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
/
Examining presence as an influence on learning engagement
(USC Thesis Other)
Examining presence as an influence on learning engagement
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Running head: PRESENCE AND ENGAGEMENT i
EXAMINING PRESENCE AS AN INFLUENCE ON LEARNING ENGAGEMENT
by
Scott A. Smith
A Dissertation Presented to the
FACULTY OF THE USC ROSSIER SCHOOL OF EDUCATION
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF EDUCATION
August 2014
Copyright 2014 Scott A. Smith
PRESENCE AND ENGAGEMENT ii
Acknowledgments
This research would not be possible without the kind support and efforts of many
people. Grateful acknowledgement is made to those who participated in this study and to
consultants Dr. Laila-Angela Hasan and Dr. Margo Pensavalle, as well as Drs. Linton,
Jenkins, McGregor, Bernstein, Schmaltz, and Foulk. Their time and assistance is greatly
appreciated.
Thank you so much to my chair, Dr. Patricia Tobey, for her limitless energy and
ideas. Her belief in my ability to handle new challenges inspired me to tackle each one. I
am also very grateful for my committee members Dr. Kenneth Yates and Dr. Sean Early.
My committee’s willingness to let me run with an unusual topic and analysis plan, and
their unflagging guidance and support every step of the way were essential. They have
each gone beyond the call of duty in offering their time, advice, and guidance.
While this dissertation represents only a first step in connecting presence to
learning engagement, it represents the last step on a long road toward earning a doctorate,
along which I have had the support of a veritable army, to which I am grateful:
Thank you so very much to Caitlin and Claire. Their support of my studies and
willingness to tolerate long hours of classes, studying, and writing is appreciated beyond
words. They are the antidotes to my most stressful moments and my reasons for pushing
on when things have been tough.
Thank you so much to Mona for setting me on this course, and for her steadfast
support and willing accommodation of the time demands this process has required.
Thank you to my cohort of Ed.D. students — I has been my honor to be in their
company, and I have learned as much from their insights and experiences as from my
PRESENCE AND ENGAGEMENT iii
formal studies. I am especially grateful to the members of my thematic group, Heather,
Rebecca, Maryse, Catalina, and Robert, for their advice and weekly service as a sounding
board. Special thanks as well to Rebecca and the other Friday Night Writes attendees,
Abi, Karolina, Heather D., and Catalina. Your support, encouragement, and camaraderie
made it possible to write huge chunks of this opus, while also remaining sane.
Finally, this project would not be possible without the efforts of my parents and
grandparents. My parents’ support of my education and my every aspiration have always
been a steady source of comfort and inspiration, from my earliest learning to this degree.
I also especially wish to thank my grandparents for their encouragement and financial
support of my studies.
This work is dedicated in memory of my grandfather, who taught my father and
me to ‘work smarter, not harder,’ with the hope that this research may one day help
educators to harness new pedagogical strategies in exactly that spirit.
PRESENCE AND ENGAGEMENT iv
Table of Contents
Acknowledgements ii
List of Tables viii
List of Figures ix
Abstract x
Chapter One: Overview of the Study 11
Statement of the Problem 11
Relevance of the Problem 14
Research Questions 16
Methodology 18
Assumptions 19
Limitations 19
Delimitations 19
Glossary of Terms 20
Classroom Environment 20
Cognitive Load 21
Cognitive Theory of Multimedia Learning 22
Distance Education 22
Information Processing Model 23
Learning Engagement 24
Multimedia 25
Multimedia Instruction/Learning 25
Presence 25
Virtual Learning Environment 27
Chapter Two: Literature Review 28
Introduction 28
Learning Engagement 32
Conceptualizing Learning Engagement 32
Factors that Contribute to Learning Engagement 36
Conduciveness to attention and active processing 36
Emotional and social interaction 37
Collaboration 38
Authenticity of task 38
Management of cognitive load 39
Measuring Learning Engagement 40
Models of Contemporary Learning Environments 42
Classroom Environments 43
Virtual Learning Environments 47
Virtual learning environment design models 47
Learning engagement and virtual learning environments 49
The Sense of Presence 51
Defining Presence 52
Evolution of the Term Presence 58
PRESENCE AND ENGAGEMENT v
Domains of Presence 61
Degrees of Presence 63
Presence from Different Perspectives 64
Flow Theory 66
Cognitive Theory for Multimedia Learning and the
Multimedia Principle 71
Measures of Presence 76
Presence in the Context of Learning Engagement 79
Social presence and co-presence 82
Mental immersion 84
Richness, realism 85
Summary and Conclusion 86
Chapter Three: Methodology 88
Research Design 88
Participants and Setting 90
Instrumentation 92
Variables 93
Learning Engagement 93
Sense of Presence 94
Social presence 95
Physical presence 95
Virtual Environment Design 96
Classroom Design 96
Demographics 97
Procedure 97
Chapter Four: Results 100
Descriptive Statistics & Demographic Variables 101
Analysis of Statistical Consistency 103
Learning Engagement 103
Sense of Presence 106
Analysis of Research Questions 110
Research Question 1: Does the degree of presence sensed
by a student predict the degree of learning engagement? 110
Research Question 1a: Does the degree of presence sensed by
a student predict the degree of learning engagement when
controlling for the type of classroom environment
(on-campus versus distance learning)? 114
Research Question 2: Does the sense of presence for learners
attending distance education and on-campus classes differ
when measured by the Early Sense of Presence Inventory? 115
Research Question 2a: How does the sense of presence for
learners attending distance education and on-campus classes
differ when measured by the Early Sense of Presence
Inventory, accounting for instructor and content? 117
PRESENCE AND ENGAGEMENT vi
Research Question 3: Does on-campus learners’ sense of presence
differ under different classroom configurations when
accounting for instructor and content? 120
Research Question 4: Does distance education learners’ sense
of presence differ under different screen interface
configurations when accounting for instructor and content? 120
Structural Equation Modeling 121
Introduction 121
Preliminary Model 124
Model Revisions 130
Additional Survey Information 134
Overall Reactions 135
Text Chat as a Second Channel 135
Text Chat as Presence Enrichment 136
Text Chat as a Utility 137
Age and Text Chat 137
Text Chat in an On-Campus Setting 138
Summary 139
Chapter Five: Discussion 141
Internal Consistency 141
Research Questions 142
Research Question 1 142
Research Question 1a 143
Research Question 2 143
Research Question 2a 144
Research Question 3 145
Research Question 4 146
Structural Equation Models 147
Follow-up Survey 149
Limitations of the Present Study 151
Instrumentation 151
Sample Size 152
Sample Composition 153
Consistency of Experimental Conditions 153
Model Complexity 154
On-Campus Intervention Fidelity 154
Implications for Practice 155
Recommendations for Future Research 157
Larger Studies 158
Wider SEM Exploration 158
New Directions for VLE Design 159
Technology Enhancement of the On-Campus Classroom 160
Demographic-Related Research 161
Development of New Instruments 162
PRESENCE AND ENGAGEMENT vii
Physiological and Qualitative Research of Presence in the
Classroom and VLE 163
Expanded Venues 164
Summary 166
References 168
Appendices
Appendix A: Original Early Sense of Presence Inventory 196
Appendix B: Modified Early Sense of Presence Inventory
(as distributed) 201
Appendix C: Early Sense of Presence Inventory Subscales 205
PRESENCE AND ENGAGEMENT viii
List of Tables
Table 1: Comparison of salient characteristics of flow and presence 68
Table 2: Comparison of salient characteristics of presence
and learning engagement 81
Table 3: Summary of reported demographic information 102
Table 4: Psychometrics for learning engagement subscales in the current study 104
Table 5: Summary statistics for learning engagement 105
Table 6: Psychometrics for presence subscales in the current study 107
Table 7: Summary statistics for sense of presence 109
Table 8: Selected output of Research Question 1 multiple regression analysis 111
Table 9: Research Question 1 correlations 113
Table 10: Results of independent-samples t-tests for Instructor 1 118
Table 11: Results of independent-samples t-tests for Instructor 2 118
Table 12: Results of independent-samples t-tests for Instructor 3 119
Table 13: Selected AMOS output for Model A: summary notes 131
Table 14: Selected AMOS output for Model A: goodness-of-fit statistics 131
Table 15: Selected AMOS output for Model B: summary notes 133
Table 16: Selected AMOS output for Model B: goodness-of-fit statistics 133
PRESENCE AND ENGAGEMENT ix
List of Figures
Figure 1: Participant flow chart showing overall research design 90
Figure 2: A representation of the appearance of the VLE in use 91
Figure 3: Concept representation in its most basic form 125
Figure 4: Concept representation with disturbance added 125
Figure 5: Preliminary hypothesized model 127
Figure 6: Structural Equation Model A 130
Figure 7: Structural Equation Model B 132
PRESENCE AND ENGAGEMENT x
Abstract
The sense of presence—a feeling of engagement with a mediated or artificial experience
that so closely mirrors the experience of its natural, unmediated analogue that the
perceived difference becomes blurred or lost completely (Lee, 2004)—has been
extensively studied in fields such as computer science and telecommunications, but rarely
in an education context. The present study sought to examine the influence of presence
on learning engagement, using mixed methods to assess both constructs among master’s
degree students in both online and on-campus class settings. A self-report instrument that
combined an established learning engagement measure with an established presence
measure was distributed to participants (n = 148; online n = 102, on campus n = 46).
