9
Temp 9.22 0.000 7.67 0.000 7.19 0.000
Lighting -0.0589 0.033 -0.0628 0.023
Acoustic 0.300 0.122
S 10.2617 10.2507 10.2463
R-sq 1.68% 1.95% 2.09%
R-sq(adj) 1.62% 1.83% 1.91%
R-sq(pred) 1.46% 1.62% 1.65%
Mallows’ Cp 7.72 5.16 4.76
Stepwise regression with the dataset: office zone C1– day1 (mid-age)
For each analysis: α to enter = 0.15, α to remove = 0.15
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2-----
Coef P Coef P
Constant 10.66 10.04
Skin Temp -0.2381 0.000 -0.2242 0.000
EDA 11.23 0.025
S 0.972209 0.968153
R-sq 3.68% 4.68%
R-sq(adj) 3.49% 4.29%
R-sq(pred) 2.90% 3.40%
Mallows’ Cp 5.21 2.15
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 3.866 1.490 -3.99
EDA -55.83 0.000 -42.65 0.000 -42.57 0.000
HR 0.02601 0.000 0.02235 0.000
Skin Temp 0.1930 0.002
S 1.10153 1.04620 1.03714
R-sq 16.83% 25.13% 26.58%
R-sq(adj) 16.66% 24.82% 26.12%
R-sq(pred) 16.13% 24.29% 25.60%
Mallows’ Cp 63.26 10.83 3.37
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
123
Coef P Coef P Coef P
Constant -2.264 -1.383 -1.257
Skin Temp 0.2379 0.000 0.1833 0.000 0.1831 0.000
HR 0.009189 0.000 0.008498 0.000
EDA -3.47 0.016
S 0.293621 0.267278 0.265954
R-sq 29.52% 41.72% 42.41%
R-sq(adj) 29.37% 41.48% 42.06%
R-sq(pred) 28.78% 40.83% 41.44%
Mallows’ Cp 107.28 7.27 3.46
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 10.64 9.59
Skin Temp -0.2430 0.001 -0.2194 0.002
EDA 19.01 0.002
S 1.22208 1.21178
R-sq 2.46% 4.30%
R-sq(adj) 2.26% 3.90%
R-sq(pred) 1.70% 3.29%
Mallows’ Cp 13.22 5.91
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------ -------Step 3-------
Coef P Coef P Coef P
Constant -0.01561 -0.4009 -0.3944
Lighting 0.000338 0.000 0.000444 0.000 0.000451 0.000
Temp 0.01616 0.000 0.01692 0.000
Acoustic -0.000449 0.052
S 0.0086981 0.0084042 0.0083897
R-sq 14.69% 20.46% 20.83%
R-sq(adj) 14.59% 20.26% 20.53%
R-sq(pred) 12.80% 18.22% 18.31%
Mallows’ Cp 62.98 6.05 4.26
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2-----
Coef P Coef P
Constant 21.301 29.61
Lighting 0.07864 0.000 0.07932 0.000
CO2 -0.01367 0.001
S 2.14476 2.13186
R-sq 13.30% 14.44%
R-sq(adj) 13.19% 14.23%
R-sq(pred) 12.12% 13.16%
Mallows’ Cp 10.98 2.11
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2-----
Coef P Coef P
Constant 1.4 14.8
Acoustic 1.469 0.002 1.554 0.001
Lighting -0.1836 0.001
S 15.7864 15.6709
R-sq 1.48% 3.07%
R-sq(adj) 1.32% 2.76%
R-sq(pred) 0.83% 1.98%
Mallows’ Cp 10.35 2.17
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
124
----Step 1---- -----Step 2---- -----Step 3----- -----Step 4----
Coef P Coef P Coef P Coef P
Constant -57.0 -170.5 -161.6 124
Acoustic 2.253 0.004 2.449 0.002 2.457 0.002 2.650 0.001
CO2 0.1677 0.004 0.1749 0.002 0.1384 0.026
Lighting -0.1407 0.135 -0.215 0.043
Temp -11.48 0.130
S 23.6136 23.4394 23.4107 23.3807
R-sq 1.61% 3.24% 3.67% 4.11%
R-sq(adj) 1.41% 2.86% 3.10% 3.35%
R-sq(pred) 0.87% 2.12% 2.08% 2.07%
Mallows’ Cp 12.17 5.55 5.30 5.00
Stepwise regression with the dataset: office zone C1 – day2
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- -------Step 3------ -------Step 4------
Coef P Coef P Coef P Coef P
Constant 3.6692 -4.106 -4.415 -10.9
EDA 0.17788 0 0.1174 0 0.14459 0 0.09297 0
Skin Temp 0.2619 0 0.2842 0 0.3816 0
Stress Level -0.0106 0 -0.02437 0
HR 0.05604 0
S 0.837388 0.797892 0.749545 0.636862
R-sq 27.57% 34.29% 42.06% 58.20%
R-sq(adj) 27.52% 34.20% 41.93% 58.08%
R-sq(pred) 27.42% 34.05% 41.72% 57.86%
Mallows’ Cp 1015.86 794.79 539.01 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- -------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.2921 4.6466 0.578 -1.819
EDA 0.10068 0 0.13302 0 0.09913 0 0.08103 0
Stress Level -0.0106 0 -0.0258 0 -0.02661 0
HR 0.06328 0 0.06564 0
Skin Temp 0.0759 0
S 0.926553 0.885154 0.757027 0.753897
R-sq 9.06% 17.06% 39.38% 39.92%
R-sq(adj) 8.99% 16.94% 39.25% 39.75%
R-sq(pred) 8.88% 16.76% 39.08% 39.52%
Mallows’ Cp 712.09 529.15 15.56 5
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 3.7834 4.1075 1.303 -6.44
EDA 0.1803 0.000 0.2098 0.000 0.1865 0.000 0.1280 0.000
Stress Level -0.00970 0.000 -0.02018 0.000 -0.02274 0.000
HR 0.04362 0.000 0.05124 0.000
Skin Temp 0.2450 0.000
S 1.17221 1.14536 1.10117 1.07694
R-sq 16.63% 20.47% 26.54% 29.79%
R-sq(adj) 16.57% 20.35% 26.38% 29.58%
R-sq(pred) 16.46% 20.14% 26.06% 29.25%
Mallows’ Cp 258.99 185.23 67.19 5.00
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- -------Step 2------ -------Step 3------ -------Step 4------
125
Coef P Coef P Coef P Coef P
Constant 4.575 4.8675 2.273 3.708
EDA 0.0604 0 0.08707 0 0.06547 0 0.0763 0
Stress Level -0.0088 0 -0.0184 0 -0.01797 0
HR 0.04035 0 0.03894 0
Skin Temp -0.0454 0
S 0.583099 0.537385 0.450789 0.448902
R-sq 8.30% 22.17% 45.27% 45.77%
R-sq(adj) 8.23% 22.06% 45.15% 45.61%
R-sq(pred) 8.12% 21.85% 44.99% 45.44%
Mallows’ Cp 957.97 604.97 15.7 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 2.8301 10.354 9.159
EDA 0.29905 0.000 0.3576 0.000 0.3405 0.000
Skin Temp -0.2534 0.000 -0.2400 0.000
HR 0.01101 0.000
S 1.01628 0.986214 0.981021
R-sq 42.21% 45.62% 46.23%
R-sq(adj) 42.17% 45.54% 46.12%
R-sq(pred) 42.08% 45.41% 45.92%
Mallows’ Cp 104.44 18.37 4.61
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- ------Step 3-----
Coef P Coef P Coef P
Constant -2.23 -7.87 -3.45
Lighting 0.0368 0.001 0.0375 0.001 0.0367 0.001
Acoustic 0.1016 0.017 0.0908 0.033
CO2 -0.00619 0.049
S 2.51275 2.51028 2.50877
R-sq 0.47% 0.71% 0.87%
R-sq(adj) 0.43% 0.63% 0.74%
R-sq(pred) 0.29% 0.43% 0.50%
Mallows’ Cp 8.72 4.96 3.07
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----
Coef P
Constant -27.20
Temp 2.480 0.000
S 2.16278
R-sq 5.55%
R-sq(adj) 5.52%
R-sq(pred) 5.41%
Mallows’ Cp 0.88
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 52.8 40.8
CO2 0.0383 0.034 0.0393 0.030
Lighting 0.1169 0.077
S 14.7271 14.7208
R-sq 0.18% 0.31%
R-sq(adj) 0.14% 0.23%
R-sq(pred) 0.02% 0.06%
Mallows’ Cp 2.38 1.26
Stress level vs IEQ
126
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2----
Coef P Coef P
Constant -21.5 -66.1
Acoustic 1.093 0.053 1.203 0.034
CO2 0.0639 0.128
S 30.3399 30.3302
R-sq 0.18% 0.30%
R-sq(adj) 0.14% 0.20%
R-sq(pred) 0.00% 0.00%
Mallows’ Cp 3.63 3.31
Stepwise regression with the dataset: office zone C1 – day2 (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-------Step 1------
-------
Step 2---
---
-------Step 3------ -------Step 4------
Coef P Coef P Coef P Coef P
Constant 3.993 3.6424 3.355 1.837
Stress Level -0.0185 0 -0.0136 0 -0.0148 0 -0.01544 0
EDA 0.2734 0 0.2509 0 0.2521 0
HR 0.00482 0.008 0.00799 0
Skin Temp 0.0444 0.004
S 0.366757 0.326843 0.325325 0.323498
R-sq 61.75% 69.67% 70.00% 70.38%
R-sq(adj) 61.69% 69.57% 69.86% 70.19%
R-sq(pred) 61.51% 69.26% 69.44% 69.72%
Mallows’ Cp 184.61 16.28 11.22 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 3.5956 0.712 -0.513 3
EDA 0.6525 0 0.663 0 0.1596 0.012 0.1568 0.013
HR 0.03977 0 0.07519 0 0.06785 0
Stress Level -0.026 0 -0.02451 0
Skin Temp -0.1028 0.017
S 1.05603 0.998999 0.903326 0.900013
R-sq 18.04% 26.76% 40.21% 40.74%
R-sq(adj) 17.91% 26.54% 39.93% 40.37%
R-sq(pred) 17.68% 26.33% 39.70% 40.01%
Mallows’ Cp 243.13 151.29 8.71 5
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 4.9055 4.155 9.97 11.91
Stress Level -0.0288 0 -0.0183 0 -0.019 0 -0.01629 0
EDA 0.5855 0 0.5158 0 0.5529 0
Skin Temp -0.194 0 -0.2366 0
HR -0.01125 0.092
S 1.09185 1.03249 1.01607 1.0146
R-sq 30.60% 38.04% 40.09% 40.36%
R-sq(adj) 30.50% 37.85% 39.81% 39.98%
R-sq(pred) 30.21% 37.50% 39.28% 39.33%
Mallows’ Cp 103.17 25.73 5.85 5
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
127
------Step 1----- -------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 14.728 14.531 7.931 7.953
Skin Temp -0.3517 0 -0.3289 0 -0.1822 0 -0.1811 0
Stress Level -0.0118 0 -0.018 0 -0.01592 0
HR 0.03467 0 0.03207 0
EDA 0.0892 0.018
S 0.653635 0.583591 0.541124 0.539181
R-sq 22.25% 38.12% 46.88% 47.34%
R-sq(adj) 22.13% 37.93% 46.63% 47.01%
R-sq(pred) 21.63% 37.53% 46.28% 46.70%
Mallows’ Cp 302.55 112.6 8.61 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.7016 1.037 5.31 4.32
EDA 0.4777 0 0.4838 0 0.4448 0 0.3664 0
HR 0.02295 0 0.01636 0 0.02358 0
Skin Temp -0.1274 0.001 -0.1042 0.007
Stress Level -0.00441 0.029
S 0.837658 0.814334 0.807555 0.80515
R-sq 15.79% 20.53% 21.97% 22.56%
R-sq(adj) 15.65% 20.29% 21.61% 22.07%
R-sq(pred) 15.43% 19.98% 21.23% 21.49%
Mallows’ Cp 54.72 17.66 7.82 5
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1-----
Coef P
Constant 23.33
Temp -1.000 0.000
S 0.820277
R-sq 6.10%
R-sq(adj) 6.01%
R-sq(pred) 5.72%
Mallows’ Cp 2.79
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- -----Step 3----- -----Step 4-----
Coef P Coef P Coef P Coef P
Constant -32.53 -44.53 -46.71 -38.93
Temp 2.706 0.000 2.541 0.000 2.791 0.000 2.740 0.000
CO2 0.02611 0.000 0.02566 0.000 0.02465 0.000
Lighting -0.0336 0.003 -0.0339 0.003
Acoustic -0.1092 0.010
S 1.88187 1.83407 1.82884 1.82496
R-sq 8.50% 13.16% 13.71% 14.14%
R-sq(adj) 8.43% 13.03% 13.52% 13.89%
R-sq(pred) 8.30% 12.79% 13.17% 13.42%
Mallows’ Cp 87.58 16.51 9.74 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- -----Step 3-----
Coef P Coef P Coef P
Constant 38.4 69.0 83.7
CO2 0.0601 0.013 0.0557 0.023 0.0552 0.024
Acoustic -0.509 0.122 -0.535 0.105
Lighting -0.1351 0.119
128
S 15.0400 15.0329 15.0255
R-sq 0.42% 0.58% 0.74%
R-sq(adj) 0.35% 0.44% 0.54%
R-sq(pred) 0.16% 0.20% 0.26%
Mallows’ Cp 3.87 3.48 3.04
Stepwise regression with the dataset: office zone C1 – day2 (male)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant -8.228 -8.049 -10.176 -10.509
Skin Temp 0.4212 0 0.3781 0 0.4451 0 0.445 0
HR 0.01484 0 0.01728 0 0.02281 0
EDA -0.0363 0 -0.02943 0.001
Stress Level -0.00284 0.091
S 0.605583 0.585423 0.578422 0.577701
R-sq 40.04% 44.04% 45.44% 45.65%
R-sq(adj) 39.96% 43.89% 45.23% 45.36%
R-sq(pred) 39.69% 43.55% 44.89% 44.88%
Mallows’ Cp 76.04 23.13 5.87 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -0.268 -2.213 -2.351
Skin Temp 0.1698 0.000 0.2365 0.000 0.2345 0.000
EDA -0.03271 0.000 -0.03558 0.000
HR 0.00270 0.031
S 0.358754 0.348360 0.347507
R-sq 23.61% 28.07% 28.52%
R-sq(adj) 23.51% 27.88% 28.23%
R-sq(pred) 22.48% 26.79% 27.10%
Mallows’ Cp 50.21 5.69 3.02
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant -14.882 -14.602 -17.5 -17.95
Skin Temp 0.6326 0 0.5654 0 0.6566 0 0.6565 0
HR 0.02316 0 0.02648 0 0.03402 0
EDA -0.0494 0 -0.0401 0.004
Stress Level -0.00387 0.123
S 0.90645 0.873543 0.864938 0.86414
R-sq 40.20% 44.54% 45.70% 45.87%
R-sq(adj) 40.12% 44.39% 45.48% 45.58%
R-sq(pred) 39.85% 44.05% 45.15% 45.11%
Mallows’ Cp 77.14 19.37 5.38 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 2.8879 12.91 13.10 14.93
EDA 0.2876 0.000 0.3616 0.000 0.3140 0.000 0.2958 0.000
Skin Temp -0.3355 0.000 -0.3502 0.000 -0.3430 0.000
Stress Level 0.01062 0.000 0.02219 0.000
HR -0.03163 0.000
S 1.14006 1.10270 1.07491 1.06167
R-sq 44.84% 48.46% 51.09% 52.35%
129
R-sq(adj) 44.76% 48.32% 50.90% 52.10%
R-sq(pred) 44.58% 48.05% 50.53% 51.61%
Mallows’ Cp 116.70 61.94 22.75 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -5.29 4.74 -5.40 -20.4
Lighting 0.0741 0.000 0.0733 0.000 0.0750 0.000 0.0647 0.001
CO2 -0.01646 0.002 -0.01495 0.005 -0.01563 0.004
Acoustic 0.1655 0.023 0.1735 0.017
Temp 0.702 0.111
S 3.16901 3.15878 3.15381 3.15198
R-sq 1.15% 1.87% 2.25% 2.43%
R-sq(adj) 1.08% 1.72% 2.03% 2.14%
R-sq(pred) 0.84% 1.38% 1.59% 1.69%
Mallows’ Cp 16.43 8.75 5.55 5.00
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -22.13 -12.84 -22.26 -20.70
Temp 2.266 0.000 2.382 0.000 2.440 0.000 2.255 0.000
CO2 -0.01974 0.000 -0.01855 0.000 -0.01823 0.000
Acoustic 0.1352 0.013 0.1364 0.013
Lighting 0.0244 0.098
S 2.39649 2.37571 2.37119 2.36965
R-sq 3.83% 5.56% 5.99% 6.18%
R-sq(adj) 3.76% 5.42% 5.78% 5.90%
R-sq(pred) 3.49% 5.09% 5.40% 5.45%
Mallows’ Cp 32.64 9.88 5.75 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- ----Step 2----
Coef P Coef P
Constant 34.01 -8.5
Lighting 0.448 0.000 0.458 0.000
Acoustic 0.759 0.034
S 13.9351 13.9108
R-sq 1.94% 2.38%
R-sq(adj) 1.84% 2.19%
R-sq(pred) 1.50% 1.72%
Mallows’ Cp 5.24 2.75
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2---- ----Step 3----
Coef P Coef P Coef P
Constant -11.8 -114.2 -291
Lighting 0.495 0.025 0.518 0.019 0.435 0.055
Acoustic 1.831 0.023 1.948 0.016
Temp 7.83 0.144
S 29.2973 29.2271 29.2080
R-sq 0.58% 1.17% 1.41%
R-sq(adj) 0.47% 0.94% 1.07%
R-sq(pred) 0.14% 0.49% 0.51%
Mallows’ Cp 7.25 4.06 3.91
Stepwise regression with the dataset: office zone C1 – day2 (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- -------Step 2------ -------Step 3------ -------Step 4------
130
Coef P Coef P Coef P Coef P
Constant 3.8952 4.2616 0.253 -8.318
EDA 0.1433 0 0.17481 0 0.1319 0 0.07815 0
Stress Level -0.0107 0 -0.0233 0 -0.02619 0
HR 0.06154 0 0.06759 0
Skin Temp 0.2753 0
S 0.775596 0.730602 0.592514 0.545849
R-sq 23.94% 32.56% 55.68% 62.42%
R-sq(adj) 23.87% 32.45% 55.57% 62.29%
R-sq(pred) 23.74% 32.23% 55.39% 62.09%
Mallows’ Cp 1220.67 948.84 216.88 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.2045 4.7189 0.153 -3.329
EDA 0.11408 0 0.15832 0 0.10945 0 0.08761 0
Stress Level -0.0151 0 -0.0294 0 -0.03056 0
HR 0.07009 0 0.07254 0
Skin Temp 0.1118 0
S 0.937779 0.863479 0.713197 0.70729
R-sq 12.00% 25.46% 49.19% 50.07%
R-sq(adj) 11.93% 25.33% 49.06% 49.90%
R-sq(pred) 11.79% 25.13% 48.90% 49.69%
Mallows’ Cp 908.51 589.06 24.03 5
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- -------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.254 4.6856 0.372 5.37
EDA 0.1083 0 0.14542 0 0.09924 0 0.13059 0
Stress Level -0.0126 0 -0.0262 0 -0.0245 0
HR 0.06622 0 0.0627 0
Skin Temp -0.1605 0
S 0.886362 0.831596 0.692842 0.679901
R-sq 12.10% 22.69% 46.38% 48.41%
R-sq(adj) 12.02% 22.56% 46.25% 48.24%
R-sq(pred) 11.88% 22.36% 46.07% 48.05%
Mallows’ Cp 838.65 595.73 49.89 5
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- -------Step 2------ -------Step 3------ -------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.474 4.8524 2.017 1.145
EDA 0.07584 0 0.10838 0 0.07804 0 0.07257 0
Stress Level -0.0111 0 -0.02 0 -0.02027 0