Multiple regression and structural equation modeling (SEM) indicate that presence is a
significant predictor of learning engagement (with 42% of the variance explained).
Analysis also indicates that the location itself (online vs. on-campus) does not
significantly influence presence. By making a meaningful change in each environment,
however, the present study found that the text chat element in a real-time online learning
environment makes a significant positive contribution to presence. SEM analysis of the
data produced two models explaining the relationship between the sense of presence,
learning engagement, and the significant components of each. Mental immersion, active
interpersonal social presence, and richness emerged as key component factors of
presence. The results support strategic application of presence concepts in online and on-
campus class environments, and suggest further study among different populations, in
additional environments, and in more complex structural contexts, as well as further
exploration of the role of text chat and other online classroom design elements.
PRESENCE AND ENGAGEMENT 11
CHAPTER ONE: OVERVIEW OF THE STUDY
Statement of the Problem
As human beings, we can learn from the unique bits of information we discover in
nearly any environment. The nature of cognition is to take in new information from our
surroundings through the use of our senses, then process that acquired data, integrating it
into our previous understanding to shape new knowledge for immediate or future action.
To teach, however, is to take control of this process by thoughtfully crafting an
experience from which someone can learn specifically desired information. Just as
teachers design the structure of their lessons and make decisions about the content they
present to learners, they also make both conscious and unconscious choices that shape the
learning environment itself, and that environment has its own effects on the depth,
quality, and ultimate success of the learning experience.
In any teaching setting, a sense of shared space is established between the learners
and the instructor, whether that space is formal and rigidly-bounded or informal and
bounded more by sociological constraints. This shared space allows for learners and
instructors to engage with each other, making a communicative and social connection
that allows each to adapt to the other, and allows the instructor to optimize the pedagogy
to align appropriately with the learner’s zone of proximal development (Alexander,
Schallert, & Reynolds, 2009). The shared space and connection also enhance cognitive
processing and the formation of long-term memory by reducing extraneous cognitive
load (Chandler & Sweller, 1991; Mayer & Moreno, 2003; Park, Moreno, Seufert, &
Brünken, 2011; Sweller, Van Merrienboer, & Paas, 1998). Further, the bounds of this
shared space help to establish an affective focus on the learning process (Appleton,
PRESENCE AND ENGAGEMENT 12
Christenson, Kim, & Reschly, 2006; Skinner, Furrer, Marchand, & Kindermann, 2008).
Piaget, Vygotsky, Bandura and other key education theorists assert that attention is
necessary for learning and that active attention is superior to passive attention—a
carefully constructed classroom environment helps to ensure that learners’ attention is as
focused and undivided as possible (Bandura, 1977, 1991; Piaget, 1971; Tudge &
Winterhoff, 2010; Vygotsky, 1926/1997).
As institutions of higher education embark on distance education programs that
include classes conducted through an entirely mediated online environment, the
opportunity—and urgency—arises to explore the fundamental elements of a classroom
environment that make it conducive to efficient and effective learning with increased
specificity. In both the real classroom and the virtual space, elements of the environment
are chosen and designed—whether they be the placement of desks and size of the space
in a physical classroom or the placement of on-screen information in a virtual learning
interface—forming a mediated environment. The wholly artificially-constructed nature
of the online classroom represents a new kind of setting where all of the design choices
are by necessity explicit. Accordingly, awareness of and attention to the details that
optimize the shared sense of space are essential to encouraging engagement and optimal
learning. Knowledge of these environmental elements may help to refine the real
classroom environment, as well, where the same factors may exist, but are more easily
assumed or overlooked. In these physical spaces, more conscious and concerted efforts
to shape the environment might also achieve significant enhancement of learning
engagement.
PRESENCE AND ENGAGEMENT 13
The details that create a space for learners to interact with instructors and fellow
learners are recognized in the literature as establishing presence: a feeling of engagement
with a mediated or artificial experience that so closely mirrors the feeling of engagement
in a natural, unmediated experience that the perceived difference between the two
becomes blurred or lost completely (Lee, 2004). On some level, participants remain
aware of the artificial elements of their experience, but they overlook this knowledge
during the experience itself (International Society for Presence Research, 2000; Lombard
& Ditton, 1997). The effect is personal and internal, and its physiological mechanism is
still not entirely clear (International Society for Presence Research, 2000). The idea of
presence is widely noted in disciplines as far-flung as neuroscience, philosophy, business,
engineering, computer science, and the arts, but is known by different names in different
fields (Lee, 2004; Lombard & Ditton, 1997). Presence is conceptualized in many
different ways, such as suspension of disbelief in media studies, suture in cinematic arts,
or engagement in communications, but its study is largely fragmented by the variety of
interests and nascent shared literature (Lombard & Ditton, 1997; Lombard & Jones,
2007). As a common understanding emerges, unifying different perspectives of presence,
it is becomes clear that in all settings where people interact, a sense of presence is key to
establishing a rich and robust experience (International Society for Presence Research,
2000; Lombard & Ditton, 1997).
When a cinema audience becomes fully immersed in the experience of a film,
audience members can feel similar emotions as they would if they were observing real
people, real environments, and real situations rather than just projected images (even
though outside of the experience, they would easily discern the difference). Many subtle
PRESENCE AND ENGAGEMENT 14
but critical details help to establish this sense of presence, and if any of these elements
fail, the sense of presence—and its emotional and cognitive benefits—is broken
(Lombard & Ditton, 1997). In the cinema, if the audio suddenly falls out of
synchronization with the picture, the effect of presence is quickly lost, and audiences are
immediately ‘pulled out’ of the experience, reminded that they are sitting in the artificial
environment of a theatre rather than participating in the film’s storyline and character
interactions (Biocca, Harms, & Burgoon, 2003; Lombard & Ditton, 1997).
The interaction between instructor and learner is similarly affected by the quality
of presence. As noted by Lombard and Ditton (1997), when a controlled environment
fails to create a strong sense of presence, the effect is jarring, distracting, and a potential
source of extraneous cognitive load—a condition that can undermine the learning
process. When the environment is carefully constructed to maximize presence, the
structure of the environment and its potential cognitive and affective distractions fade
away, allowing learners to fully engage with pedagogy, devoting the maximum amount
of their mental and emotional resources to the learning process.
Relevance of the Problem
The development and expansion of distance education and online course offerings
are a primary interest for American colleges and universities. An estimated 6 million
undergraduate students will engage in some type of online, computer-based distance
learning program offered by United States higher education institutions in 2013-2014—
representing nearly 1 in 4 undergraduates (Radford, 2011). From 2000 to 2008, the
number of students participating in distance education coursework at the postsecondary
level grew from 8 percent to 20 percent (Radford, 2011). Further, the increased
PRESENCE AND ENGAGEMENT 15
availability and affordability of communications technology have catalyzed the growth of
international programs and cross-border distance education (Spring, 2008). In aggregate,
distance education and globalization represent the most active area of innovation in
higher education, and institutions face increasing pressure to respond to their peers,
establishing and nurturing new distance learning programs (Armstrong, 2007).
Distance education employs technology, such as Internet-based computer
environments, to allow students to study or collaborate at geographically remote
institutions. The nature of such entirely computer-mediated environments requires
educators to rethink even the most basic elements of classroom design in adapting them
for the limitations of the computer interface. The pressures of meeting the demand for
distance learning programs, though, have caused many institutions to rush into
programming with less-than-ideal consideration of the most effective means of crafting a
computer-based instructional environment. Many programs attempt a simple translation
of the classroom experience, recording and broadcasting traditional classroom lectures,
often pairing professorial speeches with text-heavy PowerPoint slide shows.
Increasingly, however, institutions are implementing virtual learning environments
(VLEs) that provide real-time interaction between students and their professors, as well
as their peers. Many are also undertaking the hard work of redesigning visual and
interactive elements of these interfaces to bolster the quality of their offerings.
Emerging in parallel, an explosion of interest in computerized ‘apps’ for use on
smartphones and tablet computers such as the iPad has also opened the door to virtual
learning interfaces. Influenced by this new generation of interactive, multimedia-based
personal media devices, colleges and universities are seeking to enhance online distance
PRESENCE AND ENGAGEMENT 16
education programs with the inclusion of interactive applications for mobile devices
(Culén & Gasparini, 2011; Foote, 2010; Lennon & Girard, 2012). While most higher
education institutions have yet to develop academic apps for their programs, many are in
the design stages of apps or enhanced podcasts that integrate with their distance
education offerings.
The concept of presence, while still relatively new in the literature, offers a fresh
conceptualization and aggregation of the effects of learning environment elements on
engagement. While its terminology may seem new to the field of education, its concepts
have long been core to many other fields, and the cross-application of presence concepts
may offer important new ideas for optimizing the learning experience, especially in
highly-mediated environments such as these increasingly important distance learning and
mobile applications within higher education.
Research Questions
The Early Sense of Presence Inventory (ESoPI) (Early, 2008) is a self-report
survey instrument designed to measure the sense of presence felt by subjects. It includes
items designed to test the major recognized categories of presence as cited in the
literature (Lee, 2004; Lombard & Ditton, 1997), including physical presence, social
presence, mental immersion, and social richness.
This study pairs the Early Sense of Presence Inventory with a similarly detailed
inventory of learning engagement, the Student Course Engagement Questionnaire
(SCEQ) (Handelsman, Briggs, Sullivan, & Towler, 2005), to examine both a physical
classroom environment and a virtual learning environment, each in two configurations, to
examine the relationship between the degree of presence sensed and the level of learning
PRESENCE AND ENGAGEMENT 17
engagement and to explore potential factors in each environment that affect the learners’
sense of presence.
The research questions for this exploration are:
1. Does the degree of presence sensed by a student (as measured by the Early Sense
of Presence Inventory) predict the degree of learning engagement (as measured by
the Student Course Engagement Questionnaire)?
1a. Does the degree of presence sensed by a student (as measured by the Early
Sense of Presence Inventory) predict the degree of learning engagement (as
measured by the Student Course Engagement Questionnaire) when
controlling for the type of classroom environment (on-campus versus
distance learning)?
2. Does the sense of presence for learners attending distance education and on-
campus classes differ when measured by the Early Sense of Presence Inventory?
2a. How does the sense of presence for learners attending distance education
and on-campus classes differ when measured by the Early Sense of Presence
Inventory, accounting for instructor and content?
3. Does on-campus learners’ sense of presence differ under different classroom
configurations when accounting for instructor and content?
4. Does distance education learners’ sense of presence differ under different screen
interface configurations when accounting for instructor and content?
The study examined master’s degree students in an education program at a large
private research university in California. Classes were chosen that taught the same
content in different sections, some of which were held on the institution’s campus and
PRESENCE AND ENGAGEMENT 18
others of which were held via a real-time interactive distance learning software solution.
In each teaching environment (online and on campus), specific structural changes were
made, and the students were asked to complete a combined SCEQ-ESoPI instrument both
before and after the changes.
The results from each administration were compared, and two structural equation
models were subsequently developed to better understand the relationship between the
two primary latent factors: sense of presence and learning engagement.
Methodology
The present study employed a mixed-methods approach with special analysis
procedures. The research utilized quantitative methods along with structural equation
modeling (SEM) and qualitative methods to enhance the interpretation of results and
provide a triangulated, multiple-perspective review of latent factors. A self-report
instrument combining the Student Course Engagement Questionnaire (SCEQ)
(Handelsman, et al., 2005) and Early Sense of Presence Inventory (ESoPI) (Early, 2008)
was employed, and is discussed at length in Chapter 3. The analysis, discussed in detail
in Chapter 4, was conducted in multi-step sequence. Demographic information and
descriptives were examined, and the psychometric validity of the instruments was
verified. Then, standard statistical methods, including multiple regression and t-tests,
were employed to answer the research questions posed above. Additionally, a
hypothesized model was constructed, analyzed, and refined using structural equation
modeling techniques. Finally, qualitative data obtained from a free-response follow-up
survey was examined for additional insight and confirmation of other findings.
PRESENCE AND ENGAGEMENT 19
Assumptions
This research assumes that all participants completed each survey item truthfully
and that all responses were tabulated and analyzed accurately. While the validity of the
instruments was verified through analysis of statistical consistency, the research does
assume that these measures accurately reflect the latent constructs they purport to
describe.
Limitations
The researcher acknowledges that this study’s methodology and scope pose
limitations. For example, the generalizability of the findings may be limited to the
demographics of the particular participants sampled. This study examined master’s
degree students at a private research university in California, and its findings should not
be presumed valid with other education levels or population characteristics, such as age,
beyond those described in Chapter 4 as a part of the sample. The self-report nature of the
instrument used in this research also presents limitations, as it requires a level of self-
awareness and understanding that may not be possible from participants. Self-report
instruments are subject to potential misinterpretation by participants, additional bias, and
unintended responses (Salkind, 2012). Additional discussion of these and other
limitations is provided in Chapter 5.
Delimitations
The scope of the current study includes only the traditional, physical classroom
and real-time, synchronous virtual learning environment (VLE). The study does not
examine asynchronous online education, as it is a poorer analogue to in-person pedagogy,
and involves dramatically different expectations for and manifestations of learning
PRESENCE AND ENGAGEMENT 20
engagement and presence. Such environments are, however, briefly described and
compared with the synchronous VLE in Chapter 2. The population chosen in the current
study consisted of master’s degree students, an education level where online learning is
well-established and prevalent, and where participants are likely able to provide accurate
self-assessment results. In the interest of conciseness, literature reviewed in Chapter 2 is
limited to that which is relevant to learning engagement and presence as defined for the
purposes of this study. Literature extending beyond relevant definitions is mentioned, but
full explorations of alternate definitions, or specific subsets of concepts (e.g., learning
engagement as marked by matriculation and retention, or presence as marked by remote
manipulation of objects) is left to the reader. Concepts such as multimedia, flow theory,
and cognitive theory for multimedia learning are addressed only in relation to the key
concepts of learning engagement and presence at issue in this research. As described
above, the scope of methodology in the study included both quantitative and qualitative
analysis, as well as structural equation modeling, to ensure the broadest set of
perspectives on the phenomena of interest.
Glossary of Terms
Classroom Environment
For the purposes of this study, the classroom environment is defined as the
physical space in which instruction is presented. The classroom environment, in this
study, refers only to the traditional, in-person design of a classroom where students and
instructors are all present in the same physical room, with students typically seated at
desks or tables. The term on campus is used throughout this study to aid in
PRESENCE AND ENGAGEMENT 21
differentiation between the classroom environment and the online or virtual learning
environment.
Cognitive Load
For the purposes of this study, cognitive load theory is grounded in the limitations
of working memory described by the Information Processing Model, and includes three
separate types of load that make demands of these finite available resources (Kalyuga,
2011; P. Kirschner et al., 2011b; Mayer, 2009, 2011; Paas et al., 2010). The three types
of load, intrinsic, germane and extraneous or extrinsic load, are of keen interest for
improving instructional effectiveness (Mayer, 2011). While educators would seek to
increase the amount of germane load—that which is necessary for the storage of new
knowledge and schema in long-term memory—they should seek to minimize the
existence of distracting extraneous load—resources devoted to processing information
irrelevant to the desired learning, and manage intrinsic load—that which is inherent in the
complexity and format of the material itself (Kalyuga, 2011; P. Kirschner et al., 2011b;
Mayer, 2009, 2011; Paas et al., 2010). When information is presented in ways that force
the learner to split his or her attention—such as text separated from related explanatory
images—cognitive load is increased, until reaching the point of cognitive overload, the
condition where the available resources of working memory are exhausted and
information can no longer be effectively processed either for current use or long-term
memory (Chandler & Sweller, 1991; Mousavi, Low, & Sweller, 1995; Thompson,
Schellenberg, & Letnic, 2012).