HR 0.04352 0 0.04413 0
Skin Temp 0.028 0.066
S 0.599268 0.535346 0.441861 0.441419
R-sq 12.87% 30.52% 52.71% 52.84%
R-sq(adj) 12.79% 30.40% 52.59% 52.68%
R-sq(pred) 12.65% 30.17% 52.42% 52.51%
Mallows’ Cp 1010.3 565.66 6.39 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
131
Coef P Coef P Coef P Coef P
Constant 2.5315 1.532 1.728 3.189
EDA 0.34469 0 0.32671 0 0.32492 0 0.3341 0
HR 0.01386 0 0.01006 0.003 0.00903 0.009
Stress Level 0.00229 0.056 0.00278 0.024
Skin Temp -0.0469 0.097
S 0.828993 0.820833 0.819923 0.819321
R-sq 61.45% 62.23% 62.35% 62.43%
R-sq(adj) 61.41% 62.17% 62.25% 62.31%
R-sq(pred) 61.32% 62.03% 62.08% 62.11%
Mallows’ Cp 30.4 7.42 5.76 5
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -2.95 4.66 -3.99
Lighting 0.0490 0.001 0.0483 0.001 0.0498 0.000
CO2 -0.01250 0.002 -0.01121 0.006
Acoustic 0.1413 0.010
S 2.77305 2.76631 2.76197
R-sq 0.67% 1.20% 1.57%
R-sq(adj) 0.61% 1.09% 1.40%
R-sq(pred) 0.42% 0.84% 1.07%
Mallows’ Cp 16.10 8.38 3.79
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2----
Coef P Coef P
Constant 6.01 0.04
Temp 1.053 0.000 1.101 0.000
Acoustic 0.0893 0.002
S 1.50075 1.49738
R-sq 2.14% 2.64%
R-sq(adj) 2.09% 2.53%
R-sq(pred) 1.87% 2.27%
Mallows’ Cp 10.92 3.75
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----
Coef P
Constant 52.99
Lighting 0.2062 0.001
S 12.0856
R-sq 0.59%
R-sq(adj) 0.54%
R-sq(pred) 0.34%
Mallows’ Cp 2.57
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- ----Step 2---- -----Step 3----
Coef P Coef P Coef P
Constant -8.4 -93.4 -271.7
Lighting 0.441 0.003 0.454 0.002 0.344 0.028
Acoustic 1.530 0.007 1.656 0.003
Temp 8.01 0.019
S 27.1941 27.1408 27.1030
R-sq 0.53% 0.98% 1.32%
R-sq(adj) 0.47% 0.86% 1.14%
R-sq(pred) 0.30% 0.63% 0.84%
Mallows’ Cp 12.14 6.75 3.22
Stepwise regression with the dataset: office zone C1 – day2 (mid-age)
132
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 2.1804 2.4990 3.313
EDA 43.49 0.000 30.62 0.000 31.83 0.000
Stress Level -0.005217 0.000 -0.005303 0.000
Skin Temp -0.0287 0.060
S 0.249552 0.208355 0.206971
R-sq 69.22% 78.65% 79.05%
R-sq(adj) 69.06% 78.43% 78.72%
R-sq(pred) 68.33% 77.72% 77.85%
Mallows’ Cp 88.20 4.62 3.06
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2-----
Coef P Coef P
Constant -3.21 -5.72
Skin Temp 0.2735 0.000 0.3522 0.000
Stress Level 0.00811 0.000
S 0.694381 0.644736
R-sq 15.14% 27.22%
R-sq(adj) 14.70% 26.46%
R-sq(pred) 10.07% 21.81%
Mallows’ Cp 30.80 1.23
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -------Step 2------ -------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 1.4427 1.6951 3.204 3.684
EDA 29.46 0 19.27 0 21.51 0 22.05 0
Stress Level -0.0041 0 -0.0043 0 -0.00261 0.037
Skin Temp -0.0532 0.01 -0.0602 0.004
HR -0.00463 0.09
S 0.301776 0.281943 0.277768 0.276396
R-sq 41.37% 49.09% 50.84% 51.58%
R-sq(adj) 41.07% 48.56% 50.07% 50.56%
R-sq(pred) 39.66% 46.96% 48.22% 48.37%
Mallows’ Cp 39.07 10.78 5.9 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1-----
Coef P
Constant 3.476
HR 0.00835 0.132
S 1.09819
R-sq 1.17%
R-sq(adj) 0.66%
R-sq(pred) 0.00%
Mallows’ Cp 0.13
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ -------Step 2-------
Coef P Coef P
Constant -0.4467 -0.3915
Temp 0.02012 0.000 0.02012 0.000
CO2 -0.000091 0.000
S 0.0096247 0.0095099
R-sq 15.73% 17.86%
R-sq(adj) 15.59% 17.59%
133
R-sq(pred) 15.38% 17.16%
Mallows’ Cp 15.52 1.38
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- -----Step 3----- -----Step 4-----
Coef P Coef P Coef P Coef P
Constant -92.93 -100.37 -102.81 -94.10
Temp 5.304 0.000 5.217 0.000 5.490 0.000 5.441 0.000
CO2 0.01561 0.001 0.01516 0.002 0.01406 0.004
Lighting -0.0360 0.043 -0.0365 0.040
Acoustic -0.1256 0.060
S 2.35648 2.34402 2.33992 2.33654
R-sq 18.62% 19.57% 19.94% 20.26%
R-sq(adj) 18.53% 19.39% 19.67% 19.90%
R-sq(pred) 18.30% 19.02% 19.19% 19.28%
Mallows’ Cp 17.12 8.65 6.55 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2----
Coef P Coef P
Constant 290.9 253.4
Temp -9.05 0.013 -9.92 0.007
CO2 0.0948 0.040
S 18.0231 17.9753
R-sq 1.01% 1.70%
R-sq(adj) 0.85% 1.38%
R-sq(pred) 0.39% 0.75%
Mallows’ Cp 4.24 1.99
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- -----Step 3----
Coef P Coef P Coef P
Constant 174.4 -15.8 447
Lighting -1.237 0.001 -1.311 0.000 -0.972 0.011
CO2 0.326 0.003 0.356 0.001
Temp -22.59 0.010
S 35.9050 35.5546 35.3116
R-sq 2.73% 4.85% 6.38%
R-sq(adj) 2.49% 4.39% 5.69%
R-sq(pred) 1.78% 3.34% 4.49%
Mallows’ Cp 14.89 7.65 3.00
Stepwise regression with the dataset: office zone C1 – day3
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.9669 -6.877 -4.973 -4.962
Stress Level 0.0194 0 0.01681 0 0.0289 0 0.02813 0
Skin Temp 0.3355 0 0.3494 0 0.3602 0
HR
-
0.0373
0 -0.0397 0
EDA -1.358 0
S 1.09352 1.00213 0.973265 0.965703
R-sq 22.24% 34.73% 38.48% 39.46%
R-sq(adj) 22.19% 34.66% 38.37% 39.32%
R-sq(pred) 22.07% 34.51% 38.17% 39.08%
Mallows’ Cp 474.74 131.67 30.31 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
134
Coef P Coef P Coef P Coef P
Constant 3.4629 -3.355 -3.466 -2.682
Stress Level 0.01728 0 0.01542 0 0.01292 0 0.0178 0
Skin Temp 0.2327 0 0.2465 0 0.2532 0
EDA -2.226 0 -2.361 0
HR -0.01551 0
S 1.07682 1.03348 1.0133 1.00867
R-sq 19.19% 25.61% 28.53% 29.23%
R-sq(adj) 19.14% 25.52% 28.40% 29.05%
R-sq(pred) 19.02% 25.38% 27.82% 28.42%
Mallows’ Cp 230.82 84.51 19.04 5
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -6.470 -7.035 -9.720
Skin Temp 0.3436 0.000 0.3020 0.000 0.3120 0.000
HR 0.02404 0.000 0.06648 0.000
Stress Level -0.01994 0.000
S 1.31671 1.28692 1.24392
R-sq 10.71% 14.75% 20.40%
R-sq(adj) 10.65% 14.65% 20.26%
R-sq(pred) 10.51% 14.43% 20.02%
Mallows’ Cp 199.16 118.00 3.88
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 0.14 -6.079 -5.792 -6.242
HR 0.05295 0 0.04816 0 0.04078 0 0.04783 0
Skin Temp 0.2221 0 0.2387 0 0.2408 0
EDA -2.515 0 -2.577 0
Stress Level -0.00339 0.004
S 0.876232 0.82741 0.794812 0.793069
R-sq 34.03% 41.21% 45.79% 46.06%
R-sq(adj) 33.99% 41.14% 45.69% 45.93%
R-sq(pred) 33.90% 41.03% 45.34% 45.54%
Mallows’ Cp 363.54 147.84 11.2 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------ ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 2.5117 -9.074 -9.165 -9.986
Stress Level 0.0247 0 0.02155 0 0.0195 0 0.01439 0
Skin Temp 0.3954 0 0.4068 0 0.3998 0
EDA -1.822 0 -1.68 0
HR 0.01625 0
S 1.08637 0.956404 0.941889 0.936347
R-sq 32.30% 47.56% 49.17% 49.80%
R-sq(adj) 32.26% 47.50% 49.08% 49.68%
R-sq(pred) 32.15% 47.36% 48.86% 49.43%
Mallows’ Cp 568.92 73.88 23.43 5
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ -------Step 2------
Coef P Coef P
Constant -0.4033 -0.4131
CO2 0.000729 0.000 0.000821 0.000
135
Lighting -0.000564 0.088
S 0.104688 0.104649
R-sq 2.26% 2.37%
R-sq(adj) 2.22% 2.29%
R-sq(pred) 2.14% 2.17%
Mallows’ Cp 3.27 2.35
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----
Coef P
Constant -27.66
Temp 2.546 0.000
S 1.45965
R-sq 9.65%
R-sq(adj) 9.62%
R-sq(pred) 9.54%
Mallows’ Cp 1.42
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 60.84 113.2
Lighting 0.1647 0.000 0.1536 0.000
Temp -2.29 0.094
S 13.0495 13.0449
R-sq 0.80% 0.90%
R-sq(adj) 0.76% 0.83%
R-sq(pred) 0.65% 0.68%
Mallows’ Cp 3.00 2.19
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----
Coef P
Constant 19.61
Lighting 0.2026 0.012
S 28.2768
R-sq 0.27%
R-sq(adj) 0.23%
R-sq(pred) 0.10%
Mallows’ Cp 3.46
Stepwise regression with the dataset: office zone C1 – day3 (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 2.5714 8.278 7.854 8.955
Stress Level 0.01873 0 0.04562 0 0.04657 0 0.046504 0
HR -0.0889 0 -0.0853 0 -0.08467 0
EDA 1.313 0 1.403 0
Skin Temp -0.0397 0.013
S 0.724764 0.418215 0.410834 0.409596
R-sq 43.25% 81.12% 81.81% 81.94%
R-sq(adj) 43.18% 81.08% 81.74% 81.85%
R-sq(pred) 43.03% 81.00% 81.63% 81.71%
Mallows’ Cp 1853.83 39.92 9.25 5
Acoustic satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
136
Constant -2.023 -0.493 -7.518 -6.95
HR 0.07148 0 0.05567 0 0.05234 0 0.03938 0
EDA -3.621 0 -4.221 0 -3.826 0
Skin Temp 0.2518 0 0.2565 0
Stress Level 0.00632 0
S 0.845124 0.815304 0.785402 0.777543
R-sq 51.57% 54.98% 58.27% 59.15%
R-sq(adj) 51.52% 54.88% 58.13% 58.96%
R-sq(pred) 51.39% 54.69% 57.95% 58.74%
Mallows’ Cp 159.63 89.35 21.62 5
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 11.07 11.56 10.08 7.88
Skin Temp -0.2646 0.000 -0.2901 0.000 -0.3409 0.000 -0.3590 0.000
EDA 2.763 0.000 6.220 0.000 4.696 0.000
HR 0.03492 0.000 0.08501 0.000
Stress Level -0.02445 0.000
S 1.18263 1.16178 1.11747 1.02755
R-sq 3.80% 7.27% 14.30% 27.62%
R-sq(adj) 3.69% 7.05% 14.01% 27.29%
R-sq(pred) 3.42% 6.59% 13.45% 26.75%
Mallows’ Cp 284.09 244.58 162.38 5.00
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant -0.963 0.771 -0.218 -4.342
HR 0.06347 0 0.04555 0 0.06377 0 0.06127 0
EDA -4.105 0 -4.678 0 -5.015 0
Stress Level -0.0089 0 -0.00868 0
Skin Temp 0.1488 0
S 0.735748 0.690841 0.672187 0.659822
R-sq 52.56% 58.22% 60.49% 61.97%
R-sq(adj) 52.50% 58.12% 60.35% 61.80%
R-sq(pred) 52.38% 57.90% 60.09% 61.53%
Mallows’ Cp 213.5 86.53 36.8 5
Thermal satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 2.2311 0.178 -2.99 -3
Stress Level 0.02649 0 0.01681 0 0.01719 0 0.0177 0
HR 0.03197 0 0.03078 0 0.03285 0
Skin Temp 0.1113 0.001 0.1035 0.002
EDA 0.737 0.14
S 0.908928 0.884122 0.879115 0.878515
R-sq 49.21% 52.00% 52.60% 52.72%
R-sq(adj) 49.15% 51.89% 52.43% 52.50%
R-sq(pred) 49.01% 51.68% 52.16% 52.13%
Mallows’ Cp 63.21 14.12 5.18 5
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant -3.198 -3.320 -3.524
Temp 0.1466 0.000 0.1360 0.000 0.1463 0.000
CO2 0.000532 0.000 0.000357 0.002
137
Lighting 0.000987 0.005
S 0.0761661 0.0753332 0.0751350
R-sq 11.62% 13.61% 14.13%
R-sq(adj) 11.55% 13.47% 13.93%
R-sq(pred) 11.28% 13.16% 13.54%
Mallows’ Cp 36.83 8.97 3.15
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2------
Coef P Coef P
Constant -32.10 -30.38
Temp 2.727 0.000 2.872 0.000
CO2 -0.00736 0.000
S 0.835951 0.821013
R-sq 27.22% 29.85%
R-sq(adj) 27.17% 29.75%
R-sq(pred) 27.04% 29.51%
Mallows’ Cp 49.23 2.08
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- -----Step 3---- -----Step 4----
Coef P Coef P Coef P Coef P
Constant 44.96 22.5 114.1 150.6
Lighting 0.3361 0.000 0.2736 0.000 0.2311 0.000 0.2229 0.000
CO2 0.0418 0.020 0.0549 0.004 0.0502 0.009
Temp -4.30 0.026 -4.68 0.016
Acoustic -0.442 0.145
S 13.2986 13.2788 13.2609 13.2559
R-sq 3.05% 3.40% 3.72% 3.86%
R-sq(adj) 2.98% 3.27% 3.53% 3.60%
R-sq(pred) 2.77% 3.01% 3.19% 3.20%
Mallows’ Cp 11.52 8.11 5.13 5.00
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1----
Coef P
Constant 9.3
Lighting 0.311 0.005
S 29.8677
R-sq 0.56%
R-sq(adj) 0.49%
R-sq(pred) 0.27%
Mallows’ Cp 1.73
Stepwise regression with the dataset: office zone C1 – day3 (male)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -9.760 -9.006 -8.924 -10.384
Skin Temp 0.4625 0.000 0.3375 0.000 0.3468 0.000 0.3454 0.000
HR 0.04098 0.000 0.03870 0.000 0.06683 0.000
EDA -2.094 0.000 -2.024 0.000
Stress Level -0.01420 0.000
S 1.13009 1.04938 1.02352 1.00780
R-sq 26.98% 37.11% 40.25% 42.14%
R-sq(adj) 26.89% 36.96% 40.03% 41.86%
R-sq(pred) 26.57% 36.59% 39.54% 41.23%
Mallows’ Cp 208.96 70.64 29.22 5.00
Lighting satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4-----
138
Coef P Coef P Coef P Coef P
Constant -0.853 -9.747 -9.575 -11.071
HR 0.06709 0.000 0.04879 0.000 0.04654 0.000 0.07419 0.000
Skin Temp 0.3395 0.000 0.3460 0.000 0.3469 0.000
EDA -2.282 0.000 -2.181 0.000
Stress Level -0.01406 0.000
S 1.27415 1.18579 1.15805 1.14457
R-sq 26.67% 36.57% 39.58% 41.06%
R-sq(adj) 26.57% 36.41% 39.35% 40.75%
R-sq(pred) 26.30% 36.05% 38.95% 40.22%
Mallows’ Cp 185.51 59.17 22.12 5.00
IAQ satisfaction vs bio-signals
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 1.119 -4.397 -5.514
HR 0.04442 0.000 0.03306 0.000 0.05373 0.000
Skin Temp 0.2106 0.000 0.2114 0.000
Stress Level -0.01055 0.000
EDA
S 0.876424 0.827465 0.816755
R-sq 25.20% 33.41% 35.21%
R-sq(adj) 25.10% 33.24% 34.95%
R-sq(pred) 24.82% 32.87% 34.44%
Mallows’ Cp 138.29 41.14 21.44
------Step 4------
Coef P
Constant -5.365
HR 0.05143 0.000
Skin Temp 0.2145 0.000
Stress Level -0.00993 0.000
EDA -1.109 0.000
S 0.807603
R-sq 36.73%
R-sq(adj) 36.40%
R-sq(pred) 35.83%
Mallows’ Cp 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -14.926 -14.495 -14.379
Skin Temp 0.6196 0.000 0.5518 0.000 0.5562 0.000
HR 0.02187 0.000 0.02036 0.000
EDA -1.530 0.000
S 0.921480 0.893259 0.876863
R-sq 50.78% 53.81% 55.55%
R-sq(adj) 50.72% 53.69% 55.38%
R-sq(pred) 50.50% 53.42% 54.96%
Mallows’ Cp 81.28 31.20 3.31
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 2.075 1.871 2.272
Temp -0.0885 0.000 -0.1059 0.000 -0.1261 0.000
CO2 0.000877 0.000 0.001221 0.000
Lighting -0.001932 0.001
S 0.124678 0.123285 0.122806
R-sq 1.74% 4.00% 4.82%
R-sq(adj) 1.67% 3.85% 4.60%
139
R-sq(pred) 1.49% 3.68% 4.30%
Mallows’ Cp 41.60 12.68 3.46
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2-----
Coef P Coef P
Constant -23.22 -24.96
Temp 2.366 0.000 2.218 0.000
CO2 0.00747 0.001
S 1.80729 1.80076
R-sq 5.68% 6.43%
R-sq(adj) 5.61% 6.29%
R-sq(pred) 5.43% 5.96%
Mallows’ Cp 11.27 2.73
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1-----
Coef P
Constant 112.7
CO2 -0.0540 0.002
S 12.5093
R-sq 0.88%
R-sq(adj) 0.79%
R-sq(pred) 0.50%
Mallows’ Cp -0.43
Stepwise regression with the dataset: office zone C1 – day3 (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 3.0072 8.174 1.211
Stress Level 0.03061 0.000 0.05367 0.000 0.04902 0.000
HR -0.08073 0.000 -0.08183 0.000
Skin Temp 0.2405 0.000
EDA
S 0.985584 0.845490 0.792718
R-sq 38.66% 54.90% 60.39%
R-sq(adj) 38.61% 54.82% 60.29%
R-sq(pred) 38.47% 54.67% 60.11%
Mallows’ Cp 874.31 335.69 154.95
------Step 4------
Coef P
Constant 0.426
Stress Level 0.04793 0.000
HR -0.08575 0.000
Skin Temp 0.2898 0.000
EDA -3.446 0.000
S 0.745977
R-sq 64.95%
R-sq(adj) 64.83%
R-sq(pred) 64.64%
Mallows’ Cp 5.00
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -12.177 -13.183 -10.880 -11.654
Skin Temp 0.5360 0.000 0.5946 0.000 0.5036 0.000 0.5006 0.000
EDA -7.071 0.000 -6.471 0.000 -6.373 0.000