PRESENCE AND ENGAGEMENT 22
In this way, any element of instruction that creates increased extraneous load
should be eliminated or reduced as much as possible, and elements that reduce extraneous
load or assist in managing intrinsic load should be considered for inclusion in pedagogy.
Cognitive Theory of Multimedia Learning
The Cognitive Theory of Multimedia Learning or CTML as outlined by Richard
Mayer (1998, 2005, 2009, 2011) describes multimedia elements and techniques that aid a
learner in the process of acquiring new information and then organizing and storing that
information for later recall. This theory is framed by the Information Processing Model
and its description of the functioning of the human mind (Mayer, 2005; Mayer &
Moreno, 1998; Mayer, Heiser, & Lonn, 2001). At its core, CTML assumes that coupling
images and sound with words is a more educationally valuable proposition than using
words alone (Mayer, 2005). A full discussion of the relationship of CTML and presence
is provided in Chapter 2.
Distance Education
As colleges and universities seek new ways to increase their reach and
enrollments, the popularity of distance education programs continues to grow. In
2007–08, nearly 4.3 million undergraduate students in the United States took at least one
distance education course—representing 20% of all U.S. undergraduates (Radford, 2011).
Four percent of all undergraduates took their entire program through distance education
in the 2007-2008 school year (Radford, 2011).
Radford (2011) defines participation in a distance education class as reporting:
that [the student] took a course for credit during the academic year that was not a
correspondence course but was primarily delivered using live, interactive audio or
videoconferencing, pre-recorded instructional videos, webcasts, CD-ROM or
PRESENCE AND ENGAGEMENT 23
DVD, or computer-based systems delivered over the Internet. (Radford, 2011,
p. 2)
For the purposes of this study, distance education is defined as any educational
instruction provided by an institution of higher education through a computer or
computerized device interface rather than in a physical classroom space.
Information Processing Model
The Information Processing Model (IPM) of learning suggests that the brain
functions much like an orderly machine, processing information obtained through the five
senses, working with that information, and, when appropriate, encoding and storing that
information for later retrieval (Klein, O’Neil, & Baker, 1998; Mayer, 2011; Merrill,
2002). This model holds that learning takes place when information is successfully
processed and stored by this ‘mental machinery’ (Mayer, 2011).
The IPM envisions three distinct segments of the cognitive process: short-term
memory, working memory, and long-term memory (Mayer, 2011). Sensory input,
chiefly from the eyes and ears, flows into short-term memory (also called sensory
memory), and is not stored for any length of time. Items, or bits, of information that are
of interest to the learner move into one of a limited number (usually defined as four to
seven) of ‘slots’ in working memory, where they are held temporarily for processing
(Mayer, 2011). The idea that working memory is constrained in this way is the basis for
cognitive load theory (Kalyuga, 2011; P. Kirschner, Kirschner, & Paas, 2011b; Mayer,
2009, 2011; Paas, Gog, & Sweller, 2010). Information is only held in working memory
for a limited time, and some bits of information are schematized and stored in long term
memory for later recall, while other bits are dropped (Mayer, 2011). Long term memory
is seen as permanent storage, limited only by the learner’s capacity for efficiently
PRESENCE AND ENGAGEMENT 24
recalling bits of information stored there, a process that is facilitated by robust schema
construction (Mayer, 2009, 2011). The Information Processing Model highlights three
key cognitive processes that allow for information to be integrated with prior knowledge
and schematized for efficient storage in long-term memory: organization, attention, and
repetition (Mayer, 2009, 2011).
As envisioned by the Information Processing Model, effective instruction is that
which assists the efficient processing of relevant information from working memory to
long-term memory (Mayer, 2009, 2011). Such instruction focuses learners’ attention,
repeats important bits of information, and organizes information for learners—connecting
it to prior knowledge and other related bits of information (helping to encourage the
creation of robust schema for ease of later recall) (Mayer, 2009, 2011).
Learning Engagement
Learning engagement is a broad concept with a wide range of definitions,
primarily influenced by the scope of the learning addressed. For the purposes of this
study, learning engagement refers to the active commitment of cognitive resources to the
process of acquiring, organizing, and storing new knowledge (Fredricks, Blumenfeld, &
Paris, 2004; Fredricks et al., 2011; Furrer & Skinner, 2003). Learning engagement
requires that students remain attentive and focused on the task of learning, and exhibit a
degree of dedication to the educational process marked by active participation, consistent
attendance, and completion of required tasks. Engaged learners thoughtfully consider
presented information, and respond to it in discussion, writing, and other activities.
A more detailed description of learning engagement, its conceptualization, and its
operationalization in this study is provided in Chapter 2.
PRESENCE AND ENGAGEMENT 25
Multimedia
For the purposes of this study, the term multimedia will be used to describe
presentations and experiences that employ a variety of visual, aural, and interactive
elements to express an idea. While multimedia does not always require electronics, for
the purposes of this study, multimedia will be narrowly defined as technologically-based.
Multimedia Instruction/Learning
Multimedia instruction and multimedia learning will be used interchangeably to
describe the pedagogical strategy of employing a coordinated visual, aural, and
interactive presentation to aid in teaching concepts (Mayer, 2009). Multimedia
instruction assumes careful consideration and an effort to optimize the use of technology
in forming a coherent multimedia presentation.
Presence
Presence, or the sense of presence, is viewed from a number of perspectives in the
literature, each with its own slightly different definition (Lee, 2004; Lombard & Jones,
2007). For the purposes of this study, presence is defined as the perception of
engagement with a mediated or artificial experience that so closely resembles
engagement in a fully natural, unmediated experience that the difference is overlooked.
When presence is strong, individuals interact with the people and environments they are
exposed to as if they were unmediated and authentic, even when they are, in fact,
mediated or wholly artificial (Lee, 2004; Lombard & Ditton, 1997). When presence is
weak, individuals are aware of the mediated or false nature of interactions, and the
distracting awareness of such mediation causes them to withdraw from interacting to at
least a small degree.
PRESENCE AND ENGAGEMENT 26
Lombard and Ditton (1997) offer the example of an audience engrossed in the
experience of watching a motion picture. The experience, with realistic movement,
synchronized and spatially-distributed sound, and realistic characterizations, leads some
audience members to ignore the fact that they are watching flat, projected photographs of
people who are pretending to be other people. When we receive enough cues—sounds
timed to match actions, characters behaving in ways that are familiar to us—we suspend
our disbelief, and begin to process the experience as if we are watching real people
grappling with real events that are playing out in front of us. If a technical malfunction
occurs with the projection, however, and the sound falls out of synchronization with the
images, the presence is greatly diminished, and audience members are “pulled out” of
their engagement, suddenly distracted by the reminder of the artificiality of the
experience (Lombard & Ditton, 1997).
When an experience fails to provide enough sensory and social cues for
individuals to overlook artificiality and mediation from the beginning, individuals may
not engage at all with the experience, or may quickly become distracted (Biocca et al.,
2003; Lombard & Ditton, 1997). In successful, high-presence situations, individuals feel
a sense of ‘truth’ and often must make a prompted effort to recall what elements made the
situation feel realistic to them (Biocca, Burgoon, Harms, & Stoner, 2001; Lombard &
Ditton, 1997).
A more complete definition for the purposes of this study, including
differentiation from the term telepresence, along with a more thorough exploration of the
concept of presence, its history, conceptualizations, and measurement is provided in
Chapter 2.
PRESENCE AND ENGAGEMENT 27
Virtual Learning Environment
For the purposes of this study, the virtual learning environment, or VLE, is
defined as the electronic interface utilized for connecting instructors and students
engaged in presenting and receiving course instruction in a distance education program in
real time. The virtual learning environment is typically a representation of the class
participants on a computer screen, and uses the Internet or other computer network
architecture to provide connectivity between each participant’s computer and the other
participants’ computers. In this study, the virtual learning environment includes both
audiovisual elements and a text-based ‘chat’ system (referred to as simply text chat) for
ancillary communication.
PRESENCE AND ENGAGEMENT 28
CHAPTER TWO: LITERATURE REVIEW
Given the emerging significance of online classroom environments and their
nascent structural paradigm, it is especially important to consider the effect of classroom
design on the quality of the learning experience as a whole. To that end, this chapter
explores the broad preexisting literature related to learning engagement and classroom
design, and considers the ways in which these concepts relate to the canon of research
and theory surrounding the sense of presence, which is also intrinsically related to and
influenced by the design of an environment. A brief description of key instruments for
measuring both learning engagement and presence is also presented.