Stress Level 0.01118 0.000 0.00748 0.000
140
HR 0.01326 0.001
S 1.05419 0.888647 0.850044 0.846249
R-sq 31.77% 51.56% 55.72% 56.15%
R-sq(adj) 31.71% 51.48% 55.60% 56.00%
R-sq(pred) 31.57% 51.32% 55.43% 55.79%
Mallows’ Cp 627.22 119.27 14.17 5.00
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 0.404 -5.159 -5.528
HR 0.05226 0.000 0.04315 0.000 0.03856 0.000
Skin Temp 0.2086 0.000 0.2417 0.000
EDA -2.726 0.000
Stress Level
S 1.01243 0.977455 0.953990
R-sq 20.23% 25.71% 29.30%
R-sq(adj) 20.16% 25.58% 29.11%
R-sq(pred) 19.93% 25.30% 28.75%
Mallows’ Cp 158.61 71.98 16.00
------Step 4------
Coef P
Constant -6.827
HR 0.05004 0.000
Skin Temp 0.2640 0.000
EDA -2.805 0.000
Stress Level -0.00626 0.000
S 0.948970
R-sq 30.10%
R-sq(adj) 29.85%
R-sq(pred) 29.42%
Mallows’ Cp 5.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant -8.262 -9.113 -9.137 -10.965
Skin Temp 0.4142 0.000 0.4637 0.000 0.3936 0.000 0.4249 0.000
EDA -5.976 0.000 -5.348 0.000 -5.459 0.000
HR 0.02809 0.000 0.04426 0.000
Stress Level -0.00881 0.000
S 0.899151 0.760793 0.717098 0.702971
R-sq 27.65% 48.25% 54.06% 55.90%
R-sq(adj) 27.59% 48.16% 53.94% 55.74%
R-sq(pred) 27.42% 47.99% 53.75% 55.50%
Mallows’ Cp 722.62 196.88 49.92 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -12.855 -9.628 -10.202
Skin Temp 0.5442 0.000 0.4191 0.000 0.4179 0.000
Stress Level 0.01626 0.000 0.01357 0.000
HR 0.00950 0.027
S 0.997569 0.920732 0.919150
R-sq 34.90% 44.59% 44.83%
R-sq(adj) 34.84% 44.49% 44.69%
R-sq(pred) 34.68% 44.27% 44.41%
Mallows’ Cp 202.44 5.93 3.04
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
141
------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant -3.039 -3.241 -3.476
Temp 0.13995 0.000 0.12281 0.000 0.1347 0.000
CO2 0.000867 0.000 0.000666 0.000
Lighting 0.001135 0.000
S 0.0772193 0.0749438 0.0746664
R-sq 10.34% 15.60% 16.27%
R-sq(adj) 10.29% 15.50% 16.13%
R-sq(pred) 10.11% 15.29% 15.86%
Mallows’ Cp 123.28 15.90 3.92
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2-----
Coef P Coef P
Constant -9.81 -10.52
Temp 1.763 0.000 1.703 0.000
CO2 0.00305 0.090
S 1.62720 1.62633
R-sq 3.96% 4.12%
R-sq(adj) 3.90% 4.01%
R-sq(pred) 3.78% 3.76%
Mallows’ Cp 3.78 2.90
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----
Coef P
Constant 306.2
Temp -10.35 0.000
S 10.9358
R-sq 3.19%
R-sq(adj) 3.13%
R-sq(pred) 2.99%
Mallows’ Cp 1.07
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 483.9 469.1
Temp -20.12 0.000 -21.42 0.000
CO2 0.0651 0.015
S 23.6531 23.6200
R-sq 2.49% 2.82%
R-sq(adj) 2.43% 2.70%
R-sq(pred) 2.28% 2.49%
Mallows’ Cp 5.93 2.06
Stepwise regression with the dataset: office zone C1 – day3 (mid-age)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 1.7140 1.8920 4.995
Stress Level 0.025403 0.000 0.023417 0.000 0.023516 0.000
EDA -1.113 0.000 -1.465 0.000
Skin Temp -0.1063 0.000
HR
S 0.548759 0.536075 0.525479
R-sq 69.21% 70.68% 71.88%
R-sq(adj) 69.15% 70.56% 71.71%
R-sq(pred) 69.04% 69.92% 70.71%
142
Mallows’ Cp 73.88 48.53 28.05
------Step 4-----
Coef P
Constant 4.104
Stress Level 0.01416 0.000
EDA -1.418 0.000
Skin Temp -0.1170 0.000
HR 0.02201 0.000
S 0.513305
R-sq 73.22%
R-sq(adj) 73.01%
R-sq(pred) 71.99%
Mallows’ Cp 5.00
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 2.542 1.298
Skin Temp 0.0790 0.000 0.1199 0.000
EDA 0.974 0.000
S 0.540339 0.529557
R-sq 2.54% 6.58%
R-sq(adj) 2.35% 6.20%
R-sq(pred) 2.01% 5.32%
Mallows’ Cp 33.23 13.11
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 1.4299 10.26 11.75 10.79
Stress Level 0.01975 0.000 0.02186 0.000 0.01974 0.000 0.00974 0.012
Skin Temp -0.3079 0.000 -0.3523 0.000 -0.3638 0.000
EDA -1.356 0.002 -1.305 0.003
HR 0.02356 0.005
S 1.04652 0.993125 0.984502 0.977785
R-sq 27.20% 34.57% 35.83% 36.83%
R-sq(adj) 27.06% 34.31% 35.45% 36.33%
R-sq(pred) 26.73% 33.86% 34.93% 35.77%
Mallows’ Cp 75.20 18.88 10.91 5.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3-----
Coef P Coef P Coef P
Constant 2.4733 6.110 5.042
Stress Level 0.028171 0.000 0.029040 0.000 0.01703 0.000
Skin Temp -0.1269 0.000 -0.1425 0.000
HR 0.02805 0.000
EDA
S 0.663168 0.649464 0.633422
R-sq 65.43% 66.91% 68.59%
R-sq(adj) 65.37% 66.78% 68.40%
R-sq(pred) 65.21% 66.55% 68.08%
Mallows’ Cp 58.84 36.87 11.71
------Step 4-----
Coef P
Constant 5.964
Stress Level 0.01604 0.000
Skin Temp -0.1689 0.000
HR 0.02738 0.000
EDA -0.816 0.003
S 0.628602
143
R-sq 69.13%
R-sq(adj) 68.88%
R-sq(pred) 68.27%
Mallows’ Cp 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -1.585 0.140 -8.17 -7.44
HR 0.06677 0.000 0.04840 0.000 0.04703 0.000 0.02739 0.000
EDA -4.812 0.000 -3.902 0.000 -3.755 0.000
Skin Temp 0.2880 0.000 0.2965 0.000
Stress Level 0.00988 0.006
S 1.08982 0.959024 0.913800 0.907890
R-sq 45.60% 57.95% 61.90% 62.47%
R-sq(adj) 45.49% 57.79% 61.67% 62.17%
R-sq(pred) 45.09% 54.38% 58.93% 59.32%
Mallows’ Cp 223.79 61.14 10.54 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 4.442 4.926 5.115
Temp -0.1950 0.000 -0.2085 0.000 -0.2417 0.000
Lighting -0.001957 0.003 -0.003683 0.000
CO2 0.001055 0.000
S 0.135265 0.134639 0.133363
R-sq 6.91% 7.88% 9.72%
R-sq(adj) 6.80% 7.66% 9.40%
R-sq(pred) 6.57% 7.35% 9.03%
Mallows’ Cp 25.66 18.53 3.15
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -63.36 -61.97 -63.79
Temp 4.114 0.000 4.231 0.000 4.323 0.000
CO2 -0.00593 0.000 -0.00749 0.000
Lighting 0.00878 0.039
S 0.751341 0.740740 0.739354
R-sq 51.32% 52.74% 52.97%
R-sq(adj) 51.27% 52.63% 52.81%
R-sq(pred) 51.01% 52.34% 52.48%
Mallows’ Cp 29.83 5.56 3.29
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- -----Step 3----
Coef P Coef P Coef P
Constant -227.2 -329.2 -255.0
Temp 13.75 0.000 16.59 0.000 15.54 0.000
Lighting 0.4133 0.000 0.3786 0.000
Acoustic -0.867 0.065
S 15.4648 15.1931 15.1693
R-sq 2.72% 6.23% 6.65%
R-sq(adj) 2.60% 5.99% 6.28%
R-sq(pred) 2.22% 5.48% 5.68%
Mallows’ Cp 31.69 4.91 3.49
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- ----Step 2----
Coef P Coef P
144
Constant -1040 -1226
Temp 48.83 0.000 54.46 0.000
Lighting 0.653 0.000
S 31.4838 31.1748
R-sq 7.38% 9.34%
R-sq(adj) 7.22% 9.04%
R-sq(pred) 6.75% 8.40%
Mallows’ Cp 12.78 2.07
Stepwise regression with the dataset: office zone C1 – day4
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 2.382 2.044 -1.342
HR 0.02764 0.000 0.03039 0.000 0.02665 0.000
EDA 0.5349 0.000 0.6051 0.000
Skin Temp 0.1191 0.000
Stress Level
S 0.755242 0.733958 0.720851
R-sq 15.54% 20.29% 23.16%
R-sq(adj) 15.48% 20.18% 23.00%
R-sq(pred) 15.36% 19.64% 22.32%
Mallows’ Cp 164.49 75.50 22.42
------Step 4-----
Coef P
Constant -1.016
HR 0.01823 0.000
EDA 0.6588 0.000
Skin Temp 0.1246 0.000
Stress Level 0.00538 0.000
S 0.716329
R-sq 24.18%
R-sq(adj) 23.97%
R-sq(pred) 23.03%
Mallows’ Cp 5.00
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.2073 3.9822 0.280
Stress Level 0.01264 0.000 0.01512 0.000 0.01392 0.000
EDA 0.6799 0.000 0.7537 0.000
Skin Temp 0.1211 0.000
HR
S 0.970377 0.944252 0.933307
R-sq 8.55% 13.47% 15.52%
R-sq(adj) 8.49% 13.35% 15.35%
R-sq(pred) 8.35% 12.90% 14.80%
Mallows’ Cp 122.60 41.10 8.23
------Step 4------
Coef P
Constant 0.392
Stress Level 0.01657 0.000
EDA 0.7683 0.000
Skin Temp 0.1343 0.000
HR -0.00771 0.022
S 0.931928
R-sq 15.83%
R-sq(adj) 15.60%
R-sq(pred) 14.94%
145
Mallows’ Cp 5.00
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 4.0098 2.663 0.823
EDA 0.7225 0.000 0.8164 0.000 0.6215 0.000
HR 0.01760 0.000 0.04922 0.000
Stress Level -0.02117 0.000
S 1.06814 1.05027 0.998441
R-sq 5.08% 8.29% 17.18%
R-sq(adj) 5.01% 8.16% 17.00%
R-sq(pred) 3.29% 6.33% 15.68%
Mallows’ Cp 207.76 154.31 3.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 2.194 1.809 -2.332
HR 0.03091 0.000 0.03397 0.000 0.02891 0.000
EDA 0.6275 0.000 0.7150 0.000
Skin Temp 0.1468 0.000
Stress Level
S 0.824260 0.796882 0.778752
R-sq 15.26% 20.85% 24.46%
R-sq(adj) 15.20% 20.74% 24.30%
R-sq(pred) 14.98% 20.00% 23.45%
Mallows’ Cp 175.67 71.73 5.26
------Step 4------
Coef P
Constant -2.452
HR 0.03198 0.000
EDA 0.6954 0.000
Skin Temp 0.1449 0.000
Stress Level -0.00201 0.133
S 0.778408
R-sq 24.58%
R-sq(adj) 24.37%
R-sq(pred) 23.44%
Mallows’ Cp 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.1142 4.0119 1.164
Stress Level 0.01329 0.000 0.01441 0.000 0.01349 0.000
EDA 0.3092 0.000 0.3660 0.000
Skin Temp 0.0932 0.000
HR
S 0.941033 0.935779 0.929391
R-sq 9.90% 10.97% 12.24%
R-sq(adj) 9.84% 10.84% 12.05%
R-sq(pred) 9.71% 10.68% 11.82%
Mallows’ Cp 44.05 28.62 9.80
------Step 4-----
Coef P
Constant 1.036
Stress Level 0.01049 0.000
EDA 0.3494 0.000
Skin Temp 0.0783 0.000
HR 0.00875 0.009
146
S 0.927514
R-sq 12.65%
R-sq(adj) 12.41%
R-sq(pred) 12.16%
Mallows’ Cp 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -1.286 -0.526 -6.43
CO2 0.002384 0.000 0.001867 0.000 0.001647 0.000
Lighting -0.00500 0.000 -0.00630 0.000
Temp 0.2682 0.000
S 0.338162 0.336642 0.335833
R-sq 2.39% 3.30% 3.80%
R-sq(adj) 2.35% 3.22% 3.69%
R-sq(pred) 2.26% 3.14% 3.56%
Mallows’ Cp 37.32 15.06 3.73
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2-----
Coef P Coef P
Constant 35.38 37.58
CO2 -0.00833 0.000 -0.00985 0.000
Lighting -0.01434 0.014
S 1.94829 1.94640
R-sq 0.88% 1.11%
R-sq(adj) 0.84% 1.04%
R-sq(pred) 0.75% 0.90%
Mallows’ Cp 5.32 1.25
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2----
Coef P Coef P
Constant -93.1 -100.8
Temp 7.40 0.012 6.92 0.020
Acoustic 0.344 0.087
S 14.1327 14.1273
R-sq 0.25% 0.36%
R-sq(adj) 0.21% 0.28%
R-sq(pred) 0.10% 0.14%
Mallows’ Cp 2.01 1.08
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2-----
Coef P Coef P
Constant -11.15 16.9
Lighting 0.4004 0.000 0.3562 0.000
CO2 -0.0383 0.135
S 23.3714 23.3642
R-sq 1.46% 1.57%
R-sq(adj) 1.41% 1.47%
R-sq(pred) 1.26% 1.27%
Mallows’ Cp 1.67 1.44
Stepwise regression with the dataset: office zone C1 – day4 (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.7594 2.9372 1.944
147
Stress Level 0.01665 0.000 0.02645 0.000 0.02082 0.000
EDA 2.818 0.000 2.977 0.000
HR 0.01444 0.005
Skin Temp
S 0.874797 0.709604 0.706186
R-sq 21.99% 48.74% 49.31%
R-sq(adj) 21.89% 48.60% 49.10%
R-sq(pred) 21.69% 48.38% 48.71%
Mallows’ Cp 397.06 14.97 8.89
------Step 4-----
Coef P
Constant 0.125
Stress Level 0.02043 0.000
EDA 3.120 0.000
HR 0.01470 0.004
Skin Temp 0.0589 0.016
S 0.703812
R-sq 49.72%
R-sq(adj) 49.44%
R-sq(pred) 48.95%
Mallows’ Cp 5.00
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.5913 2.6196 1.384
Stress Level 0.01801 0.000 0.02968 0.000 0.02280 0.000
EDA 3.325 0.000 3.524 0.000
HR 0.01793 0.012
S 1.15859 0.982838 0.979150
R-sq 14.43% 38.51% 39.06%
R-sq(adj) 14.31% 38.33% 38.80%
R-sq(pred) 14.08% 38.09% 38.39%
Mallows’ Cp 283.52 9.03 4.74
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.2334 14.141 14.339
EDA 3.191 0.000 2.248 0.000 2.474 0.000
Skin Temp -0.3547 0.000 -0.3660 0.000
Stress Level 0.00448 0.000
HR
S 0.756212 0.654279 0.646251
R-sq 43.23% 57.57% 58.66%
R-sq(adj) 43.15% 57.44% 58.48%
R-sq(pred) 42.95% 57.28% 58.23%
Mallows’ Cp 265.72 24.14 7.55
------Step 4------
Coef P
Constant 15.038
EDA 2.362 0.000
Skin Temp -0.3663 0.000
Stress Level 0.00833 0.000
HR -0.01004 0.033
S 0.644616
R-sq 58.93%
R-sq(adj) 58.69%
R-sq(pred) 58.31%
Mallows’ Cp 5.00
IAQ satisfaction vs bio
148
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.1068 3.5074 0.830
Stress Level 0.01168 0.000 0.01887 0.000 0.01842 0.000
EDA 2.051 0.000 2.258 0.000
Skin Temp 0.0876 0.000
S 0.734976 0.630094 0.623599
R-sq 14.98% 37.60% 38.97%
R-sq(adj) 14.86% 37.42% 38.71%
R-sq(pred) 14.64% 37.19% 38.30%
Mallows’ Cp 275.95 18.53 4.86
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 3.4512 9.968 9.29
Stress Level 0.02023 0.000 0.02318 0.000 0.01709 0.000
Skin Temp -0.2180 0.000 -0.2277 0.000
HR 0.01473 0.034
S 1.04626 1.01648 1.01396
R-sq 20.70% 25.26% 25.73%
R-sq(adj) 20.59% 25.04% 25.41%
R-sq(pred) 20.39% 24.68% 24.92%
Mallows’ Cp 46.98 6.11 3.62
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant -0.657 -0.205 0.094
CO2 0.001342 0.000 0.001032 0.000 0.001071 0.000
Lighting -0.002953 0.001 -0.002770 0.002
Acoustic -0.00619 0.146
S 0.210042 0.209237 0.209146
R-sq 1.96% 2.79% 2.95%
R-sq(adj) 1.89% 2.64% 2.72%
R-sq(pred) 1.68% 2.38% 2.40%
Mallows’ Cp 12.42 3.51 3.39
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 38.21 32.42 30.27
CO2 -0.01352 0.000 -0.01386 0.000 -0.01191 0.000
Acoustic 0.1094 0.006 0.0973 0.016
Lighting 0.01821 0.033
Temp
S 1.99127 1.98629 1.98357
R-sq 2.21% 2.77% 3.11%
R-sq(adj) 2.13% 2.62% 2.89%
R-sq(pred) 1.96% 2.43% 2.65%
Mallows’ Cp 18.69 13.14 10.58
------Step 4-----
Coef P
Constant 67.6
CO2 -0.01061 0.000
Acoustic 0.1036 0.010
Lighting 0.02537 0.005
Temp -1.702 0.006
S 1.97852
R-sq 3.68%
149
R-sq(adj) 3.38%
R-sq(pred) 3.12%
Mallows’ Cp 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- ----Step 2----
Coef P Coef P
Constant -248 -259
Temp 14.14 0.001 13.35 0.003
Acoustic 0.525 0.074
S 15.6455 15.6335
R-sq 0.71% 0.93%
R-sq(adj) 0.64% 0.79%
R-sq(pred) 0.45% 0.54%
Mallows’ Cp 3.75 2.54
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- ----Step 2---- -----Step 3----
Coef P Coef P Coef P
Constant -793 -710 -698
Temp 35.57 0.000 30.88 0.000 27.75 0.001
Lighting 0.279 0.010 0.384 0.001
CO2 0.0803 0.037
S 25.9527 25.8868 25.8484
R-sq 1.58% 2.16% 2.54%
R-sq(adj) 1.49% 1.99% 2.28%
R-sq(pred) 1.24% 1.65% 1.85%
Mallows’ Cp 10.89 6.14 3.81
Stepwise regression with the dataset: office zone C1 – day4 (male)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 2.7836 2.2855 0.291
HR 0.02659 0.000 0.03634 0.000 0.03283 0.000
Stress Level -0.009705 0.000 -0.009289 0.000
Skin Temp 0.07231 0.000
S 0.304137 0.252925 0.241871
R-sq 43.72% 61.13% 64.50%
R-sq(adj) 43.64% 61.02% 64.35%
R-sq(pred) 43.36% 60.71% 64.01%
Mallows’ Cp 424.08 69.98 3.01
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -0.962 -2.360 -8.150
HR 0.07147 0.000 0.09884 0.000 0.08866 0.000
Stress Level -0.02725 0.000 -0.02604 0.000
Skin Temp 0.2099 0.000
S 0.862069 0.719908 0.687119
R-sq 41.12% 59.00% 62.70%
R-sq(adj) 41.04% 58.88% 62.54%
R-sq(pred) 40.74% 58.53% 62.16%
Mallows’ Cp 421.55 74.88 4.68
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 1.014 0.352 -4.245
HR 0.04839 0.000 0.06134 0.000 0.05325 0.000
150
Stress Level -0.01288 0.000 -0.01192 0.000
Skin Temp 0.1667 0.000
EDA
S 0.822942 0.792255 0.773943
R-sq 26.00% 31.51% 34.73%
R-sq(adj) 25.90% 31.33% 34.46%
R-sq(pred) 25.26% 30.66% 33.82%
Mallows’ Cp 112.45 51.94 17.42
------Step 4------
Coef P
Constant -4.988
HR 0.05004 0.000
Stress Level -0.01020 0.000
Skin Temp 0.1947 0.000
EDA 0.2660 0.000
S 0.766896
R-sq 36.00%
R-sq(adj) 35.65%
R-sq(pred) 34.98%
Mallows’ Cp 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 4.9863 4.6733 4.6785
Stress Level -0.002058 0.000 -0.003191 0.000 -0.002907 0.000
HR 0.00451 0.000 0.00417 0.000
EDA 0.0482 0.022
S 0.242086 0.239038 0.238342
R-sq 2.85% 5.41% 6.09%
R-sq(adj) 2.72% 5.15% 5.71%
R-sq(pred) 2.17% 4.42% 4.99%
Mallows’ Cp 24.92 7.10 3.84
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -1.903 -0.869 -13.97
CO2 0.003409 0.000 0.002709 0.000 0.002201 0.000
Lighting -0.00686 0.000 -0.01007 0.000
Temp 0.597 0.000
S 0.427093 0.424880 0.421530
R-sq 3.04% 4.12% 5.70%
R-sq(adj) 2.97% 3.97% 5.48%
R-sq(pred) 2.81% 3.83% 5.23%
Mallows’ Cp 34.77 22.31 3.09
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 33.878 39.84 -7.9
Lighting -0.03850 0.000 -0.04716 0.000 -0.05796 0.000
CO2 -0.00822 0.000 -0.01001 0.000
Temp 2.172 0.000
Acoustic
S 1.76159 1.75397 1.74326
R-sq 2.28% 3.20% 4.45%
R-sq(adj) 2.21% 3.05% 4.23%
R-sq(pred) 1.87% 2.70% 3.81%
Mallows’ Cp 36.40 25.81 10.62
151
------Step 4-----
Coef P
Constant -5.0
Lighting -0.05547 0.000
CO2 -0.00948 0.000
Temp 2.250 0.000
Acoustic -0.0975 0.006
S 1.73885
R-sq 5.01%
R-sq(adj) 4.72%
R-sq(pred) 4.24%
Mallows’ Cp 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----
Coef P
Constant 70.27
Lighting 0.0822 0.111
S 11.8008
R-sq 0.23%
R-sq(adj) 0.14%
R-sq(pred) 0.00%
Mallows’ Cp 0.29
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 175.3 641 809
CO2 -0.2394 0.000 -0.2359 0.000 -0.1910 0.000
Temp -20.40 0.002 -30.46 0.000
Lighting 0.408 0.000
S 18.9734 18.8755 18.7307
R-sq 7.07% 8.14% 9.64%
R-sq(adj) 6.97% 7.92% 9.33%
R-sq(pred) 6.65% 7.51% 8.77%
Mallows’ Cp 23.73 15.56 3.12
Stepwise regression with the dataset: office zone C1 – day4 (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.7594 2.9372 1.944
Stress Level 0.01665 0.000 0.02645 0.000 0.02082 0.000
EDA 2.818 0.000 2.977 0.000
HR 0.01444 0.005
Skin Temp
S 0.874797 0.709604 0.706186
R-sq 21.99% 48.74% 49.31%
R-sq(adj) 21.89% 48.60% 49.10%
R-sq(pred) 21.69% 48.38% 48.71%
Mallows’ Cp 397.06 14.97 8.89
------Step 4-----
Coef P
Constant 0.125
Stress Level 0.02043 0.000
EDA 3.120 0.000
HR 0.01470 0.004
Skin Temp 0.0589 0.016
S 0.703812
R-sq 49.72%
152
R-sq(adj) 49.44%
R-sq(pred) 48.95%
Mallows’ Cp 5.00
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.5913 2.6196 1.384
Stress Level 0.01801 0.000 0.02968 0.000 0.02280 0.000
EDA 3.325 0.000 3.524 0.000
HR 0.01793 0.012
S 1.15859 0.982838 0.979150
R-sq 14.43% 38.51% 39.06%
R-sq(adj) 14.31% 38.33% 38.80%
R-sq(pred) 14.08% 38.09% 38.39%
Mallows’ Cp 283.52 9.03 4.74
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.2334 14.141 14.339
EDA 3.191 0.000 2.248 0.000 2.474 0.000
Skin Temp -0.3547 0.000 -0.3660 0.000
Stress Level 0.00448 0.000
HR
S 0.756212 0.654279 0.646251
R-sq 43.23% 57.57% 58.66%
R-sq(adj) 43.15% 57.44% 58.48%
R-sq(pred) 42.95% 57.28% 58.23%
Mallows’ Cp 265.72 24.14 7.55
------Step 4------
Coef P
Constant 15.038
EDA 2.362 0.000
Skin Temp -0.3663 0.000
Stress Level 0.00833 0.000
HR -0.01004 0.033
S 0.644616
R-sq 58.93%
R-sq(adj) 58.69%
R-sq(pred) 58.31%
Mallows’ Cp 5.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.1068 3.5074 0.830
Stress Level 0.01168 0.000 0.01887 0.000 0.01842 0.000
EDA 2.051 0.000 2.258 0.000
Skin Temp 0.0876 0.000
S 0.734976 0.630094 0.623599
R-sq 14.98% 37.60% 38.97%
R-sq(adj) 14.86% 37.42% 38.71%
R-sq(pred) 14.64% 37.19% 38.30%
Mallows’ Cp 275.95 18.53 4.86
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 3.4512 9.968 9.29
Stress Level 0.02023 0.000 0.02318 0.000 0.01709 0.000
153
Skin Temp -0.2180 0.000 -0.2277 0.000
HR 0.01473 0.034
S 1.04626 1.01648 1.01396
R-sq 20.70% 25.26% 25.73%
R-sq(adj) 20.59% 25.04% 25.41%
R-sq(pred) 20.39% 24.68% 24.92%
Mallows’ Cp 46.98 6.11 3.62
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant -0.657 -0.205 0.094
CO2 0.001342 0.000 0.001032 0.000 0.001071 0.000
Lighting -0.002953 0.001 -0.002770 0.002
Acoustic -0.00619 0.146
S 0.210042 0.209237 0.209146
R-sq 1.96% 2.79% 2.95%
R-sq(adj) 1.89% 2.64% 2.72%
R-sq(pred) 1.68% 2.38% 2.40%
Mallows’ Cp 12.42 3.51 3.39
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 38.21 32.42 30.27
CO2 -0.01352 0.000 -0.01386 0.000 -0.01191 0.000
Acoustic 0.1094 0.006 0.0973 0.016
Lighting 0.01821 0.033
Temp
S 1.99127 1.98629 1.98357
R-sq 2.21% 2.77% 3.11%
R-sq(adj) 2.13% 2.62% 2.89%
R-sq(pred) 1.96% 2.43% 2.65%
Mallows’ Cp 18.69 13.14 10.58
------Step 4-----
Coef P
Constant 67.6
CO2 -0.01061 0.000
Acoustic 0.1036 0.010
Lighting 0.02537 0.005
Temp -1.702 0.006
S 1.97852
R-sq 3.68%
R-sq(adj) 3.38%
R-sq(pred) 3.12%
Mallows’ Cp 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- ----Step 2----
Coef P Coef P
Constant -248 -259
Temp 14.14 0.001 13.35 0.003
Acoustic 0.525 0.074
S 15.6455 15.6335
R-sq 0.71% 0.93%
R-sq(adj) 0.64% 0.79%
R-sq(pred) 0.45% 0.54%
Mallows’ Cp 3.75 2.54
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- ----Step 2---- -----Step 3----
154
Coef P Coef P Coef P
Constant -793 -710 -698
Temp 35.57 0.000 30.88 0.000 27.75 0.001
Lighting 0.279 0.010 0.384 0.001
CO2 0.0803 0.037
S 25.9527 25.8868 25.8484
R-sq 1.58% 2.16% 2.54%
R-sq(adj) 1.49% 1.99% 2.28%
R-sq(pred) 1.24% 1.65% 1.85%
Mallows’ Cp 10.89 6.14 3.81
Stepwise regression with the dataset: office zone C1 – day4 (mid-age)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 2.6070 2.0566 -0.818
HR 0.024225 0.000 0.03602 0.000 0.03247 0.000
Stress Level -0.008658 0.000 -0.007677 0.000
Skin Temp 0.1017 0.000
S 0.234933 0.198304 0.184951
R-sq 77.83% 84.25% 86.34%
R-sq(adj) 77.76% 84.15% 86.22%
R-sq(pred) 77.64% 83.92% 85.87%
Mallows’ Cp 206.53 52.54 3.76
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1-----
Coef P
Constant 6.477
Skin Temp -0.0497 0.001
S 0.204211
R-sq 3.32%
R-sq(adj) 3.01%
R-sq(pred) 2.23%
Mallows’ Cp 0.64
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 0.838 -0.090 -4.18
HR 0.02542 0.000 0.04540 0.000 0.03990 0.000
Stress Level -0.01484 0.000 -0.01339 0.000
Skin Temp 0.1458 0.021
S 0.712120 0.677276 0.672509
R-sq 28.76% 35.77% 36.87%
R-sq(adj) 28.53% 35.35% 36.26%
R-sq(pred) 27.90% 34.39% 35.17%
Mallows’ Cp 39.63 7.43 4.03
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 0.779 -17.16 -20.48
HR 0.04390 0.000 0.02872 0.000 0.02910 0.000
Skin Temp 0.6250 0.000 0.7306 0.000
EDA 0.515 0.000
S 1.09556 1.02479 0.993527
R-sq 33.72% 42.19% 45.84%
R-sq(adj) 33.51% 41.82% 45.32%
R-sq(pred) 32.76% 41.02% 44.54%
155
Mallows’ Cp 68.07 21.98 3.28
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 2.508 2.080
Skin Temp 0.0794 0.000 0.0932 0.000
EDA 0.0640 0.015
S 0.232162 0.230324
R-sq 6.36% 8.13%
R-sq(adj) 6.05% 7.54%
R-sq(pred) 5.01% 6.62%
Mallows’ Cp 10.01 5.99
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -16.66 -20.74 -20.32
Temp 0.732 0.000 0.946 0.000 0.876 0.000
Lighting -0.00979 0.000 -0.00782 0.001
CO2 0.001601 0.029
S 0.443552 0.438868 0.437894
R-sq 2.62% 4.78% 5.31%
R-sq(adj) 2.51% 4.56% 4.98%
R-sq(pred) 2.13% 4.25% 4.65%
Mallows’ Cp 24.04 6.65 3.86
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 128.1 109.2 105.5
Temp -4.284 0.000 -3.286 0.000 -3.396 0.000
Lighting -0.0463 0.000 -0.0485 0.000
Acoustic 0.1163 0.030
S 2.18643 2.16569 2.16109
R-sq 3.67% 5.59% 6.10%
R-sq(adj) 3.56% 5.38% 5.78%
R-sq(pred) 3.36% 5.12% 5.50%
Mallows’ Cp 21.89 6.01 3.29
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 17.3 -414
CO2 0.1031 0.001 0.0978 0.001
Temp 18.92 0.003
S 17.8201 17.7244
R-sq 1.62% 2.82%
R-sq(adj) 1.48% 2.54%
R-sq(pred) 1.08% 2.05%
Mallows’ Cp 9.22 2.60
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2----
Coef P Coef P
Constant -1478 -1540
Temp 66.4 0.000 65.9 0.000
CO2 0.1141 0.086
S 27.7224 27.6532
R-sq 6.26% 6.97%
R-sq(adj) 6.02% 6.49%
156
R-sq(pred) 5.43% 5.60%
Mallows’ Cp 4.18 3.22
Stepwise regression with the dataset: office zone C2 – day1
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant -3.534 -4.065 -4.251
Skin Temp 0.2315 0.000 0.2577 0.000 0.2643 0.000
Stress Level -0.011952 0.000 -0.011955 0.000
EDA -0.1066 0.094
S 0.699931 0.629512 0.629090
R-sq 13.75% 30.28% 30.43%
R-sq(adj) 13.69% 30.18% 30.27%
R-sq(pred) 13.56% 30.04% 30.16%
Mallows’ Cp 322.79 4.32 3.51
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -8.567 -10.260 -10.367
Skin Temp 0.4012 0.000 0.4704 0.000 0.4114 0.000
Stress Level -0.01971 0.000 -0.02770 0.000
HR 0.02927 0.000
S 1.26073 1.15504 1.13997
R-sq 11.97% 26.17% 28.14%
R-sq(adj) 11.90% 26.05% 27.97%
R-sq(pred) 11.70% 25.82% 27.66%
Mallows’ Cp 278.10 36.18 4.29
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.2490 3.395 6.411
Stress Level -0.01662 0.000 -0.02031 0.000 -0.02132 0.000
HR 0.01285 0.007 0.01903 0.000
Skin Temp -0.1098 0.000
S 1.15488 1.15199 1.14635
R-sq 12.29% 12.80% 13.72%
R-sq(adj) 12.22% 12.66% 13.51%
R-sq(pred) 12.03% 12.37% 13.14%
Mallows’ Cp 20.22 14.94 3.77
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 5.0611 4.003 4.032
Stress Level -0.017941 0.000 -0.022508 0.000 -0.022321 0.000
HR 0.01591 0.000 0.01513 0.000
EDA 0.1244 0.014
S 0.524449 0.513217 0.512180
R-sq 44.18% 46.59% 46.85%
R-sq(adj) 44.14% 46.51% 46.72%
R-sq(pred) 43.94% 46.25% 46.48%
Mallows’ Cp 61.42 7.52 3.52
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -9.977 -10.346 -9.90
Skin Temp 0.4395 0.000 0.4546 0.000 0.4391 0.000
157
Stress Level -0.00430 0.002 -0.00426 0.002
EDA 0.230 0.072
S 1.26307 1.25874 1.25760
R-sq 13.98% 14.64% 14.86%
R-sq(adj) 13.91% 14.50% 14.65%
R-sq(pred) 13.70% 14.27% 14.46%
Mallows’ Cp 12.41 4.88 3.64
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- -------Step 3------
Coef P Coef P Coef P
Constant -0.496 -1.409 -1.763
Acoustic 0.01259 0.000 0.01565 0.000 0.01511 0.000
Temp 0.03127 0.000 0.0483 0.000
Lighting -0.000101 0.012
S 0.260191 0.259109 0.258769
R-sq 1.43% 2.30% 2.60%
R-sq(adj) 1.38% 2.20% 2.46%
R-sq(pred) 1.02% 1.85% 2.11%
Mallows’ Cp 25.10 9.14 4.83
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ ------Step 2------
Coef P Coef P
Constant 31.4875 35.504
Lighting -0.000884 0.000 -0.001014 0.000
CO2 -0.004017 0.000
S 1.14429 1.13934
R-sq 2.35% 3.24%
R-sq(adj) 2.30% 3.14%
R-sq(pred) 2.17% 2.96%
Mallows’ Cp 19.97 3.37
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 71.443 118.0 75.1
Lighting 0.01358 0.000 0.01198 0.000 0.00447 0.033
CO2 -0.0466 0.000 -0.0606 0.000
Temp 2.450 0.000
S 12.1385 12.0763 11.9998
R-sq 4.84% 5.87% 7.11%
R-sq(adj) 4.79% 5.76% 6.96%
R-sq(pred) 4.61% 5.53% 6.67%
Mallows’ Cp 45.25 27.24 4.96
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3----- -----Step 4-----
Coef P Coef P Coef P Coef P
Constant 18.39 117.4 39.4 -4.2
Lighting 0.02428 0.000 0.02103 0.000 0.00806 0.099 0.00962 0.052
CO2 -0.0991 0.000 -0.1244 0.000 -0.1162 0.000
Temp 4.45 0.000 4.67 0.000
Acoustic 0.551 0.049
S 25.2806 25.1477 25.0303 25.0044
R-sq 3.55% 4.63% 5.59% 5.85%
R-sq(adj) 3.48% 4.49% 5.38% 5.58%
R-sq(pred) 3.24% 4.20% 5.02% 5.13%
Mallows’ Cp 33.00 19.03 6.89 5.00
Stepwise regression with the dataset: office zone C2 – day1 (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
158
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 5.740 10.650 10.659
HR -0.01920 0.000 -0.01702 0.000 -0.02053 0.000
Skin Temp -0.1591 0.000 -0.1527 0.000
Stress Level 0.00296 0.009
EDA
S 0.439465 0.412924 0.411218
R-sq 14.46% 24.58% 25.31%
R-sq(adj) 14.33% 24.37% 24.99%
R-sq(pred) 14.02% 23.97% 24.48%
Mallows’ Cp 103.67 10.17 5.30
------Step 4------
Coef P
Constant 10.437
HR -0.02036 0.000
Skin Temp -0.1457 0.000
Stress Level 0.00303 0.007
EDA -0.0658 0.130
S 0.410838
R-sq 25.56%
R-sq(adj) 25.13%
R-sq(pred) 24.66%
Mallows’ Cp 5.00
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------
Coef P Coef P
Constant 3.449 3.395
Skin Temp 0.0453 0.006 0.0464 0.005
Stress Level 0.001393 0.121
S 0.412295 0.411868
R-sq 1.09% 1.44%
R-sq(adj) 0.95% 1.15%
R-sq(pred) 0.67% 0.77%
Mallows’ Cp 1.73 1.32
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1-----
Coef P
Constant 15.94
Skin Temp -0.3669 0.000
S 1.14486
R-sq 8.58%
R-sq(adj) 8.44%
R-sq(pred) 8.12%
Mallows’ Cp 0.32
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2---- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant -2.90 -1.69 -1.17 -1.27
Skin Temp 0.2279 0.000 0.1872 0.000 0.2012 0.000 0.2263 0.000
EDA 0.365 0.002 0.404 0.001 0.389 0.001
HR -0.01342 0.004 -0.02504 0.000
Stress Level 0.00973 0.002
S 1.12538 1.11833 1.11248 1.10521
R-sq 3.61% 4.95% 6.08% 7.44%
R-sq(adj) 3.47% 4.67% 5.67% 6.90%
R-sq(pred) 2.94% 4.17% 5.06% 6.25%
159
Mallows’ Cp 27.12 19.27 12.98 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- -------Step 3------
Coef P Coef P Coef P
Constant -0.859 -2.260 -2.762
Acoustic 0.01995 0.000 0.02468 0.000 0.02392 0.000
Temp 0.0480 0.000 0.0721 0.000
Lighting -0.000144 0.027
S 0.324209 0.322169 0.321649
R-sq 2.30% 3.61% 4.00%
R-sq(adj) 2.22% 3.45% 3.76%
R-sq(pred) 1.63% 2.87% 3.18%
Mallows’ Cp 21.79 7.32 4.40
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 31.9873 31.124 29.32
Lighting -0.000989 0.000 -0.000936 0.000 -0.000872 0.000
Acoustic 0.0158 0.150 0.0193 0.083
CO2 0.001611 0.102
S 0.915070 0.914664 0.914030
R-sq 4.50% 4.66% 4.87%
R-sq(adj) 4.42% 4.50% 4.64%
R-sq(pred) 4.12% 4.11% 4.18%
Mallows’ Cp 3.80 3.72 3.04
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 70.596 27.6 54.9
Lighting 0.01519 0.000 0.00997 0.000 0.00737 0.007
Temp 1.850 0.004 2.301 0.000
CO2 -0.0378 0.009
S 12.3264 12.2838 12.2503
R-sq 6.03% 6.76% 7.35%
R-sq(adj) 5.94% 6.59% 7.10%
R-sq(pred) 5.60% 6.17% 6.56%
Mallows’ Cp 14.70 8.12 3.18
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1-----
Coef P
Constant 9.046
Lighting 0.02650 0.000
S 16.8567
R-sq 9.54%
R-sq(adj) 9.41%
R-sq(pred) 8.83%
Mallows’ Cp 1.02