Introduction
Every learner has limited resources available to leverage in a learning experience,
and the optimization of a learner’s investment of attention and cognition is of critical
concern to pedagogy (Baddeley, 1992; Mayer, 2011; Norman & Bobrow, 1975). At
every age and experience level, the cognitive processing that is fundamental to the act of
learning new information requires students to commit themselves mentally and focus on
the information being presented to them. Students must also be willing participants in the
interpersonal processing of new information through discussion, group activities, and
self-expression through writing or audiovisual presentations. Learning engagement is a
broadly-accepted construct that describes a learner’s commitment of mental and
emotional resources to the learning experience, comprising behavioral, cognitive and
affective components. However, scholars such as Coates (2006) warn that, “despite an
explosion of studies,” we have only scratched the surface of understanding the ways
PRESENCE AND ENGAGEMENT 29
students engage “with activities and conditions that are likely to promote learning and
development” (p. 6).
As Mayer (2005, 2011) notes, attention is a requirement for the active processing
of information, and extraneous influences compete for limited available cognitive
resources. Similarly, Vygotsky emphasizes attention, sensation, and perception as three
of the four elementary mental functions required for learning (Alexander et al., 2009;
Tudge & Winterhoff, 2010; Vygotsky, 1926/1997). “Attention, in fact, serves as a
strategist, i.e., a director and organizer, a guide and commander of the battle, without,
however, participating directly in the combat itself” (Vygotsky, 1926/1997, p. 118).
Bandura (1981, 2001) also emphasizes the important influences of the environment and
others’ behavior around an individual (D. Smith, 2002; Tudge & Winterhoff, 2010).
Both Vygotsky and Bandura also posit the significant importance of high-quality social
interaction in the learning process. Given the significance of the environment and the
behavior of peers, a student’s attention and mental focus are not only important for his or
her own learning, but also for the effectiveness of his or her peers’ learning experience.
For all of these reasons, the classroom environment itself is an essential element
of the learning experience and its design must be thoughtfully considered and carefully
implemented. In the past few decades, however, the concept of the classroom itself has
been broadening, and a new type of classroom interaction has surfaced: the virtual
learning environment. The virtual learning environment is the interface at the core of
online and distance education that connects physically disparate students and teachers in
an electronic learning space that attempts to mimic the most important parts of the
traditional, physical classroom experience. Online environments are also becoming more
PRESENCE AND ENGAGEMENT 30
heavily favored for corporate training and certification programs (Joo, Joung, & Kim,
2013). With such interest and sustained growth, it is all the more important for the
quality of the virtual learning environment to be as carefully considered as the quality of
the traditional classroom.
In an effort to optimize the conditions for learning engagement in both the
traditional classroom and in virtual learning environments, scholars and practitioners
have considered a variety of approaches developed from educational theory and built on
our understanding of learning engagement. The sense of presence, while not yet widely
explored in the context of educational instruction, offers a framework for aggregating
many theories into a single model of mental and emotional engagement with an
environment. Presence is defined as a feeling of engagement with a mediated or artificial
experience that so closely mirrors the feeling of engagement in a natural, unmediated
experience that the perceived difference between the two becomes blurred or lost
completely. It describes a state of heightened mental investment that should be
conducive to learning. Because presence is sensitive to extraneous elements that distract
from commitment to a mediated environment in the same way as to an analogous
unmediated one, the principles of optimizing for presence would necessarily mirror the
elements influencing extraneous cognitive load when attempting to optimize learning.
While scholarship aggregating and integrating theory into a single unified concept
termed presence is relatively recent, the literature that the concept is built on is extensive
and broad, crossing many fields of science humanities, and the arts. From the earliest
theorizing in philosophy and psychology, scholars have sought to accurately describe the
nature of our human perception of reality and existence. While early scholars such as
PRESENCE AND ENGAGEMENT 31
Spinoza focused on the nature of truth and self consciousness, more recent scholarship
has explored how the basic notions of perception and the human understanding of reality
might be applied in the advancement of more convincingly realistic simulations of reality
(Lee, 2004).
As authors such as Lombard and Ditton (1997) and Lee (2004) have sought to
braid the existing broad strands of theory, a few key instruments have also been
developed to measure the sense of presence. While physiological testing offers a real-
time view of subjects’ reactions, the internal, psychological nature of presence makes it
difficult to measure without the additional support of self-report instruments. These
instruments, however, are not without their own challenges and drawbacks.
This chapter will explore the literature in each of these critical areas of
understanding. It will begin by introducing the key scholarship related to learning
engagement and the factors understood to promote learning engagement in the classroom.
Key measurement methods and instruments will also be discussed. The chapter will also
present design models of both traditional classroom environments and virtual learning
environments. Then the chapter will examine the body of literature related to the sense of
presence, beginning with the key theories that define this concept. It will also explore the
theorized components of presence, its relationship to flow theory and cognitive theory of
multimedia learning, and its potential relationship to learning engagement. Finally, this
chapter will consider the challenges of measuring presence, and examine the principal
measurement instruments available.
PRESENCE AND ENGAGEMENT 32
Learning Engagement
Conceptualizing Learning Engagement
Learning engagement is a broad term with meanings that vary considerably from
author to author. Despite more than two decades of active research and scholarly debates
related to engagement in educational contexts, the term is still ill-defined (Appleton,
Christenson, & Furlong, 2008). While the term engagement is usually assumed to
reference the degree to which a learner is invested in the education process, the definition
of how this investment is manifested differs considerably, and many conceive of it as a
metaconstruct, encompassing many dimensions of a student’s interest in and affinity for a
school, class, or the educational process at large (Fredricks et al., 2004; Skinner et al.,
2008). Some authors view learning engagement as an internal, mental allocation of
resources, while others describe engagement as more readily-observable outward,
physical actions taken by learners. Skinner, Furrer, Marchand, and Kindermann (2008)
further caution that indicators of learning engagement are all too easily confused with
facilitators of learning engagement. Indicators are those observable markers which are
themselves only signs of engagement, whereas facilitators are outward activities that
serve as causal agents to bring about or influence engagement (Sinclair, Christenson,
Lehr, & Anderson, 2003; Skinner et al., 2008).
For the purposes of this study, learning engagement refers to student attention to
and involvement in the process of taking in, considering, categorizing, and storing new
information for later retrieval (Fredricks et al., 2004, 2011; Furrer & Skinner, 2003).
Learning engagement is marked by concentration, effort, the use of cognitive and
metacognitive strategies, and persistence (Fredricks et al., 2004; Schunk, 2008; Skinner
PRESENCE AND ENGAGEMENT 33
et al., 2008). While the term has also been used to describe a learner’s ultimate
persistence and satisfaction with his or her educational experience as a whole (Astin,
1984; Fredricks et al., 2011), this study seeks to examine the more immediate investment
of effort in a class setting, and considers learning engagement to be a largely
psychological, internal phenomenon—not directly observable, but with observable
indicators.
Fredricks, Blumenfeld, and Paris (2004) argue that engagement is a potentially
critical feature of school quality, and a significant influence on satisfaction and retention.
The authors characterize learning engagement as a multidimensional construct, viewed in
an overall framework of school engagement which can be broken down into smaller
facets (Fredricks et al., 2004). Engagement is broadly “seen as an antidote to… signs of
student alienation” (Fredricks et al., 2004, p. 60).
Fredricks et al. (2004) describe three distinct categories of engagement:
behavioral, affective, and cognitive. Behavioral engagement is marked by participation
in both in-class and extracurricular activities related to the academic experience
(Fredricks et al., 2004). Emotional or affective engagement pertains to the student’s
reactions to teachers, peers, and the general student environment (Fredricks et al., 2004).
While these two subcategories cover broadly observable elements of the learning process,
cognitive engagement is a more internal quality. Fredricks et al. describe it as the
learner’s investment: “it incorporates thoughtfulness and willingness to exert the effort
necessary to comprehend complex ideas and master difficult skills” (p. 60).