Stepwise regression with the dataset: office zone C2 – day1 (male)
For each analysis: α to enter = 0.15, α to remove = 0.15
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.711 2.978 1.974
EDA -5.506 0.000 -5.862 0.000 -5.613 0.000
HR 0.01053 0.030 0.02696 0.007
Stress Level -0.00710 0.060
S 1.22521 1.22110 1.21829
160
R-sq 12.11% 12.86% 13.41%
R-sq(adj) 11.95% 12.54% 12.94%
R-sq(pred) 11.48% 12.00% 12.14%
Mallows’ Cp 8.16 5.42 3.87
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2-----
Coef P Coef P
Constant 4.2264 15.81
EDA -6.209 0.000 -5.203 0.000
Skin Temp -0.3815 0.000
S 0.962609 0.875347
R-sq 22.11% 35.71%
R-sq(adj) 21.97% 35.47%
R-sq(pred) 21.55% 35.09%
Mallows’ Cp 120.58 5.51
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 4.8726 13.47 12.72
Stress Level -0.019731 0.000 -0.01177 0.000 -0.01808 0.000
Skin Temp -0.2890 0.000 -0.3077 0.000
HR 0.02081 0.000
S 0.611673 0.578326 0.568731
R-sq 48.12% 53.70% 55.31%
R-sq(adj) 48.02% 53.53% 55.06%
R-sq(pred) 47.76% 53.25% 54.64%
Mallows’ Cp 87.96 21.26 3.53
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- -----Step 4-----
Coef P Coef P Coef P Coef P
Constant 0.280 0.284 1.575 6.02
HR 0.03823 0.000 0.04688 0.000 0.02575 0.007 0.02814 0.003
EDA -4.605 0.000 -4.925 0.000 -4.999 0.000
Stress Level 0.00914 0.011 0.01267 0.001
Skin Temp -0.1541 0.032
S 1.22367 1.16808 1.16235 1.15853
R-sq 10.71% 18.78% 19.72% 20.40%
R-sq(adj) 10.54% 18.49% 19.29% 19.82%
R-sq(pred) 10.10% 18.10% 18.76% 19.09%
Mallows’ Cp 65.82 12.12 7.63 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant -0.0185 -0.120 -0.315
CO2 0.000150 0.108 0.000167 0.076 0.000170 0.069
Acoustic 0.00157 0.137 0.00222 0.045
Temp 0.00654 0.051
Lighting
S 0.0741628 0.0741069 0.0739767
R-sq 0.32% 0.59% 1.07%
R-sq(adj) 0.20% 0.35% 0.70%
R-sq(pred) 0.00% 0.00% 0.09%
Mallows’ Cp 7.51 7.28 5.44
-------Step 4------- -------Step 5-------
Coef P Coef P
Constant -0.361 -0.264
CO2 0.000115 0.253
161
Acoustic 0.00197 0.077 0.00177 0.108
Temp 0.01161 0.013 0.01286 0.005
Lighting -0.000031 0.119 -0.000038 0.035
S 0.0739105 0.0739246
R-sq 1.37% 1.21%
R-sq(adj) 0.87% 0.84%
R-sq(pred) 0.24% 0.36%
Mallows’ Cp 5.00 4.31
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2------
Coef P Coef P
Constant 40.60 42.70
CO2 -0.01010 0.000 -0.01196 0.000
Lighting -0.001115 0.000
S 1.05319 1.03103
R-sq 6.72% 10.71%
R-sq(adj) 6.60% 10.49%
R-sq(pred) 6.31% 10.13%
Mallows’ Cp 36.30 2.23
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 167.0 104.2 60.3
CO2 -0.0924 0.000 -0.0941 0.000 -0.0885 0.000
Temp 2.698 0.000 3.162 0.000
Acoustic 0.504 0.005
S 11.7313 11.5289 11.4747
R-sq 4.68% 8.06% 9.05%
R-sq(adj) 4.55% 7.81% 8.68%
R-sq(pred) 4.19% 7.40% 8.18%
Mallows’ Cp 34.13 9.05 3.16
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 284.4 151.8 17.4
CO2 -0.2527 0.000 -0.2638 0.000 -0.2493 0.000
Temp 6.02 0.000 7.39 0.000
Acoustic 1.616 0.000
S 28.5767 28.1891 27.9440
R-sq 5.73% 8.40% 10.13%
R-sq(adj) 5.58% 8.13% 9.72%
R-sq(pred) 5.20% 7.66% 9.04%
Mallows’ Cp 31.40 13.70 3.04
Stepwise regression with the dataset: office zone C2 – day1 (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 4.2727 0.033 -0.269
Stress Level -0.014640 0.000 -0.014707 0.000 -0.014719 0.000
Skin Temp 0.1344 0.000 0.1452 0.000
EDA -0.1761 0.004
S 0.628345 0.609709 0.607678
R-sq 28.55% 32.79% 33.30%
R-sq(adj) 28.48% 32.66% 33.11%
R-sq(pred) 28.32% 32.45% 32.92%
Mallows’ Cp 74.79 9.41 3.31
Acoustic satisfaction vs bio
162
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 4.6591 -11.718 -11.577
Stress Level -0.02231 0.000 -0.02393 0.000 -0.03289 0.000
Skin Temp 0.5184 0.000 0.4443 0.000
HR 0.03276 0.000
S 1.11337 0.982552 0.959643
R-sq 23.00% 40.09% 42.91%
R-sq(adj) 22.92% 39.97% 42.74%
R-sq(pred) 22.72% 39.67% 42.38%
Mallows’ Cp 345.29 50.44 3.49
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.2632 3.176 5.86
Stress Level -0.01943 0.000 -0.02398 0.000 -0.02498 0.000
HR 0.01619 0.002 0.02081 0.000
Skin Temp -0.0947 0.011
S 1.13091 1.12595 1.12288
R-sq 18.01% 18.81% 19.33%
R-sq(adj) 17.93% 18.64% 19.09%
R-sq(pred) 17.72% 18.31% 18.65%
Mallows’ Cp 15.50 7.68 3.25
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 5.0822 3.750 3.779
Stress Level -0.019735 0.000 -0.025304 0.000 -0.025068 0.000
HR 0.01983 0.000 0.01897 0.000
EDA 0.1382 0.007
S 0.525958 0.508129 0.506509
R-sq 51.16% 54.46% 54.80%
R-sq(adj) 51.11% 54.37% 54.66%
R-sq(pred) 50.91% 54.11% 54.40%
Mallows’ Cp 78.94 8.55 3.20
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -13.68 -13.96 -13.45
Skin Temp 0.5544 0.000 0.5679 0.000 0.5501 0.000
Stress Level -0.00598 0.000 -0.00594 0.000
EDA 0.253 0.037
S 1.20110 1.19064 1.18864
R-sq 18.01% 19.52% 19.87%
R-sq(adj) 17.93% 19.35% 19.63%
R-sq(pred) 17.68% 19.06% 19.36%
Mallows’ Cp 21.96 5.37 3.03
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- -------Step 3------
Coef P Coef P Coef P
Constant -0.607 -1.702 -2.126
Acoustic 0.01507 0.000 0.01876 0.000 0.01812 0.000
Temp 0.03749 0.000 0.0579 0.000
Lighting -0.000122 0.014
CO2
S 0.285110 0.283692 0.283249
163
R-sq 1.71% 2.74% 3.11%
R-sq(adj) 1.65% 2.62% 2.93%
R-sq(pred) 1.20% 2.18% 2.49%
Mallows’ Cp 24.83 9.54 5.47
-------Step 4------
Coef P
Constant -2.486
Acoustic 0.01887 0.000
Temp 0.0532 0.000
Lighting -0.000092 0.082
CO2 0.000426 0.116
S 0.283120
R-sq 3.26%
R-sq(adj) 3.02%
R-sq(pred) 2.57%
Mallows’ Cp 5.00
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 31.8302 35.737 33.61
Lighting -0.001220 0.000 -0.001346 0.000 -0.001230 0.000
CO2 -0.003907 0.000 -0.003400 0.001
Acoustic 0.0296 0.009
Temp
S 1.07240 1.06745 1.06550
R-sq 4.96% 5.89% 6.29%
R-sq(adj) 4.90% 5.78% 6.12%
R-sq(pred) 4.71% 5.53% 5.83%
Mallows’ Cp 24.70 10.59 5.64
------Step 4------
Coef P
Constant 32.09
Lighting -0.001456 0.000
CO2 -0.00379 0.000
Acoustic 0.0318 0.005
Temp 0.0770 0.104
S 1.06496
R-sq 6.45%
R-sq(adj) 6.22%
R-sq(pred) 5.86%
Mallows’ Cp 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2-----
Coef P Coef P
Constant -17.89 54.5
Temp 3.905 0.000 3.917 0.000
CO2 -0.0732 0.000
S 12.3879 12.2225
R-sq 6.34% 8.88%
R-sq(adj) 6.27% 8.76%
R-sq(pred) 6.07% 8.49%
Mallows’ Cp 41.30 2.63
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant -158.2 0.7 -62.3
Temp 7.73 0.000 7.742 0.000 8.41 0.000
CO2 -0.1606 0.000 -0.1534 0.000
164
Acoustic 0.738 0.025
S 26.6551 26.2919 26.2427
R-sq 5.12% 7.77% 8.20%
R-sq(adj) 5.03% 7.60% 7.95%
R-sq(pred) 4.77% 7.28% 7.53%
Mallows’ Cp 35.37 6.07 3.00
Stepwise regression with the dataset: office zone C2 – day1 (mid-age)
For each analysis: α to enter = 0.15, α to remove = 0.15
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 3.256 4.004 -8.05
Stress Level 0.02606 0.000 0.02963 0.000 0.02409 0.000
EDA -8.99 0.000 -9.68 0.000
Skin Temp 0.406 0.001
S 1.51095 1.44994 1.41952
R-sq 9.59% 17.09% 20.86%
R-sq(adj) 9.21% 16.39% 19.86%
R-sq(pred) 8.35% 14.85% 18.21%
Mallows’ Cp 33.71 13.17 3.82
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 3.9558 4.349
Stress Level 0.01272 0.001 0.01460 0.000
EDA -4.72 0.001
S 1.08164 1.05945
R-sq 4.70% 8.95%
R-sq(adj) 4.30% 8.19%
R-sq(pred) 3.52% 6.95%
Mallows’ Cp 11.47 2.34
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1-----
Coef P
Constant 4.6112
EDA 2.516 0.000
S 0.347997
R-sq 11.16%
R-sq(adj) 10.79%
R-sq(pred) 9.00%
Mallows’ Cp 3.97
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 3.173 3.527 -3.47
Stress Level 0.01921 0.000 0.02090 0.000 0.01769 0.001
EDA -4.25 0.025 -4.65 0.014
Skin Temp 0.236 0.049
S 1.42423 1.41223 1.40370
R-sq 6.09% 8.05% 9.54%
R-sq(adj) 5.70% 7.28% 8.40%
R-sq(pred) 4.84% 6.24% 7.15%
Mallows’ Cp 8.84 5.68 3.77
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------ -------Step 3-------
Coef P Coef P Coef P
Constant -0.0282 -0.2405 -0.0797
165
Acoustic 0.002164 0.001 0.002870 0.000 0.002659 0.000
Temp 0.00729 0.000 0.00718 0.000
CO2 -0.000148 0.009
Lighting
S 0.0321639 0.0316966 0.0314652
R-sq 2.72% 5.76% 7.36%
R-sq(adj) 2.48% 5.29% 6.67%
R-sq(pred) 2.00% 4.44% 5.63%
Mallows’ Cp 29.44 18.07 13.02
-------Step 4-------
Coef P
Constant -0.1349
Acoustic 0.002362 0.000
Temp 0.01331 0.000
CO2 -0.000215 0.000
Lighting -0.000037 0.002
S 0.0311144
R-sq 9.65%
R-sq(adj) 8.74%
R-sq(pred) 7.65%
Mallows’ Cp 5.00
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ ------Step 2------
Coef P Coef P
Constant 35.17 38.24
CO2 -0.00498 0.000 -0.00549 0.000
Acoustic -0.0476 0.000
S 0.653262 0.643294
R-sq 4.37% 7.50%
R-sq(adj) 4.13% 7.04%
R-sq(pred) 3.54% 6.13%
Mallows’ Cp 20.89 9.10
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 68.357 143.1
Lighting 0.01161 0.000 0.02130 0.000
Temp -3.220 0.000
S 9.89378 9.72361
R-sq 4.69% 8.19%
R-sq(adj) 4.42% 7.68%
R-sq(pred) 3.57% 6.67%
Mallows’ Cp 13.19 1.53
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 11.79 173.4
Lighting 0.01861 0.000 0.03753 0.000
Temp -6.94 0.000
S 16.4230 15.9161
R-sq 4.70% 10.78%
R-sq(adj) 4.39% 10.20%
R-sq(pred) 3.20% 8.68%
Mallows’ Cp 21.99 3.11
Stepwise regression with the dataset: office zone C2 – day2
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
166
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 16.353 14.908 15.440
Skin Temp -0.3955 0.000 -0.3370 0.000 -0.3557 0.000
Stress Level -0.01382 0.000 -0.01314 0.000
EDA 0.0694 0.083
S 1.22643 1.18956 1.18847
R-sq 21.47% 26.18% 26.39%
R-sq(adj) 21.39% 26.05% 26.19%
R-sq(pred) 21.18% 25.77% 25.85%
Mallows’ Cp 72.93 5.23 4.22
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -10.929 -9.659 -7.107
Skin Temp 0.4901 0.000 0.5106 0.000 0.4547 0.000
HR -0.02677 0.000 -0.04398 0.000
Stress Level 0.01570 0.000
EDA
S 0.802285 0.767765 0.704771
R-sq 48.10% 52.52% 60.03%
R-sq(adj) 48.05% 52.42% 59.91%
R-sq(pred) 47.89% 52.23% 59.70%
Mallows’ Cp 308.92 197.24 5.90
------Step 4------
Coef P
Constant -7.492
Skin Temp 0.4666 0.000
HR -0.04331 0.000
Stress Level 0.01516 0.000
EDA -0.0410 0.089
S 0.704120
R-sq 60.14%
R-sq(adj) 59.98%
R-sq(pred) 59.77%
Mallows’ Cp 5.00
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -11.493 -10.258 -7.623
Skin Temp 0.5075 0.000 0.5274 0.000 0.4697 0.000
HR -0.02603 0.000 -0.04380 0.000
Stress Level 0.01621 0.000
S 0.909450 0.880951 0.823007
R-sq 43.61% 47.14% 53.91%
R-sq(adj) 43.55% 47.03% 53.77%
R-sq(pred) 43.37% 46.82% 53.54%
Mallows’ Cp 230.81 154.07 5.06
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -3.032 -1.865 -0.549
Skin Temp 0.2468 0.000 0.2657 0.000 0.2368 0.000
HR -0.02463 0.000 -0.03349 0.000
Stress Level 0.008091 0.000
S 0.602263 0.562750 0.540561
R-sq 29.43% 38.45% 43.26%
R-sq(adj) 29.36% 38.33% 43.09%
167
R-sq(pred) 29.15% 38.08% 42.81%
Mallows’ Cp 251.52 90.18 4.97
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 0.858 1.554
Skin Temp 0.0903 0.001 0.0665 0.029
EDA 0.0969 0.039
S 1.42699 1.42473
R-sq 0.99% 1.40%
R-sq(adj) 0.89% 1.20%
R-sq(pred) 0.56% 0.87%
Mallows’ Cp 6.54 4.26
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 5.68 3.35 4.26
Temp -0.2423 0.001 -0.1834 0.018 -0.2288 0.005
Acoustic 0.01917 0.025 0.01941 0.023
Lighting 0.000612 0.064
S 0.951645 0.950743 0.950195
R-sq 0.52% 0.76% 0.92%
R-sq(adj) 0.47% 0.66% 0.78%
R-sq(pred) 0.38% 0.51% 0.54%
Mallows’ Cp 7.48 4.46 3.01
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -20.13 -28.85 -37.12
Temp 2.294 0.000 2.718 0.000 2.849 0.000
Lighting -0.005233 0.000 -0.004756 0.000
CO2 0.00514 0.050
S 2.44343 2.41988 2.41835
R-sq 6.46% 8.30% 8.45%
R-sq(adj) 6.42% 8.22% 8.33%
R-sq(pred) 6.27% 8.03% 8.10%
Mallows’ Cp 49.88 6.83 4.98
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 68.534 114.4
Lighting 0.02342 0.000 0.01804 0.000
CO2 -0.0435 0.001
S 11.7972 11.7670
R-sq 1.87% 2.42%
R-sq(adj) 1.81% 2.32%
R-sq(pred) 1.65% 2.10%
Mallows’ Cp 12.15 3.67
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 23.56 154.2 103.5
Lighting 0.03558 0.000 0.04349 0.000 0.04350 0.000
Temp -6.00 0.007 -4.73 0.045
Acoustic 0.423 0.103
S 22.6676 22.6139 22.5995
168
R-sq 1.11% 1.65% 1.85%
R-sq(adj) 1.03% 1.50% 1.63%
R-sq(pred) 0.81% 1.20% 1.25%
Mallows’ Cp 10.86 5.62 4.96
Stepwise regression with the dataset: office zone C2 – day2 (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 22.85 22.93 23.29 24.20
Skin Temp -0.5951 0.000 -0.5710 0.000 -0.6300 0.000 -0.6586 0.000
Stress Level -0.02539 0.000 -0.03055 0.000 -0.02953 0.000
HR 0.02380 0.000 0.02282 0.000
EDA 0.0771 0.041
S 1.25733 1.10534 1.09214 1.08974
R-sq 24.68% 41.87% 43.33% 43.65%
R-sq(adj) 24.57% 41.71% 43.09% 43.34%
R-sq(pred) 24.34% 41.46% 42.79% 43.02%
Mallows’ Cp 240.81 23.76 7.17 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -----Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 13.01 15.49 15.82 15.49
Skin Temp -0.2821 0.000 -0.3636 0.000 -0.4145 0.000 -0.4394 0.000
EDA 0.2049 0.000 0.2126 0.000 0.1474 0.002
HR 0.01823 0.002 0.04223 0.000
Stress Level -0.01641 0.000
S 1.39612 1.37837 1.36960 1.33475
R-sq 5.69% 8.20% 9.49% 14.16%
R-sq(adj) 5.55% 7.94% 9.11% 13.67%
R-sq(pred) 5.15% 7.58% 8.61% 13.11%
Mallows’ Cp 68.79 50.09 41.45 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 10.98 13.04 8.81
Temp -0.478 0.000 -0.579 0.000 -0.473 0.000
Lighting 0.001406 0.015 0.001428 0.014
Acoustic 0.0356 0.014
S 1.22378 1.22134 1.21884
R-sq 1.25% 1.72% 2.20%
R-sq(adj) 1.17% 1.56% 1.96%
R-sq(pred) 1.01% 1.21% 1.51%
Mallows’ Cp 10.93 7.02 3.01
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -36.81 -46.74 -59.49
Temp 3.085 0.000 3.569 0.000 3.770 0.000
Lighting -0.00596 0.000 -0.00523 0.000
CO2 0.00793 0.046
S 2.86027 2.83459 2.83143
R-sq 8.36% 10.07% 10.33%
R-sq(adj) 8.29% 9.93% 10.13%
R-sq(pred) 8.05% 9.64% 9.74%
Mallows’ Cp 30.03 6.43 4.42
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
169
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 69.550 10.0
Lighting 0.03009 0.000 0.02724 0.000
Temp 2.73 0.043
S 11.6276 11.6109
R-sq 3.02% 3.39%
R-sq(adj) 2.93% 3.21%
R-sq(pred) 2.64% 2.83%
Mallows’ Cp 7.08 4.97
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2---- -----Step 3----
Coef P Coef P Coef P
Constant 29.91 -18.6 99.3
Lighting 0.0370 0.002 0.0409 0.001 0.0463 0.000
Acoustic 0.887 0.004 0.700 0.034
Temp -4.94 0.100
S 23.2669 23.1720 23.1497