Interaction with the environment is seen as a key influence on learning
engagement, including both physical spaces and the existence of other learners within
PRESENCE AND ENGAGEMENT 34
those spaces (Furrer & Skinner, 2003). Furrer and Skinner (2003) assert that engagement
hinges on relationships among peers and between students and instructors, and describe
engagement as growing out of interaction with the world surrounding a student: “active,
goal-directed, flexible, constructive, persistent, focused interactions with the social and
physical environments” (p. 149). Their research found that the sense of relatedness is
vital to motivation for students in third through sixth grades. Meyer and Turner (2002)
further note the core importance of emotion and social interaction between instructors
and students, and among peers, emphasizing the importance of classroom context in
establishing involvement, affect, and flow. The instructors themselves also represent a
key influence on learning engagement (Skinner & Belmont, 1993). In this way, it is
important to take a broad view of the term environment when considering its effects on
learners. Physical and social characteristics both play important roles in the learning
process and its success.
Interactions are also at the heart of Kolb and Kolb’s (2005) Experiential Learning
Theory, which paints learning as a complex system of physical and emotional
interactions. This model places environment and experience at the center of the
educational experience. Kolb and Kolb emphasize that the space where teaching takes
place is an important determinant of the quality of the experience and learning.
Skinner et al. (2008) offer a useful set of markers for engagement including
concentration, effort, use of cognitive strategies and metacognitive strategies such as
planning, monitoring, and persistence in the initiation and execution of a cognitive task
(p. 766). The behavioral engagement is linked to parallel markers of emotional
PRESENCE AND ENGAGEMENT 35
engagement, such as enthusiasm and passion, that are often favored as results of
successful learning (Skinner et al., 2008).
When students are highly engaged in the learning process, their concentration
increases—they focus more precisely on instruction or the learning exercise at hand
(Skinner et al., 2008). These students are more highly resistant to distraction, and are less
frequently the source of distraction for others.
They are also willing to exert more effort, and are generally more active as a part
of activities and discussions (Skinner et al., 2008). Jimerson, Campos, and Greif (2003),
for example, frame engagement as the full set of outward behaviors directly related to
academic achievement, and contrast engagement with disaffection. The effort and
commitment of engaged learners, coupled with engagement’s inverse relationship with
disaffection, leads scholars to link engagement with motivation and make it a key
consideration in explorations of classroom dynamics (Finn, 1993; Furrer & Skinner,
2003; Skinner & Belmont, 1993; Skinner et al., 2008). Skinner et al. (2008) observe that
the engaged learner persists in the face of obstacles and difficulties, which Bandura
(1977, 1993) links to growth of self-efficacy, suggesting that the benefits of learning
engagement go beyond simply absorbing the lesson the instructor is offering.
Slightly more difficult to gauge, however, are the engaged qualities of cognitive
and metacognitive strategizing. These strategies include planning and monitoring of the
learning process, as well as taking notes, and synthesizing concepts as they are expressed
in the classroom (Fredricks et al., 2004).
Involvement exists as a continuum, and many factors can contribute to where an
individual’s level of engagement falls along that continuum (Astin, 1984; Finn, 1993).
PRESENCE AND ENGAGEMENT 36
Astin (1984) observes that “different students manifest different degrees of involvement
in a given object, and the same student manifests different degrees of involvement in
different objects at different times” (p. 519). This variability is a fundamental property of
learning engagement, and animates the core principle that an educator has the ability to
engender change and improvement in a student’s degree of involvement. Conversely, the
educator or the educational environment may unintentionally contribute the degradation
of a student’s level of involvement (Astin, 1984; Furrer & Skinner, 2003; A. Y. Kolb &
Kolb, 2005).
Factors that Contribute to Learning Engagement
While there is no definitive consensus regarding the bounds and
operationalization of learning engagement, for the purposes of this study, there are
several salient factors that are well-documented influences on learning engagement.
Consideration of these factors is essential, as they represent the principal ‘levers and
valves’ which educators may operate to influence and improve engagement.
Conduciveness to attention and active processing. Scholarship regards active
processing and attention as critical components of learning itself (Mayer, 2005; Sweller
et al., 1998). Sweller (1994) notes that the less experienced a learner is with a task, the
more “close and constant attention” he or she must devote (p. 298). Sweller’s model
builds on prior scholarship such as Schneider and Shiffrin (1977), Shiffrin and Schneider
(1977), and Kotovsky, Hayes, and Simon (1985), which explores information processing
as either controlled, where attention is required, or automatic, where conscious effort is
not required. Automatic processing, however, can only be achieved once a learner is
familiar with given material; whereas new information always requires the effortful
PRESENCE AND ENGAGEMENT 37
attention of controlled processing (Kotovsky et al., 1985; Schneider & Shiffrin, 1977;
Shiffrin & Schneider, 1977; Sweller, 1994; Sweller et al., 1998). If learning
engagement—as effortful attention—is accepted as a proxy for such controlled
processing, it is therefore key to successful acquisition of new information. Hence, the
classroom experience must be tailored to encourage such attention and active processing,
especially when teaching new material.
Emotional and social interaction. In examining the patterns within their
previous bodies of research on student–teacher interactions, Meyer and Turner (2002)
highlight the often under-appreciated importance of emotion as a driver of cognition and
motivation. The design of the classroom must provide ample opportunities for, and
facilitate, emotional and social interaction (D. K. Meyer & Turner, 2002). Free and
unfettered emotional exchange both among peers and between students and instructors is
key to engagement and motivation, and therefore must be supported and encouraged by
the physical and emotional space established by the classroom (D. K. Meyer & Turner,
2002). While the physical space, defined through the boundaries of classroom walls and
furniture, is important, it does not eclipse the more visceral landscape of the emotional
space. The emotional space is established through protocol, expectation, and culture, as
it occurs organically or as it is imposed by the instructor or by agreement of peers (D. K.
Meyer & Turner, 2002; Pintrich, 2003; Schunk, Pintrich, & Meece, 2008). This
established classroom culture forms the backbone of a community that can support more
meaningful social exchanges and higher-level cognition (Garrison & Anderson, 2003).
As a part of these established standards and community norms, the expectations
and obligations of the individual student are together an important contribution to strong
PRESENCE AND ENGAGEMENT 38
learning engagement. Kirschner, Kester, and Corbalan (2011) note that learners must be
assigned an active role in their own learning process to facilitate engagement. Related
motivational theory also reinforces the notion that an individual given personal
responsibility, with a strong sense of personal obligation, will invest effort and become
more closely engaged in the learning process (Bandura, 1991, 1993; Oyserman, Terry, &
Bybee, 2002; Schunk et al., 2008).
Collaboration. The ability to effectively collaborate with peers is also a critical
component of learning engagement (Coates, 2006). Coates (2006) emphasizes the
availability of such collaboration in both campus-based classrooms and online or
‘blended’ environments. Although Garrison and Cleveland-Innes (2005) oversimplify in
conflating basic social interaction with presence, they do provide that “interaction for
cognitive success (i.e., high levels of learning) depends on structure (i.e., design) and
leadership (i.e., facilitation and direction)” (p. 144) that facilitate a deeper, higher-quality
level of presence.
Authenticity of task. This emphasis on high-quality, structured social interaction
is buttressed by Kearsley and Shneiderman’s (1998) framework, which establishes
collaborative participation and authentic task focus as two of the three defining elements
of student engagement. The term authentic task refers to those tasks which reflect true
situations students might encounter outside of the classroom, and call upon students to
engage realistic methods with realistic limitations in completing reality-grounded goals.
The authenticity of tasks set forth in the learning environment is central not only in
Kearsley and Shneiderman’s work, but also in cognitive load research, as observed by
Kirschner, Ayres, and Chandler (Kirschner, Ayres, & Chandler, 2011a).
PRESENCE AND ENGAGEMENT 39
There is considerable benefit to cognitive processing from congruence between
the learning environment and activities with the expected out-of-class conditions and
tasks where such knowledge would be applied (Herrington, Oliver, & Reeves, 2003;
Kirschner et al., 2011a). “As society and work environments become more
interconnected and complex, it is increasingly relevant that cognition and learning
research is carried out in environments that ‘mirror’ the complexity of the real world”
(Kirschner et al., 2011a, p. 99).
Management of cognitive load. Cognitive load imposed by the instructional and
environmental design should be one of the primary considerations in crafting instruction
and classroom environments (Sweller et al., 1998). Sweller, Van Merrienboer, and Pass
(1998) argue that humans can only monitor the contents of working memory, which is
fundamentally and significantly limited. All other cognitive processes are hidden from
consciousness and must be moved into working memory, exacerbating the demands on
this resource. The severe limits of working memory make the educator’s effort to reduce
extraneous processing load not only rewarding, but essential to efficient cognitive
processing and the maximization of a learner’s cognitive potential (Sweller et al., 1998;
Sweller & Chandler, 1994). As Sweller, Van Merrienboer, and Pass note, “the
implications of working memory limitations on instructional design can hardly be
overestimated” (p. 252).