R-sq 1.05% 1.97% 2.27%
R-sq(adj) 0.94% 1.75% 1.94%
R-sq(pred) 0.65% 1.33% 1.43%
Mallows’ Cp 10.73 4.45 3.75
Stepwise regression with the dataset: office zone C2 – day2 (male)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 19.47 14.95 12.04
Skin Temp -0.5110 0.000 -0.4878 0.000 -0.3859 0.000
HR 0.05240 0.000 0.05430 0.000
EDA -1.100 0.033
S 1.05671 0.978701 0.973938
R-sq 18.08% 29.92% 30.79%
R-sq(adj) 17.86% 29.54% 30.22%
R-sq(pred) 16.97% 28.44% 29.09%
Mallows’ Cp 65.84 5.72 3.15
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 2.1556 4.450 -1.16
EDA 3.141 0.000 3.259 0.000 2.348 0.000
HR -0.03207 0.000 -0.03055 0.000
Skin Temp 0.1930 0.000
S 0.717887 0.683974 0.670721
R-sq 28.17% 35.01% 37.70%
R-sq(adj) 27.95% 34.59% 37.11%
R-sq(pred) 27.45% 33.69% 36.17%
Mallows’ Cp 47.29 14.83 3.25
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 1.9785 3.770 -0.40
EDA 3.624 0.000 3.716 0.000 3.039 0.000
HR -0.02505 0.002 -0.02391 0.002
Skin Temp 0.1434 0.056
S 0.975444 0.961518 0.957455
R-sq 22.05% 24.49% 25.37%
170
R-sq(adj) 21.80% 24.02% 24.66%
R-sq(pred) 21.17% 22.90% 23.34%
Mallows’ Cp 13.01 4.71 3.03
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 8.960 4.45 3.97
HR -0.07044 0.000 -0.07037 0.000 -0.06094 0.000
Skin Temp 0.1526 0.000 0.1500 0.000
Stress Level -0.00983 0.017
EDA
S 0.731907 0.717384 0.712004
R-sq 30.75% 33.68% 34.87%
R-sq(adj) 30.53% 33.26% 34.25%
R-sq(pred) 29.78% 32.32% 32.95%
Mallows’ Cp 24.26 11.90 8.03
------Step 4------
Coef P
Constant 1.49
HR -0.05859 0.000
Skin Temp 0.2348 0.000
Stress Level -0.01083 0.008
EDA -0.861 0.026
S 0.707489
R-sq 35.90%
R-sq(adj) 35.09%
R-sq(pred) 33.27%
Mallows’ Cp 5.00
Thermal satisfaction vs bio
Stepwise Selection of Terms
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 9.94 8.88
Skin Temp -0.2399 0.000 -0.2395 0.000
HR 0.01450 0.080
S 1.02057 1.01722
R-sq 5.10% 6.02%
R-sq(adj) 4.80% 5.43%
R-sq(pred) 4.15% 4.58%
Mallows’ Cp 2.24 1.17
EDA vs IEQ
Stepwise Selection of Terms
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -------Step 2------
Coef P Coef P
Constant -1.876 -2.133
Temp 0.0934 0.000 0.1059 0.000
Lighting -0.000154 0.004
S 0.104572 0.104142
R-sq 5.89% 6.77%
R-sq(adj) 5.78% 6.56%
R-sq(pred) 4.90% 5.66%
Mallows’ Cp 9.44 3.01
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -------Step 2------
Coef P Coef P
Constant 4.88 -2.01
Temp 1.108 0.000 1.443 0.000
171
Lighting -0.004137 0.000
S 0.991675 0.954528
R-sq 8.92% 15.71%
R-sq(adj) 8.82% 15.52%
R-sq(pred) 8.07% 14.72%
Mallows’ Cp 73.33 3.07
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1-----
Coef P
Constant 134.1
CO2 -0.0628 0.001
S 11.5471
R-sq 1.45%
R-sq(adj) 1.32%
R-sq(pred) 0.91%
Mallows’ Cp 1.67
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 10.09 222.9 317.9
Lighting 0.03482 0.000 0.04845 0.000 0.04324 0.000
Temp -9.79 0.000 -11.06 0.000
CO2 -0.0641 0.057
S 13.6257 13.3500 13.3084
R-sq 3.33% 7.42% 8.22%
R-sq(adj) 3.10% 6.98% 7.56%
R-sq(pred) 2.03% 5.77% 6.13%
Mallows’ Cp 21.95 5.20 3.56
Stepwise regression with the dataset: office zone C2 – day2 (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------
Coef P Coef P
Constant 18.765 16.720
Skin Temp -0.4680 0.000 -0.3842 0.000
Stress Level -0.01985 0.000
S 1.07876 0.988360
R-sq 32.93% 43.75%
R-sq(adj) 32.86% 43.64%
R-sq(pred) 32.69% 43.42%
Mallows’ Cp 197.67 3.30
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -11.540 -10.193 -7.627
Skin Temp 0.5098 0.000 0.5300 0.000 0.4730 0.000
HR -0.02767 0.000 -0.04484 0.000
Stress Level 0.01577 0.000
EDA
S 0.779160 0.740458 0.673731
R-sq 51.27% 56.03% 63.64%
R-sq(adj) 51.22% 55.94% 63.53%
R-sq(pred) 51.06% 55.75% 63.31%
Mallows’ Cp 333.18 208.40 8.13
------Step 4------
Coef P
Constant -8.137
172
Skin Temp 0.4888 0.000
HR -0.04395 0.000
Stress Level 0.01504 0.000
EDA -0.0523 0.024
S 0.672291
R-sq 63.83%
R-sq(adj) 63.68%
R-sq(pred) 63.46%
Mallows’ Cp 5.00
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -12.929 -11.658 -8.827
Skin Temp 0.5523 0.000 0.5713 0.000 0.5085 0.000
HR -0.02609 0.000 -0.04503 0.000
Stress Level 0.01739 0.000
EDA
S 0.891158 0.861473 0.792066
R-sq 48.55% 51.97% 59.44%
R-sq(adj) 48.50% 51.87% 59.32%
R-sq(pred) 48.34% 51.68% 59.11%
Mallows’ Cp 261.24 181.86 6.12
------Step 4------
Coef P
Constant -9.295
Skin Temp 0.5229 0.000
HR -0.04422 0.000
Stress Level 0.01673 0.000
EDA -0.0479 0.078
S 0.791196
R-sq 59.57%
R-sq(adj) 59.41%
R-sq(pred) 59.19%
Mallows’ Cp 5.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -3.899 -2.697 -1.182
Skin Temp 0.2735 0.000 0.2915 0.000 0.2579 0.000
HR -0.02468 0.000 -0.03482 0.000
Stress Level 0.009312 0.000
S 0.596283 0.555533 0.525190
R-sq 34.08% 42.84% 48.97%
R-sq(adj) 34.01% 42.72% 48.81%
R-sq(pred) 33.81% 42.49% 48.55%
Mallows’ Cp 282.12 118.63 4.91
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----
Coef P Coef P
Constant 3.6768 2.156
EDA 0.1310 0.003 0.1036 0.029
Skin Temp 0.0492 0.123
S 1.43779 1.43676
R-sq 0.91% 1.16%
R-sq(adj) 0.81% 0.95%
R-sq(pred) 0.61% 0.60%
Mallows’ Cp 2.35 1.96
EDA vs IEQ
173
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 9.11 6.10 7.47
Temp -0.3965 0.000 -0.3205 0.001 -0.389 0.000
Acoustic 0.0249 0.020 0.0253 0.018
Lighting 0.000938 0.025
S 1.05858 1.05719 1.05591
R-sq 1.13% 1.45% 1.75%
R-sq(adj) 1.07% 1.33% 1.57%
R-sq(pred) 0.96% 1.15% 1.26%
Mallows’ Cp 9.44 6.03 3.00
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant -13.85 -22.60 -32.55
Temp 2.017 0.000 2.443 0.000 2.600 0.000
Lighting -0.005255 0.000 -0.004680 0.000
CO2 0.00619 0.053
S 2.66300 2.64146 2.63944
R-sq 4.30% 5.90% 6.09%
R-sq(adj) 4.25% 5.79% 5.94%
R-sq(pred) 4.05% 5.56% 5.63%
Mallows’ Cp 34.63 6.13 4.39
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 70.267 100.3 96.4
Lighting 0.02935 0.000 0.02585 0.000 0.02569 0.000
CO2 -0.0285 0.047 -0.0360 0.016
Acoustic 0.217 0.073
S 11.1455 11.1341 11.1256
R-sq 3.23% 3.50% 3.71%
R-sq(adj) 3.17% 3.36% 3.51%
R-sq(pred) 2.96% 3.10% 3.18%
Mallows’ Cp 6.85 4.88 3.66
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 24.51 -1.7 43.9
Lighting 0.03778 0.000 0.03985 0.000 0.0345 0.001
Acoustic 0.478 0.056 0.590 0.024
CO2 -0.0490 0.142
S 22.6750 22.6508 22.6402
R-sq 1.25% 1.54% 1.71%
R-sq(adj) 1.17% 1.38% 1.47%
R-sq(pred) 0.93% 1.05% 1.05%
Mallows’ Cp 5.70 4.05 3.89
Stepwise regression with the dataset: office zone C2 – day2 (mid-age)
For each analysis: α to enter = 0.15, α to remove = 0.15
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 3.385 11.89 14.22
EDA 1.795 0.000 3.088 0.000 3.268 0.000
Skin Temp -0.293 0.016 -0.309 0.010
HR -0.0280 0.063
S 0.880343 0.846195 0.828836
174
R-sq 24.63% 31.48% 35.34%
R-sq(adj) 23.41% 29.24% 32.11%
R-sq(pred) 21.41% 25.92% 27.68%
Mallows’ Cp 10.52 6.10 4.49
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2-----
Coef P Coef P
Constant 4.3536 12.11
EDA 0.577 0.010 1.756 0.000
Skin Temp -0.2675 0.000
S 0.480532 0.418414
R-sq 10.17% 32.99%
R-sq(adj) 8.72% 30.80%
R-sq(pred) 5.96% 27.60%
Mallows’ Cp 21.10 2.49
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 2.916 -7.93 -6.13
EDA -0.895 0.004 -2.544 0.000 -2.405 0.000
Skin Temp 0.3737 0.000 0.3615 0.000
HR -0.02165 0.031
Stress Level
S 0.657818 0.568680 0.551523
R-sq 12.71% 35.81% 40.62%
R-sq(adj) 11.30% 33.71% 37.65%
R-sq(pred) 8.50% 31.04% 32.63%
Mallows’ Cp 29.90 8.10 5.15
------Step 4------
Coef P
Constant -5.35
EDA -2.316 0.000
Skin Temp 0.3265 0.000
HR -0.0165 0.117
Stress Level -0.01426 0.147
S 0.546290
R-sq 42.71%
R-sq(adj) 38.83%
R-sq(pred) 33.82%
Mallows’ Cp 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- -------Step 2------- -------Step 3-------
Coef P Coef P Coef P
Constant -7.563 -8.169 -7.576
Temp 0.3528 0.000 0.3823 0.000 0.3729 0.000
Lighting -0.000364 0.000 -0.000398 0.000
CO2 -0.000368 0.118
S 0.0998772 0.0970958 0.0969377
R-sq 49.52% 52.40% 52.66%
R-sq(adj) 49.41% 52.19% 52.35%
R-sq(pred) 48.48% 51.25% 51.25%
Mallows’ Cp 29.10 4.00 3.54
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant -45.28 -53.85 -50.79
Temp 3.4023 0.000 3.8194 0.000 3.7416 0.000
175
Lighting -0.005146 0.000 -0.005161 0.000
Acoustic -0.02489 0.007
S 0.584237 0.477628 0.474300
R-sq 72.73% 81.81% 82.11%
R-sq(adj) 72.67% 81.73% 81.99%
R-sq(pred) 72.43% 81.35% 81.53%
Mallows’ Cp 235.39 10.39 5.07
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant -142.0 -160.9 -94.0
Temp 9.33 0.000 10.26 0.000 9.13 0.000
Lighting -0.01215 0.084 -0.01587 0.032
CO2 -0.0402 0.103
S 9.28671 9.26308 9.24344
R-sq 7.38% 8.08% 8.71%
R-sq(adj) 7.14% 7.61% 8.01%
R-sq(pred) 6.39% 5.83% 6.00%
Mallows’ Cp 4.75 3.74 3.09
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1----
Coef P
Constant 204.9
Temp -8.95 0.001
S 7.87850
R-sq 13.69%
R-sq(adj) 12.48%
R-sq(pred) 4.76%
Mallows’ Cp -0.42
Stepwise regression with the dataset: office zone C2 – day3
For each analysis: α to enter = 0.15, α to remove = 0.15
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 5.4165 -0.269 -0.714
EDA -10.983 0.000 -11.491 0.000 -11.468 0.000
Skin Temp 0.1819 0.000 0.1927 0.000
Stress Level 0.002979 0.000
HR
S 0.529717 0.472119 0.466679
R-sq 64.31% 71.68% 72.35%
R-sq(adj) 64.28% 71.63% 72.28%
R-sq(pred) 64.14% 71.47% 72.10%
Mallows’ Cp 340.31 34.81 8.70
------Step 4------
Coef P
Constant -0.750
EDA -11.133 0.000
Skin Temp 0.2045 0.000
Stress Level 0.004418 0.000
HR -0.00556 0.017
S 0.465728
R-sq 72.48%
R-sq(adj) 72.39%
R-sq(pred) 72.17%
Mallows’ Cp 5.00
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
176
-----Step 1---- -----Step 2---- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.9031 0.187 0.673 -0.511
EDA -9.395 0.000 -9.816 0.000 -8.933 0.000 -7.105 0.000
Skin Temp 0.1509 0.000 0.1689 0.000 0.2559 0.000
HR -0.01515 0.000 -0.04465 0.000
Stress Level 0.01428 0.000
S 1.05362 1.03499 1.02710 0.999583
R-sq 25.00% 27.69% 28.85% 32.67%
R-sq(adj) 24.93% 27.56% 28.66% 32.44%
R-sq(pred) 24.81% 27.43% 28.47% 32.13%
Mallows’ Cp 129.91 85.98 68.18 5.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2-----
Coef P Coef P
Constant 5.2715 3.459
EDA -7.568 0.000 -7.729 0.000
Skin Temp 0.05798 0.000
S 0.409298 0.402221
R-sq 58.90% 60.34%
R-sq(adj) 58.86% 60.27%
R-sq(pred) 58.73% 60.11%
Mallows’ Cp 41.71 1.86
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.7871 4.4864 -1.129
EDA -5.974 0.000 -5.825 0.000 -6.288 0.000
Stress Level 0.00838 0.000 0.01020 0.000
Skin Temp 0.1775 0.000
HR
S 0.951136 0.928381 0.899686
R-sq 14.19% 18.32% 23.35%
R-sq(adj) 14.11% 18.17% 23.15%
R-sq(pred) 13.83% 17.83% 22.80%
Mallows’ Cp 138.94 78.94 5.26
------Step 4------
Coef P
Constant -1.173
EDA -5.881 0.000
Stress Level 0.01195 0.000
Skin Temp 0.1919 0.000
HR -0.00675 0.133
S 0.899193
R-sq 23.50%
R-sq(adj) 23.24%
R-sq(pred) 22.83%
Mallows’ Cp 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------
Coef P
Constant -0.1178
CO2 0.000169 0.000
S 0.0674072
R-sq 0.86%
R-sq(adj) 0.81%
R-sq(pred) 0.66%
Mallows’ Cp 4.63
177
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- ------Step 3------ ------Step 4------
Coef P Coef P Coef P Coef P
Constant 15.381 18.56 17.80 14.77
Temp 0.7126 0.000 0.6934 0.000 0.7399 0.000 0.7527 0.000
Acoustic -0.0512 0.022 -0.0552 0.014 -0.0489 0.031
Lighting -0.000204 0.062 -0.000282 0.015
CO2 0.00210 0.049
S 1.73593 1.73425 1.73326 1.73212
R-sq 10.59% 10.80% 10.94% 11.10%
R-sq(adj) 10.54% 10.72% 10.82% 10.94%
R-sq(pred) 10.46% 10.57% 10.66% 10.67%
Mallows’ Cp 11.66 8.38 6.88 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 73.869 57.05 37.3
Lighting 0.003660 0.000 0.002828 0.000 0.002964 0.000
Temp 0.778 0.038 0.864 0.022
Acoustic 0.332 0.049
S 11.7519 11.7412 11.7319
R-sq 1.56% 1.79% 2.00%
R-sq(adj) 1.50% 1.68% 1.84%
R-sq(pred) 1.35% 1.49% 1.59%
Mallows’ Cp 7.40 5.09 3.22
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 32.080 -23.9 -75.8
Lighting 0.00777 0.000 0.00832 0.000 0.00621 0.000
Acoustic 1.040 0.007 1.137 0.003
Temp 2.152 0.010
S 24.6694 24.6179 24.5720
R-sq 1.71% 2.19% 2.62%
R-sq(adj) 1.65% 2.06% 2.42%
R-sq(pred) 1.46% 1.80% 2.12%
Mallows’ Cp 14.19 8.87 4.27
Stepwise regression with the dataset: office zone C2 – day3 (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.3844 6.793 7.561
Stress Level 0.00764 0.000 0.01721 0.000 0.02144 0.000
HR -0.03682 0.000 -0.05318 0.000
EDA 5.82 0.000
S 1.10098 1.06395 1.04945
R-sq 2.35% 8.94% 11.54%
R-sq(adj) 2.21% 8.68% 11.15%
R-sq(pred) 1.82% 8.21% 10.61%
Mallows’ Cp 70.66 21.40 3.23
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2-----
Coef P Coef P
Constant 4.3132 4.5501
Stress Level 0.01191 0.000 0.01194 0.000
EDA -4.570 0.000
178
S 0.930288 0.914815
R-sq 7.59% 10.77%
R-sq(adj) 7.46% 10.51%
R-sq(pred) 7.07% 10.06%
Mallows’ Cp 24.65 2.10
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------
Coef P Coef P
Constant 0.1160 0.0509
Temp -0.00292 0.017 -0.00310 0.011
CO2 0.000060 0.035
S 0.0356104 0.0355576
R-sq 0.49% 0.87%
R-sq(adj) 0.40% 0.70%
R-sq(pred) 0.19% 0.30%
Mallows’ Cp 4.97 2.53
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 18.61 10.12 15.11
CO2 0.01111 0.000 0.01030 0.000 0.00978 0.000
Temp 0.4311 0.000 0.4062 0.000
Acoustic -0.0714 0.027
S 1.95086 1.91805 1.91521
R-sq 4.38% 7.64% 7.98%
R-sq(adj) 4.31% 7.50% 7.77%
R-sq(pred) 3.87% 7.09% 7.25%
Mallows’ Cp 52.36 7.82 4.91
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 37.2 13.6
CO2 0.03120 0.001 0.02980 0.002
Temp 1.155 0.005
S 11.7183 11.6825
R-sq 0.90% 1.59%
R-sq(adj) 0.81% 1.41%
R-sq(pred) 0.57% 1.11%
Mallows’ Cp 7.78 1.78
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----
Coef P Coef P
Constant -44.1 -92.9
CO2 0.0619 0.002 0.0582 0.004
Temp 2.415 0.006
S 22.5839 22.5016
R-sq 1.04% 1.87%
R-sq(adj) 0.93% 1.65%
R-sq(pred) 0.58% 1.19%
Mallows’ Cp 9.03 3.38
Stepwise regression with the dataset: office zone C2 – day3 (male)
For each analysis: α to enter = 0.15, α to remove = 0.15
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 5.3618 4.8153 7.581
EDA -12.804 0.000 -11.813 0.000 -8.074 0.000
179
Stress Level 0.00968 0.000 0.02530 0.000
HR -0.05091 0.000
Skin Temp
S 0.575948 0.534825 0.467782
R-sq 77.90% 80.98% 85.48%
R-sq(adj) 77.85% 80.90% 85.39%
R-sq(pred) 77.68% 80.69% 85.14%
Mallows’ Cp 284.21 182.75 33.82
------Step 4------
Coef P
Constant 4.878
EDA -8.083 0.000
Stress Level 0.02675 0.000
HR -0.05855 0.000
Skin Temp 0.1053 0.000
S 0.453229
R-sq 86.40%
R-sq(adj) 86.28%
R-sq(pred) 85.99%
Mallows’ Cp 5.00
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 4.7398 6.150 7.875
EDA -10.404 0.000 -9.576 0.000 -6.056 0.000
HR -0.01968 0.000 -0.05834 0.000
Stress Level 0.01851 0.000
Skin Temp
S 0.791102 0.773984 0.734374
R-sq 55.22% 57.23% 61.58%
R-sq(adj) 55.13% 57.05% 61.33%
R-sq(pred) 54.87% 56.64% 60.92%
Mallows’ Cp 78.45 56.41 6.40
------Step 4------
Coef P
Constant 6.424
EDA -6.061 0.000
HR -0.06244 0.000
Stress Level 0.01928 0.000
Skin Temp 0.0565 0.066
S 0.732453
R-sq 61.87%
R-sq(adj) 61.53%
R-sq(pred) 61.07%
Mallows’ Cp 5.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 5.1960 11.236 11.382
EDA -8.673 0.000 -7.760 0.000 -7.250 0.000
Skin Temp -0.2007 0.000 -0.2138 0.000
Stress Level 0.004400 0.000
HR
S 0.482427 0.419576 0.408744
R-sq 69.74% 77.16% 78.37%
R-sq(adj) 69.68% 77.06% 78.23%
R-sq(pred) 69.52% 76.86% 77.96%
Mallows’ Cp 287.11 106.62 78.84
------Step 4------
180
Coef P
Constant 11.819
EDA -5.151 0.000
Skin Temp -0.1702 0.000
Stress Level 0.01378 0.000
HR -0.03179 0.000
S 0.378897
R-sq 81.46%
R-sq(adj) 81.30%
R-sq(pred) 80.91%
Mallows’ Cp 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 2.9093 5.206 2.876
Stress Level 0.02286 0.000 0.03027 0.000 0.03155 0.000
HR -0.03487 0.000 -0.04154 0.000
Skin Temp 0.0909 0.005
S 0.822090 0.777688 0.771879
R-sq 30.55% 37.98% 39.04%
R-sq(adj) 30.39% 37.71% 38.64%
R-sq(pred) 29.91% 37.03% 37.75%
Mallows’ Cp 64.50 10.67 4.74
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------
Coef P Coef P
Constant -0.2198 -0.1755
CO2 0.000286 0.000 0.000246 0.002
Lighting 0.000011 0.138
S 0.0828508 0.0827939
R-sq 1.66% 1.91%
R-sq(adj) 1.55% 1.68%
R-sq(pred) 1.21% 1.13%
Mallows’ Cp 2.23 2.03
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 6.695 19.094 16.10
Temp 1.0766 0.000 1.1202 0.000 1.2003 0.000
CO2 -0.011552 0.000 -0.010410 0.000
Lighting -0.000359 0.000
S 0.868537 0.753404 0.745077
R-sq 51.94% 63.88% 64.71%
R-sq(adj) 51.89% 63.80% 64.59%
R-sq(pred) 51.70% 63.62% 64.34%
Mallows’ Cp 315.89 22.31 3.66
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 75.820 125.0 91.9
Lighting 0.00513 0.000 0.00631 0.000 0.00664 0.000
CO2 -0.0426 0.001 -0.0401 0.002
Acoustic 0.561 0.034
S 11.3989 11.3111 11.2812
R-sq 3.48% 5.10% 5.75%
R-sq(adj) 3.33% 4.81% 5.32%
R-sq(pred) 2.88% 4.24% 4.55%
181
Mallows’ Cp 15.56 6.16 3.63
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 41.49 209.2 91.1
Lighting 0.00938 0.000 0.01345 0.000 0.01426 0.000
CO2 -0.1457 0.000 -0.1387 0.000
Acoustic 2.047 0.001
S 24.4549 23.9501 23.7336
R-sq 2.74% 6.87% 8.70%
R-sq(adj) 2.57% 6.55% 8.23%
R-sq(pred) 2.12% 5.85% 7.44%
Mallows’ Cp 38.08 13.40 3.58
Stepwise regression with the dataset: office zone C2 – day3 (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2-----
Coef P Coef P
Constant 5.3971 -5.322
EDA -10.937 0.000 -10.735 0.000
Skin Temp 0.3356 0.000
S 0.610589 0.478701
R-sq 62.72% 77.11%
R-sq(adj) 62.67% 77.06%
R-sq(pred) 62.52% 76.92%
Mallows’ Cp 537.06 2.02
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2---- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 4.9327 -4.37 -4.33 -4.28
EDA -9.367 0.000 -9.192 0.000 -8.260 0.000 -6.600 0.000
Skin Temp 0.2914 0.000 0.3214 0.000 0.3780 0.000
HR -0.01456 0.001 -0.04826 0.000
Stress Level 0.01831 0.000
S 1.13242 1.08409 1.07709 1.03418
R-sq 26.40% 32.63% 33.58% 38.83%
R-sq(adj) 26.32% 32.47% 33.34% 38.55%
R-sq(pred) 26.16% 32.35% 33.14% 38.25%
Mallows’ Cp 172.35 87.50 76.32 5.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- -------Step 3------
Coef P Coef P Coef P
Constant 5.2733 2.015 2.018
EDA -7.573 0.000 -7.512 0.000 -7.483 0.000
Skin Temp 0.1020 0.000 0.1028 0.000
Stress Level -0.001079 0.100
HR
S 0.459267 0.444812 0.444368
R-sq 58.78% 61.38% 61.50%
R-sq(adj) 58.73% 61.29% 61.36%
R-sq(pred) 58.58% 61.09% 61.10%
Mallows’ Cp 61.77 6.03 5.32
-------Step 4------
Coef P
Constant 2.010
EDA -7.691 0.000
Skin Temp 0.0959 0.000
182
Stress Level -0.002059 0.025
HR 0.00367 0.128
S 0.444026
R-sq 61.60%
R-sq(adj) 61.42%
R-sq(pred) 61.11%
Mallows’ Cp 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 4.8755 4.5218 -0.989 -0.960
EDA -6.371 0.000 -6.741 0.000 -6.630 0.000 -5.896 0.000
Stress Level 0.01382 0.000 0.01353 0.000 0.01698 0.000
Skin Temp 0.1728 0.000 0.1970 0.000
HR -0.01293 0.016
S 1.06075 1.01182 0.993374 0.990609
R-sq 15.91% 23.58% 26.42% 26.92%
R-sq(adj) 15.81% 23.40% 26.16% 26.57%
R-sq(pred) 15.49% 23.04% 25.76% 26.14%
Mallows’ Cp 127.51 40.01 8.77 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant -0.2243 -0.1429 -0.0480
CO2 0.000270 0.000 0.000277 0.000 0.000241 0.000
Temp -0.00409 0.055 -0.00664 0.007
Lighting 0.000012 0.041
S 0.0722673 0.0722067 0.0721350
R-sq 1.88% 2.11% 2.36%
R-sq(adj) 1.82% 1.98% 2.18%
R-sq(pred) 1.63% 1.73% 1.83%
Mallows’ Cp 7.03 5.34 3.16
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2----- -----Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 15.76 9.25 13.14 11.07
Temp 0.7074 0.000 0.6845 0.000 0.6652 0.000 0.7237 0.000
CO2 0.00606 0.000 0.00566 0.000 0.00649 0.000
Acoustic -0.0557 0.031 -0.0578 0.026
Lighting -0.000266 0.046
S 1.78843 1.77446 1.77262 1.77111
R-sq 9.91% 11.36% 11.59% 11.79%
R-sq(adj) 9.85% 11.26% 11.44% 11.59%
R-sq(pred) 9.76% 10.98% 11.09% 11.24%
Mallows’ Cp 36.47 9.64 7.00 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 73.895 48.69 19.8
Lighting 0.004149 0.000 0.002906 0.001 0.002122 0.029
Temp 1.165 0.007 1.316 0.003
CO2 0.02223 0.021
Acoustic
S 12.0581 12.0315 12.0131
R-sq 1.93% 2.43% 2.80%
R-sq(adj) 1.86% 2.30% 2.59%
R-sq(pred) 1.66% 2.04% 2.29%
183
Mallows’ Cp 15.73 10.44 7.11
------Step 4-----
Coef P
Constant -6.0
Lighting 0.002226 0.022
Temp 1.428 0.001
CO2 0.02381 0.014
Acoustic 0.400 0.043
S 11.9999
R-sq 3.08%
R-sq(adj) 2.81%
R-sq(pred) 2.43%
Mallows’ Cp 5.00
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2----- -----Step 3----- -----Step 4-----
Coef P Coef P Coef P Coef P
Constant 25.503 -48.7 -110.3 -162.2
Lighting 0.00890 0.000 0.00958 0.000 0.00711 0.000 0.00583 0.003
Acoustic 1.380 0.001 1.486 0.000 1.565 0.000
Temp 2.574 0.004 2.805 0.002
CO2 0.0370 0.062
S 22.9062 22.8022 22.7272 22.7021
R-sq 2.60% 3.57% 4.29% 4.59%
R-sq(adj) 2.52% 3.40% 4.04% 4.25%
R-sq(pred) 2.25% 3.03% 3.58% 3.70%
Mallows’ Cp 22.43 12.98 6.50 5.00
Stepwise regression with the dataset: office zone C2 – day3 (mid-age)
For each analysis: α to enter = 0.15, α to remove = 0.15
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 6.600 7.683 2.96
EDA -51.92 0.000 -45.92 0.000 -58.79 0.000
HR -0.01826 0.001 -0.02025 0.000
Skin Temp 0.1791 0.000
S 0.661080 0.648853 0.635838
R-sq 36.36% 38.90% 41.53%
R-sq(adj) 36.14% 38.48% 40.93%
R-sq(pred) 34.77% 36.93% 37.14%
Mallows’ Cp 25.60 14.92 3.81
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 5.688 5.725 4.792
HR -0.01011 0.000 -0.00728 0.000 -0.00768 0.000
EDA -5.78 0.000 -8.33 0.000
Skin Temp 0.0354 0.022
S 0.205102 0.199000 0.197552
R-sq 13.15% 18.52% 19.97%
R-sq(adj) 12.85% 17.96% 19.15%
R-sq(pred) 10.92% 14.22% 14.53%
Mallows’ Cp 24.15 6.61 3.31
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3------
Coef P Coef P Coef P
Constant 3.8529 4.357 5.130
Stress Level 0.00949 0.000 0.01257 0.000 0.01871 0.000
184
EDA -15.98 0.000 -14.79 0.000
HR -0.01584 0.036
S 0.454770 0.432167 0.429656
R-sq 12.03% 20.83% 22.02%
R-sq(adj) 11.73% 20.29% 21.21%
R-sq(pred) 10.98% 18.32% 19.03%
Mallows’ Cp 36.44 5.60 3.17
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------- -------Step 2-------
Coef P Coef P
Constant 0.2328 0.1967
CO2 -0.000166 0.000 -0.000170 0.000
Temp 0.001830 0.002
S 0.0103251 0.0102226
R-sq 26.90% 28.51%
R-sq(adj) 26.73% 28.18%
R-sq(pred) 25.71% 27.18%
Mallows’ Cp 10.46 2.67
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 13.87 31.54 28.66
Temp 0.7334 0.000 0.7956 0.000 0.8728 0.000
CO2 -0.016463 0.000 -0.015361 0.000
Lighting -0.000346 0.001
S 0.939268 0.708462 0.700577
R-sq 30.06% 60.30% 61.27%
R-sq(adj) 29.90% 60.12% 61.00%
R-sq(pred) 29.53% 59.79% 60.62%
Mallows’ Cp 349.49 12.30 3.44
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 127.0 140.3
CO2 -0.0458 0.001 -0.0579 0.000
Lighting 0.00352 0.014
S 10.4510 10.3847
R-sq 2.66% 4.14%
R-sq(adj) 2.42% 3.65%
R-sq(pred) 1.64% 2.90%
Mallows’ Cp 7.08 3.00
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 154.7 200.9
CO2 -0.0880 0.001 -0.1302 0.000
Lighting 0.01128 0.000
S 19.3686 18.8917
R-sq 2.91% 7.88%
R-sq(adj) 2.64% 7.38%
R-sq(pred) 1.85% 6.53%
Mallows’ Cp 20.68 2.82
Stepwise regression with the dataset: office zone C2 – day4
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
185
Constant 4.2769 4.3842 5.979
EDA -3.961 0.000 -3.524 0.000 -3.506 0.000
Stress Level -0.00966 0.000 -0.00957 0.000
Skin Temp -0.0512 0.050
S 1.03844 1.02642 1.02511
R-sq 18.86% 20.80% 21.07%
R-sq(adj) 18.79% 20.66% 20.86%
R-sq(pred) 17.82% 19.64% 19.80%
Mallows’ Cp 31.61 6.17 4.32
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 4.7923 6.214 8.646
EDA -3.161 0.000 -2.624 0.000 -2.656 0.000
HR -0.02130 0.000 -0.01865 0.000
Skin Temp -0.0837 0.000
Stress Level
S 0.884844 0.868895 0.863973
R-sq 17.81% 20.82% 21.79%
R-sq(adj) 17.73% 20.67% 21.56%
R-sq(pred) 16.62% 19.61% 20.51%
Mallows’ Cp 59.57 21.56 10.68
------Step 4------
Coef P
Constant 8.376
EDA -2.607 0.000
HR -0.00945 0.049
Skin Temp -0.0926 0.000
Stress Level -0.00639 0.006
S 0.861193
R-sq 22.37%
R-sq(adj) 22.07%
R-sq(pred) 20.98%
Mallows’ Cp 5.00
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 4.5471 5.758 2.620 2.987
EDA -4.345 0.000 -3.888 0.000 -3.847 0.000 -3.913 0.000
HR -0.01813 0.000 -0.02156 0.000 -0.03405 0.000
Skin Temp 0.1080 0.001 0.1200 0.000
Stress Level 0.00868 0.006
S 1.18767 1.17945 1.17348 1.16970
R-sq 18.51% 19.71% 20.60% 21.19%
R-sq(adj) 18.43% 19.56% 20.37% 20.88%
R-sq(pred) 17.97% 19.10% 19.82% 20.23%
Mallows’ Cp 34.01 20.30 10.67 5.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 4.7342 6.684 9.891 9.511
EDA -4.324 0.000 -3.587 0.000 -3.629 0.000 -3.560 0.000
HR -0.02920 0.000 -0.02570 0.000 -0.01274 0.035
Skin Temp -0.1104 0.000 -0.1229 0.000
Stress Level -0.00900 0.002
S 1.12012 1.09631 1.08948 1.08497
R-sq 20.18% 23.61% 24.64% 25.33%
R-sq(adj) 20.11% 23.47% 24.42% 25.04%
186
R-sq(pred) 18.87% 22.31% 23.29% 23.88%
Mallows’ Cp 70.07 24.71 12.59 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 5.863 9.424 9.870
HR -0.02550 0.000 -0.02189 0.000 -0.03708 0.000
Skin Temp -0.1221 0.000 -0.1077 0.001
Stress Level 0.01023 0.001
EDA
S 1.15148 1.14345 1.13778
R-sq 3.50% 4.94% 5.97%
R-sq(adj) 3.41% 4.75% 5.70%
R-sq(pred) 3.17% 4.47% 5.35%
Mallows’ Cp 35.74 21.88 12.46
------Step 4-----
Coef P
Constant 9.735
HR -0.03283 0.000
Skin Temp -0.1103 0.000
Stress Level 0.01096 0.000
EDA -0.904 0.002
S 1.13314
R-sq 6.82%
R-sq(adj) 6.46%
R-sq(pred) 5.99%
Mallows’ Cp 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------
Coef P
Constant 0.8231
CO2 -0.000678 0.000
S 0.123563
R-sq 2.85%
R-sq(adj) 2.80%
R-sq(pred) 2.57%
Mallows’ Cp 1.04
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 19.02 27.94 28.86
CO2 0.010920 0.000 0.00924 0.000 0.00850 0.000
Temp -0.313 0.005 -0.376 0.001
Acoustic 0.0244 0.040
S 1.36254 1.36033 1.35928
R-sq 5.83% 6.18% 6.37%
R-sq(adj) 5.79% 6.09% 6.24%
R-sq(pred) 5.65% 5.92% 6.04%
Mallows’ Cp 11.89 6.07 3.84
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1----- -----Step 2----- -----Step 3-----
Coef P Coef P Coef P
Constant 152.3 67.4 59.8
CO2 -0.07417 0.000 -0.0577 0.000 -0.0512 0.000
Temp 2.95 0.008 3.51 0.002
Acoustic -0.222 0.064
S 13.1290 13.1080 13.0994
187
R-sq 3.08% 3.44% 3.62%
R-sq(adj) 3.03% 3.34% 3.46%
R-sq(pred) 2.85% 3.11% 3.19%
Mallows’ Cp 9.80 4.81 3.37
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 12.250 -81.9
Lighting 0.01613 0.000 0.01130 0.008
Temp 4.19 0.008
S 16.8708 16.8341
R-sq 1.26% 1.76%
R-sq(adj) 1.18% 1.61%
R-sq(pred) 0.95% 1.31%
Mallows’ Cp 6.84 1.88
Stepwise regression with the dataset: office zone C2 – day4 (female)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3----- ------Step 4------
Coef P Coef P Coef P Coef P
Constant 1.643 1.606 1.663 2.114
Skin Temp 0.0964 0.000 0.0991 0.000 0.0961 0.000 0.1246 0.000
EDA -0.672 0.000 -0.714 0.000 -0.471 0.005
Stress Level 0.00302 0.006 0.01038 0.000
HR -0.02088 0.000
S 0.462360 0.457098 0.454806 0.440579
R-sq 6.67% 8.92% 9.97% 15.64%
R-sq(adj) 6.53% 8.65% 9.56% 15.13%
R-sq(pred) 6.17% 8.05% 8.82% 14.54%
Mallows’ Cp 68.97 53.42 47.25 5.00
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2---- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 4.6215 -2.91 -2.80 -2.35
EDA -3.151 0.000 -3.314 0.000 -3.168 0.000 -3.049 0.000
Skin Temp 0.2425 0.000 0.2575 0.000 0.2738 0.000
HR -0.00859 0.146 -0.02437 0.004
Stress Level 0.01057 0.009
S 1.18561 1.14320 1.14217 1.13675
R-sq 7.92% 14.52% 14.82% 15.76%
R-sq(adj) 7.77% 14.25% 14.40% 15.21%
R-sq(pred) 6.93% 12.85% 13.05% 13.86%
Mallows’ Cp 56.26 10.02 9.89 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.1870 6.198 6.496
Stress Level 0.01209 0.000 0.02334 0.000 0.02457 0.000
HR -0.03116 0.000 -0.03700 0.000
EDA 1.303 0.004
S 1.20428 1.19111 1.18394
R-sq 2.64% 4.91% 6.20%
R-sq(adj) 2.48% 4.60% 5.75%
R-sq(pred) 2.21% 4.31% 5.44%
Mallows’ Cp 23.17 10.25 3.75
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
------Step 1------ -------Step 2------ -------Step 3------
188
Coef P Coef P Coef P
Constant 0.03273 0.491 1.664
Lighting 0.000187 0.000 0.000150 0.000 0.000184 0.000
CO2 -0.000416 0.000 -0.000602 0.000
Temp -0.0431 0.001
S 0.107590 0.106920 0.106399
R-sq 3.57% 4.85% 5.86%
R-sq(adj) 3.48% 4.68% 5.61%
R-sq(pred) 2.41% 3.47% 4.35%
Mallows’ Cp 27.46 14.19 4.17
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 24.39 23.42
CO2 0.00589 0.000 0.00546 0.000
Acoustic 0.0268 0.107
S 1.53062 1.52964
R-sq 1.42% 1.62%
R-sq(adj) 1.34% 1.46%
R-sq(pred) 1.11% 1.19%
Mallows’ Cp 2.00 1.40
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