Expertise in any subject matter is defined by the possession of advanced
schema—or mental models—in long-term memory and therefore, assisting learners to
build advanced schema should be a primary goal for educators (Sweller et al., 1998). To
facilitate this schema construction in light of the limits of working memory, appropriate
PRESENCE AND ENGAGEMENT 40
instruction should seek to decrease extraneous load while increasing or maximizing
germane load—that load which contributes to the construction of mental schema (P. A.
Kirschner, 2002; Sweller et al., 1998). Further, because all material includes some
intrinsic load that is impossible to reduce (without reducing the amount of material
expressed), instructional design should focus on the reduction of extraneous or extrinsic
load (Sweller & Chandler, 1994). Sweller (1994) concludes that extraneous load is fully
determined by instructional design; therefore, the manipulation of the instructional
environment and design directly affects extraneous load, which is of essential concern.
Measuring Learning Engagement
As a multidimensional construct, learning engagement presents significant
challenges for the researcher. Additionally, the disagreement about the scope of learning
engagement leads to a bifurcation in the measurement instruments that have been
developed for the concept.
Measures of learning engagement often tend toward analysis of the macro level of
engagement with academic or extracurricular life, retention, or matriculation, especially
when looking at postsecondary populations (Handelsman et al., 2005). These instruments
favor the most outward and obvious indicators of engagement, which are more easily
observed. Measurement of engagement on a more local, in-class level is more difficult to
accomplish, given its internal and temporal expression (Fredricks et al., 2011;
Handelsman et al., 2005). The most common instruments are post-interaction self-
reports, which are problematic for this application, as they naturally require participants
to further deplete the cognitive resources researchers hope to examine as they monitor
their own states and behavior (Fredricks et al., 2011).
PRESENCE AND ENGAGEMENT 41
A variety of instruments do exist, however — often concentrating on particular
aspects of engagement or specific tasks, such as reading or mathematics. Instruments are
commonly aimed at and tested with elementary and secondary school populations, and
are often focused on motivational and affective components of engagement. Measures
that examine the cognitive properties of engagement are disappointingly few and limited.
Many that do examine these properties focus on the construction and use of
metacognitive strategies (e.g. Biggs, 1987), rather than the cognitive process itself.
The Student Engagement Instrument (SEI) (Appleton et al., 2006) is one
instrument that does take cognitive factors into account. Normed against ninth graders,
the measure includes 35 items distributed across six sub scales, three of which
specifically address cognitive engagement, while the remainder focus on affective
engagement. While it has been tested on a range of middle school and high school
students, the questions refer broadly to the overall experience of attending school, rather
than moment-to-moment engagement with the immediate learning experience.
Several widely-applied national studies, such as Indiana University’s National
Survey of Student Engagement (NSSE) (2000) cull data from a broad spectrum of
students, but hew toward Astin’s (1984) more holistic view of the term engagement.
Such national surveys have given rise to smaller instruments, such as the simply-named
Student Engagement Survey (SE) (Ahlfeldt, Mehta, & Sellnow, 2005), taken from the
NSSE, which uses a more restrictive set of 14 Likert-scale items. Four of the items
concentrate on localized classroom engagement, but, on the whole, the survey is focused
on the experience across an entire course, rather than a single learning experience or
environment.
PRESENCE AND ENGAGEMENT 42
One of the most localized measures of engagement is the Experience Sampling
Method (Csikszentmihalyi & Larson, 1987), which employs special pagers that prompt
participants to stop at random times and complete a 45-item form that seeks to gauge
their current activity, location, affect, and level of cognitive engagement. This method
has evolved, and can be found as the prevailing protocol in a number of instruments, with
questions tailored for specific applications. The implementation of the Experience
Sampling Method is challenging, and is also best suited to measuring engagement with an
entire course of material.
One versatile self-report instrument normed with a range from 18 to 56 year olds
is the Student Course Engagement Questionnaire (SCEQ) (Handelsman et al., 2005).
This survey uses 23 items to explore participants’ skills, emotions, participation,
interaction, and performance in a specific course, along with a few more global
characteristics such as orientation toward an incremental theory of learning and goal
orientation. The instrument comprises Likert-scale items in four “factor” categories, with
the prompt, “To what extent do the following behaviors, thoughts, and feelings describe
you, in this course?”
The present study is concerned with the moment-to-moment, in-class definition of
learning engagement, and thus focuses on the instruments with a narrower definition of
learning engagement.
Models of Contemporary Learning Environments
One of the significant factors to be examined in the present study is the degree of
presence in both the online and in-person learning environments. This section reviews
literature available regarding the current models of contemporary learning
PRESENCE AND ENGAGEMENT 43
environments—first, the elements that characterize the on-campus classroom and
secondly, the elements that define virtual learning environments (VLEs).
Classroom Environments
Especially in higher education, classrooms are typically designed to facilitate the
typical ‘sage on the stage’ lecture format, where students sit in rows of individual seats,
desks, or at tables that face a designated ‘front’ of the room or stage, where the instructor
stands to present material. Such an environment, while firmly established, is not
conducive to the collaborative, social learning that is seen as essential for full learning
engagement, nor does it inherently optimize cognitive load.
Mayher and Brause (1986) argue that tradition is not an acceptable defense for the
maintenance of such configurations of the classroom. Classroom design should reflect
the learner’s priorities and be tuned to take advantage of perspective to heighten
humanistic interaction and encourage social engagement and collaboration (Mayher &
Brause, 1986). The structure of the furniture in the classroom and the accepted protocol
for seating constrain the nature of the activities that take place and the length of time they
last (Enyedy, 2003). The placement and design of physical objects and the spaces that
include them provide affordances that serve as cues for observers about their intended
and possible uses (Gibson, 1988). The affordances presented by devices and other
objects in the classroom environment, and the configuration of the learning environment
itself, can encourage or discourage learning engagement intentionally or unintentionally
(Eadie, 2001). So-called ‘open plan’ or other collaborative designs offer encouragement
for interaction by easing visual and aural communication between learners, as well as
facilitating group activities through flexible features such as rollers, open table space, or
PRESENCE AND ENGAGEMENT 44
clustered seating configurations (Caldwell, 1994; Niemeyer, 2003; Shield, Greenland, &
Dockrell, 2010).
Enyedy (2003) notes the dual configuration of both physical space and social
space in a classroom: physical designs affect the ethereal ‘social space.’ Enyedy
emphasizes the importance of ensuring that both spaces weigh on educators’ decisions
when planning classrooms—the physical space can encourage or discourage social
interaction and collaboration, a key enhancement for learning engagement. Smith,
Sheppard, D. W. Johnson, and Johnson (2005) find that cooperative learning provides
distinct advantages for motivation and learning achievement. Evidence also suggests that
thematic content, teaching behaviors, and quality remain unaffected by a classroom that
has been redesigned as a collaborative environment (Beery, Shell, Gillespie, & Werdman,
2013).
Facilities not only play a key role in learning engagement, but also student
achievement (da Graça, Kowaltowski, & Petreche, 2007; Uline & Tschannen-Moran,
2008). A particularly relevant study is Guardino and Fullerton’s (2010), which
discovered that at the elementary level, classroom design that deliberately minimized
extraneous load from computers and peers boosted measured engagement from 3% of the
time to 45% of the time and decreased ‘disruptive behavior.’ Other studies (e.g. Burgess
& Fordyce, 1989; Caldwell, 1994; Eadie, 2001; Weinstein, 1979) bolster this finding that
classroom configuration can affect student behavior and engagement.
Uline and Tschannen-Moran (2008) highlight the need for a basic level of
physical and emotional comfort before learning engagement can freely occur. Espey
(2008) finds that less ‘front-facing,’ constrained classroom configurations have a
PRESENCE AND ENGAGEMENT 45
significantly positive impact on student attitudes regarding the learning process and
collaborative activities. Eadie’s (2001) study of classroom design finds several factors
that allow for optimal collaboration and engagement: (a) having no ‘front’ to the
classroom, (b) wide technology availability, (c) collaborative spaces and (d) a central
screen.