----Step 1---- -----Step 2----
Coef P Coef P
Constant -31.4 -23.8
Temp 4.49 0.000 4.75 0.000
Acoustic -0.250 0.063
S 11.9963 11.9829
R-sq 1.40% 1.71%
R-sq(adj) 1.31% 1.53%
R-sq(pred) 1.05% 1.19%
Mallows’ Cp 3.97 2.52
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1-----
Coef P
Constant 9.682
Lighting 0.01587 0.001
S 15.5533
R-sq 1.33%
R-sq(adj) 1.21%
R-sq(pred) 0.83%
Mallows’ Cp 0.91
Stepwise regression with the dataset: office zone C2 – day4 (male)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 7.793 15.37 14.865
HR -0.07103 0.000 -0.06919 0.000 -0.05454 0.000
Skin Temp -0.2460 0.000 -0.2548 0.000
EDA -1.814 0.000
Stress Level
S 0.735564 0.694403 0.659910
R-sq 44.61% 50.74% 55.61%
R-sq(adj) 44.48% 50.52% 55.32%
R-sq(pred) 44.20% 50.19% 54.97%
Mallows’ Cp 144.33 80.07 29.47
189
------Step 4------
Coef P
Constant 15.110
HR -0.07501 0.000
Skin Temp -0.2248 0.000
EDA -1.983 0.000
Stress Level 0.01381 0.000
S 0.642146
R-sq 58.06%
R-sq(adj) 57.69%
R-sq(pred) 57.23%
Mallows’ Cp 5.00
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 9.354 17.06 16.76
HR -0.08102 0.000 -0.07875 0.000 -0.06143 0.000
Skin Temp -0.2510 0.000 -0.2715 0.000
EDA -1.901 0.000
Stress Level
S 0.944946 0.917112 0.887889
R-sq 37.18% 40.97% 44.80%
R-sq(adj) 37.03% 40.68% 44.40%
R-sq(pred) 36.72% 40.22% 43.96%
Mallows’ Cp 80.62 52.91 24.82
------Step 4------
Coef P
Constant 17.23
HR -0.08669 0.000
Skin Temp -0.2405 0.000
EDA -2.168 0.000
Stress Level 0.01825 0.000
S 0.866267
R-sq 47.59%
R-sq(adj) 47.08%
R-sq(pred) 46.45%
Mallows’ Cp 5.00
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 9.464 8.134 17.90
HR -0.08263 0.000 -0.05826 0.000 -0.05360 0.000
EDA -2.704 0.000 -2.899 0.000
Skin Temp -0.3211 0.000
Stress Level
S 1.10703 1.05698 1.01585
R-sq 30.96% 37.22% 42.15%
R-sq(adj) 30.80% 36.91% 41.73%
R-sq(pred) 30.42% 36.49% 41.32%
Mallows’ Cp 102.68 58.07 23.30
------Step 4------
Coef P
Constant 18.42
HR -0.08152 0.000
EDA -3.194 0.000
Skin Temp -0.2867 0.000
Stress Level 0.02017 0.000
S 0.992862
R-sq 44.87%
190
R-sq(adj) 44.33%
R-sq(pred) 43.77%
Mallows’ Cp 5.00
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 11.017 21.30 20.86
HR -0.11104 0.000 -0.10801 0.000 -0.08339 0.000
Skin Temp -0.3348 0.000 -0.3639 0.000
EDA -2.701 0.000
Stress Level
S 1.11162 1.06895 1.01726
R-sq 44.54% 48.84% 53.78%
R-sq(adj) 44.41% 48.60% 53.45%
R-sq(pred) 44.12% 48.19% 53.04%
Mallows’ Cp 113.44 74.70 29.89
------Step 4------
Coef P
Constant 21.46
HR -0.11533 0.000
Skin Temp -0.3246 0.000
EDA -3.039 0.000
Stress Level 0.02307 0.000
S 0.986721
R-sq 56.62%
R-sq(adj) 56.20%
R-sq(pred) 55.66%
Mallows’ Cp 5.00
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
------Step 1------ ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 7.826 17.12 17.40
HR -0.05888 0.000 -0.05614 0.000 -0.07088 0.000
Skin Temp -0.3026 0.000 -0.2846 0.000
Stress Level 0.00971 0.006
EDA
S 0.847660 0.801120 0.794847
R-sq 27.97% 35.82% 36.97%
R-sq(adj) 27.80% 35.51% 36.51%
R-sq(pred) 27.41% 34.98% 35.79%
Mallows’ Cp 65.98 15.90 10.25
------Step 4------
Coef P
Constant 17.30
HR -0.06539 0.000
Skin Temp -0.2910 0.000
Stress Level 0.01127 0.002
EDA -0.861 0.007
S 0.788890
R-sq 38.07%
R-sq(adj) 37.46%
R-sq(pred) 36.79%
Mallows’ Cp 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant 1.018 1.346 -0.642
CO2 -0.000828 0.000 -0.001095 0.000 -0.000774 0.000
191
Lighting -0.000243 0.000 -0.000300 0.000
Temp 0.0727 0.000
S 0.135420 0.133114 0.131958
R-sq 3.46% 6.83% 8.55%
R-sq(adj) 3.35% 6.61% 8.22%
R-sq(pred) 2.96% 6.12% 7.73%
Mallows’ Cp 45.98 17.11 3.36
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2----- -------Step 3------
Coef P Coef P Coef P
Constant 10.93 41.53 39.83
CO2 0.01849 0.000 0.01279 0.000 0.01246 0.000
Temp -1.077 0.000 -0.981 0.000
Lighting -0.000762 0.024
Acoustic
S 1.00783 0.968095 0.965730
R-sq 24.39% 30.32% 30.74%
R-sq(adj) 24.30% 30.15% 30.49%
R-sq(pred) 23.97% 29.73% 29.85%
Mallows’ Cp 79.17 9.27 6.14
-------Step 4------
Coef P
Constant 40.84
CO2 0.01176 0.000
Temp -1.049 0.000
Lighting -0.000717 0.034
Acoustic 0.0237 0.077
S 0.964497
R-sq 31.00%
R-sq(adj) 30.67%
R-sq(pred) 29.96%
Mallows’ Cp 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 216.7 228.5 136.2
CO2 -0.1318 0.000 -0.1415 0.000 -0.1256 0.000
Lighting -0.00769 0.112 -0.01068 0.038
Temp 3.33 0.097
S 14.1737 14.1596 14.1434
R-sq 8.04% 8.34% 8.67%
R-sq(adj) 7.92% 8.10% 8.31%
R-sq(pred) 7.50% 7.65% 7.76%
Mallows’ Cp 5.53 4.98 4.22
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2-----
Coef P Coef P
Constant -253.7 -112.1
Temp 12.05 0.000 8.90 0.002
CO2 -0.0647 0.037
S 17.7515 17.6973
R-sq 4.39% 5.14%
R-sq(adj) 4.21% 4.80%
R-sq(pred) 3.66% 3.98%
Mallows’ Cp 3.83 1.48
Stepwise regression with the dataset: office zone C2 – day4 (junior)
For each analysis: α to enter = 0.15, α to remove = 0.15
Overall IEQ vs bio
192
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 4.3067 6.019 5.487 6.89
EDA -4.043 0.000 -3.498 0.000 -3.451 0.000 -3.466 0.000
HR -0.02493 0.000 -0.01624 0.017 -0.01307 0.067
Stress Level -0.00546 0.085 -0.00643 0.047
Skin Temp -0.0516 0.122
S 1.13867 1.12158 1.12039 1.11955
R-sq 18.98% 21.48% 21.73% 21.93%
R-sq(adj) 18.90% 21.31% 21.48% 21.60%
R-sq(pred) 17.87% 20.31% 20.40% 20.47%
Mallows’ Cp 33.97 6.36 5.39 5.00
Acoustic satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- ------Step 2------ ------Step 3------
Coef P Coef P Coef P
Constant 4.6947 4.8143 8.526
EDA -2.941 0.000 -2.502 0.000 -2.491 0.000
Stress Level -0.01049 0.000 -0.00986 0.000
Skin Temp -0.1192 0.000
S 0.958294 0.942685 0.932996
R-sq 15.78% 18.59% 20.35%
R-sq(adj) 15.68% 18.40% 20.07%
R-sq(pred) 14.55% 17.26% 18.97%
Mallows’ Cp 48.00 19.99 3.26
Lighting satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- -----Step 2---- ------Step 3-----
Coef P Coef P Coef P
Constant 4.3653 0.97 1.13
EDA -3.931 0.000 -3.964 0.000 -3.692 0.000
Skin Temp 0.1089 0.003 0.1318 0.001
HR -0.01284 0.024
S 1.26531 1.25967 1.25665
R-sq 16.10% 16.95% 17.44%
R-sq(adj) 16.00% 16.75% 17.15%
R-sq(pred) 15.55% 16.15% 16.51%
Mallows’ Cp 14.61 7.92 4.83
IAQ satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1---- ------Step 2----- ------Step 3-----
Coef P Coef P Coef P
Constant 4.6060 4.7724 9.71
EDA -4.035 0.000 -3.423 0.000 -3.410 0.000
Stress Level -0.01459 0.000 -0.01376 0.000
Skin Temp -0.1585 0.000
S 1.21135 1.18725 1.17356
R-sq 18.07% 21.39% 23.29%
R-sq(adj) 17.98% 21.21% 23.01%
R-sq(pred) 16.72% 19.96% 21.84%
Mallows’ Cp 56.86 22.15 3.23
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1----- -----Step 2----- ------Step 3----- ------Step 4-----
Coef P Coef P Coef P Coef P
Constant 9.20 9.17 9.28 9.69
Skin Temp -0.1674 0.000 -0.1633 0.000 -0.1477 0.000 -0.1309 0.001
EDA -0.900 0.002 -0.715 0.025 -0.783 0.014
HR -0.00874 0.114 -0.02394 0.003
Stress Level 0.00941 0.010
193
S 1.22409 1.21822 1.21714 1.21315
R-sq 2.48% 3.52% 3.81% 4.55%
R-sq(adj) 2.36% 3.30% 3.47% 4.10%
R-sq(pred) 2.07% 2.91% 3.03% 3.56%
Mallows’ Cp 17.41 10.11 9.59 5.00
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------
Coef P
Constant 1.049
CO2 -0.000875 0.000
S 0.136463
R-sq 3.84%
R-sq(adj) 3.78%
R-sq(pred) 3.49%
Mallows’ Cp 1.89
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 22.59 21.78
CO2 0.00769 0.000 0.00732 0.000
Acoustic 0.0225 0.090
S 1.40701 1.40622
R-sq 2.79% 2.96%
R-sq(adj) 2.73% 2.84%
R-sq(pred) 2.56% 2.64%
Mallows’ Cp 2.53 1.66
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 141.0 47.6
CO2 -0.0627 0.000 -0.0449 0.000
Temp 3.26 0.007
S 12.7410 12.7142
R-sq 2.34% 2.82%
R-sq(adj) 2.28% 2.69%
R-sq(pred) 2.05% 2.41%
Mallows’ Cp 7.70 2.43
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1---- -----Step 2-----
Coef P Coef P
Constant -112.0 -85.7
Temp 5.64 0.000 4.42 0.013
Lighting 0.00791 0.113
S 17.5355 17.5239
R-sq 1.06% 1.27%
R-sq(adj) 0.97% 1.10%
R-sq(pred) 0.73% 0.74%
Mallows’ Cp 2.67 2.16
Stepwise regression with the dataset: office zone C2 – day4 (mid-age)
For each analysis: α to enter = 0.15, α to remove = 0.15
Thermal satisfaction vs bio
Candidate terms: EDA, Skin Temp, HR, Stress Level
-----Step 1-----
Coef P
Constant 4.9522
EDA -6.37 0.001
S 0.458852
R-sq 5.50%
194
R-sq(adj) 4.98%
R-sq(pred) 3.23%
Mallows’ Cp 0.66
EDA vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-------Step 1------ -------Step 2------ -------Step 3------
Coef P Coef P Coef P
Constant -0.0566 -0.0774 0.1521
CO2 0.000091 0.000 0.000108 0.000 0.000069 0.012
Lighting 0.000014 0.037 0.000022 0.003
Temp -0.00833 0.004
S 0.0150260 0.0149674 0.0148366
R-sq 3.45% 4.43% 6.31%
R-sq(adj) 3.23% 3.98% 5.65%
R-sq(pred) 2.68% 3.38% 4.92%
Mallows’ Cp 12.10 9.66 3.08
Skin temp vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- ------Step 2----- -------Step 3------
Coef P Coef P Coef P
Constant 90.85 62.86 58.08
Temp -2.658 0.000 -2.005 0.000 -1.738 0.000
CO2 0.01213 0.000 0.01125 0.000
Lighting -0.002023 0.000
Acoustic
S 0.963407 0.910843 0.889973
R-sq 43.92% 49.98% 52.36%
R-sq(adj) 43.78% 49.75% 52.03%
R-sq(pred) 43.40% 49.31% 51.43%
Mallows’ Cp 79.18 26.29 6.80
-------Step 4------
Coef P
Constant 59.41
Temp -1.830 0.000
CO2 0.01028 0.000
Lighting -0.001958 0.000
Acoustic 0.0335 0.052
S 0.887081
R-sq 52.78%
R-sq(adj) 52.34%
R-sq(pred) 51.73%
Mallows’ Cp 5.00
HR vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1----- -----Step 2-----
Coef P Coef P
Constant 195.2 216.6
CO2 -0.1178 0.000 -0.1103 0.000
Acoustic -0.547 0.049
S 13.5925 13.5406
R-sq 7.14% 8.09%
R-sq(adj) 6.89% 7.60%
R-sq(pred) 5.89% 6.46%
Mallows’ Cp 5.16 3.27
Stress level vs IEQ
Candidate terms: Temp, CO2, Lighting, Acoustic
-----Step 1-----
Coef P
Constant 6.22
Lighting 0.02437 0.000
S 11.5783
195
R-sq 7.16%
R-sq(adj) 6.75%
R-sq(pred) 5.08%
Mallows’ Cp 0.26
α to enter = 0.15, α to remove = 0.15
Abstract (if available)
Abstract
The comfort of occupants in office buildings is an important element of work productivity. A higher comfort level of occupants could lead to lower stress levels. Research on the relationships between an individual occupant’s comfort and indoor environmental quality (IEQ) in a single workspace is mature. However, studies on the relationships between the comfort level of a crowd and IEQ in a multioccupancy condition is being reconsidered. As biometric signals can be processed and represent the stress condition of the human body, it is meaningful to determine the relationships between IEQ and human comfort by integrating biometric signals. Four different survey groups were selected based on the building attributes. Data collection consisted of three parts: biometric signal measurement, which was conducted using wearable sensors
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Exploration for the prediction of thermal comfort & sensation with application of building HVAC automation
PDF
Human–building integration: machine learning–based and occupant eye pupil size–driven lighting control as an applicable visual comfort tool in the office environment
PDF
Nursing comfort study: are hospitals performing for nurses? IEQ analysis of working conditions inside hospitals
PDF
IEQ, sleep quality, and IoT: meta-analysis on improving IEQ and sleep quality using IoT
PDF
Human-building integration: Investigation of human eye pupil sizes as a measure of visual sensation in the workstation environment
PDF
Impacts of building performance on occupants' work productivity: a post occupancy evaluation study
PDF
Developing environmental controls using a data-driven approach for enhancing environmental comfort and energy performance
PDF
Natural ventilation in tall buildings: development of design guidelines based on climate and building height
PDF
Enhancing thermal comfort: data-driven approach to control air temperature based on facial skin temperature
PDF
Multi-occupancy environmental control for smart connected communities
PDF
Enhancing thermal comfort: air temperature control based on human facial skin temperature
PDF
Impacts of indoor environmental quality on occupants environmental comfort: a post occupancy evaluation study
PDF
Indoor environmental quality and comfort: IEQ adaptation and human physiological responses in commercial buildings
PDF
Impact of bamboo materials on indoor air quality
PDF
Quantify human experience: integrating virtual reality, biometric sensors, and machine learning
PDF
Real-time simulation-based feedback on carbon impacts for user-engaged temperature management
PDF
Streamlining sustainable design in building information modeling: BIM-based PV design and analysis tools
PDF
Adaptive façade controls: A methodology based on occupant visual comfort preferences and cluster analysis
PDF
Single & multi-family housing: the complex behind the simple: a kit of parts for minimally invasive interventions
PDF
Development of AI-driven architectural design guidelines: establishing human biometric signal-driven architectural design guideline as a function of psychological principles
Asset Metadata
Creator
Liu, Jie
(author)
Core Title
Human-building integration based on biometric signal analysis: investigation of the relationships between human comfort and IEQ in a multi-occupancy condition
School
School of Architecture
Degree
Master of Building Science
Degree Program
Building Science
Publication Date
06/17/2019
Defense Date
05/06/2019
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
biometric signals,human comfort,IEQ,multioccupancy,OAI-PMH Harvest,statistical analysis
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Choi, Joon-Ho (
committee chair
), Gil, Yolanda (
committee member
), Schiler, Marc (
committee member
)
Creator Email
amberliu0301@gmail.com,liu557@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-174716
Unique identifier
UC11660317
Identifier
etd-LiuJie-7490.pdf (filename),usctheses-c89-174716 (legacy record id)
Legacy Identifier
etd-LiuJie-7490.pdf
Dmrecord
174716
Document Type
Thesis
Format
application/pdf (imt)
Rights
Liu, Jie
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
biometric signals
human comfort
IEQ
multioccupancy
statistical analysis