Another relevant avenue in the literature is the discussion of classroom design as
it relates to the tenets of universal design. Borrowed from the architectural term of the
same name, universal design stresses the need for any environment to be constructed in
flexible, accessible ways that ensure all people have appropriate support, regardless of
their individual characteristics and abilities (T. E. Hall, Meyer, & Rose, 2012; Pliner &
Johnson, 2010; Rose & Meyer, 2002). These concepts apply to educational settings
equally as well as other environments where a diverse population can be expected — in
the classroom, universal design facilitates the learning process by reducing barriers for all
learners (Rose & Meyer, 2002). Any class of learners will necessarily have differences in
physical, psychological, and other characteristics, and should strive to thoughtfully
account for a wide range of traits and skill levels without the need for individual
accommodations. These considerations may include (though are certainly not limited to)
ensuring that the classroom environment facilitates visual and verbal interaction among
students as well as between instructor and students (T. E. Hall et al., 2012). Another
universal design concern is the limitation of individual expression in the traditional
format of class: an instructor-conducted lecture or presentation, followed by question-
and-answer period. This format limits student interaction during the lecture, and forces
PRESENCE AND ENGAGEMENT 46
all comments and questions to be held until the end and expressed verbally, with the
entire class as an audience.
It should also be noted that emerging technologies have begun to introduce more
mediated, artificial experiences into the once purely physical environment of the
classroom. As computers and other technological devices become more prevalent in the
physical space, they have a disruptive potential to add extraneous cognitive load and
interfere with personal interaction (Guardino & Fullerton, 2010; Stanton et al., 2001).
Technology can be integrated into the classroom in an immersive way, however, through
strategies that include the interaction of screen-based elements with responsive physical
‘props’ and tools in the real environment (Stanton et al., 2001).
Such tools also offer the potential to increase opportunities for student interaction
and discussion, even concurrently with presentations or lectures. Devices like the iPad
are also mobilizing technology in a way that allows it to contribute more fluidly to the
physical environment, rather than dominating an activity as desktop computers have
(Brand & Kinash, 2010; Culén & Gasparini, 2011; Lennon & Girard, 2012; Meurant,
2010). With a subtler physical presence, technology may be more unobtrusively woven
into the conduct of class, reducing its attention and cognitive load demands. Achieving a
natural integration with the classroom allows the technology to contribute to a sense of
presence, rather than detracting from it.
Technology is also being used to shift the entire purpose of the classroom and
class time, away from a presentation/lecture format and toward ‘flipped instruction,’
where the bulk of course material is consumed by students outside of class time, using
interactive or recorded materials to replace in-class lectures (Prober & Khan, 2013). This
PRESENCE AND ENGAGEMENT 47
opens the classroom time and space for a collaborative approach to learning that builds
on prior knowledge (Prober & Khan, 2013).
Virtual Learning Environments
Virtual learning environment design models. The growth in computer
technology and interconnectivity in the last three decades has given rise to a new kind of
classroom, harnessing the computer and telecommunications to bridge physically
disparate students in a single computer-mediated learning experience. These virtual
learning environments (VLEs) can vary widely, but generally consist of a computer-
based, purposefully designed interface through which instructors and learners can
interact. While much research has examined the immersive, high-technology experience
of virtual reality systems that generally require users to wear specially-designed
eyeglasses, most institutions of higher education currently employ simpler, two-
dimensional interfaces designed for the student’s and instructor’s personal computers for
most coursework in distance education (Tao & Zhang, 2013). Whether presented in two
dimensions or three, VLEs establish a sense of shared space where participants can
collaborate on tasks, be it presenting and attending a lecture or completing a group
project (Billinghurst, Weghorst, & Furness, 1998).
Also popular in higher education and industrial teaching is the asynchronous
online learning environment, where educational content is provided to students through a
computer, but such content is prerecorded, and offers only post-instruction interaction
with instructors, if at all (J. L. Hall, 2007). Such experiences offer little or no real-time
peer-to-peer interaction, as well, and should generally be considered as a wholly separate
PRESENCE AND ENGAGEMENT 48
category of computer-mediated education from virtual learning environments, which
attempt to mirror the live, natural conduct of traditional classrooms (Tao & Zhang, 2013).
Billinghurst, Weghorst, and Furness (1998) also describe the concept of the Open
Shared Workspace, where computer interface elements are integrated into a traditional
classroom environment, requiring less computer mediation of the experience by allowing
participants to use real-world tools in addition to computerized or artificial devices. This
technology-infused classroom is seen by many as becoming the norm for classroom
design in the future (Billinghurst et al., 1998; Liang et al., 2005). If such a ‘hybrid’
model of classroom design does slowly supplant traditional physical classroom designs, it
adds additional weight to the imperative to understand engagement in mediated
environments.
The most immersive virtual learning environment is the virtual reality experience,
where participants wear special glasses that are capable of filling the entire field of view
with a projection of a computer-mediated artificial environment (Billinghurst et al.,
1998). Augmented reality experiences utilize computer-connected glasses or devices
such as a smart phone with built-in cameras to allow participants to continue to see the
real world through the glasses or device, while adding projected virtual objects to the real
surroundings.
In an exploratory study (n = 18), subjects were able to collaborate on simple
space-based tasks more successfully in an augmented reality condition than in a
completely immersive condition, though the results were only significant for subjects
who used body language to express themselves (Billinghurst et al., 1998). Both
approaches, however, require specialized devices, which concern some researchers as
PRESENCE AND ENGAGEMENT 49
emphasizing the individual learner at the expense of group collaboration (Yang & Lin,
2010). Yang and Lin (2010) argue that a shared screen, with the ability for multiple
learners to interact concurrently and in a face-to-face manner, provides a collaborative
space that produces enhanced learning outcomes and more positive learner affect.
Bridging both the virtual reality concept and the shared screen design, virtual
online environments such as Second Life provide a ‘virtual world’ accessed through a
computer interface. Interactive games create high levels of social presence, including
avatar recognition, in-game and real-life meetings, and cross-cultural collaboration
(Harteveld, Thij, & Copier, 2011). These virtual worlds offer both immersive spaces and
the potential for educationally-valuable collaboration that can support more complex
tasks than simpler face-to-face conversation-based VLEs (Warburton, 2009).
Learning engagement and virtual learning environments. The virtual learning
environment presents challenges for learning engagement, since collaborative and active
learning techniques can be hampered by the heavily-mediated nature of the environment.
Research has shown, however, that when adapted for the online environment, these
approaches can yield comparable levels of engagement and interaction.
Herrington, Oliver, and Reeves (2003) posit that ‘authentic activities’—those
which mimic real-world tasks and scenarios—show a number of benefits for learners in
VLEs. Authenticity is an important enabler, as students must suspend their disbelief
before engaging in tasks and willingly immerse themselves in the simulated situation
(Herrington et al., 2003). Herrington et al. include as key characteristics of appropriately
immersive and authentic tasks the opportunity to collaborate, the opportunity to reflect,
ill-defined (realistic) boundaries, and the possibility of multiple outcomes.
PRESENCE AND ENGAGEMENT 50
In a qualitative study, Stacey (2007) examined master’s of business
administration students engaged in a distance learning program, and found that a sense of
shared space was essential to establishing group collaboration. Stacey observed a social
construction of knowledge taking place, facilitated by the shared space established in the
online software environment. “Group discussion affirmed or negated their construction
of concepts, providing a means of extending ideas beyond the level they can manage
individually” (Stacey, 2007, p. 21). Group collaboration with minimal instructional
intervention is seen as the ideal mode of learning in online settings by Oliveira, Tinoca,
and Pereira (2011). Swan (2001), on the other hand, finds that “interaction with
instructors seemed to have a much larger effect on satisfaction and perceived learning
than interaction with peers” (pp. 322-333). Perception of personal interaction between
student and instructor is an often overlooked, yet evidently important, factor in the
success of online learning (Swan, 2001). Consideration of the quality of the shared space
provided and the personalization of interactions in both collaborative tasks and teacher-
student interactions should thus form the foundation for successful VLE engagement.
Kreijns, Kirschner, and Jochems (2003) affirm that the quality of the space
provided for social interaction has a strong influence on the success of communication
and collaboration in a computer-mediated environment. Complicating the issue, though,
is evidence that the computer mediation of the space itself may adversely affect some
learners. Ghani and Deshpande (1994) find that personal attitudes toward technology and
computer use can have a significant effect on the willingness of a participant to commit
mental resources and effort to engaging in a computer-mediated task.
PRESENCE AND ENGAGEMENT 51
Warburton (2009) asserts that the highest engagement comes in the most
immersive online environments. Through three-dimensional visualization and context,
cultural and social cues, as well as the ability to easily engage in collaborative tasks in an
analogous way to those tasks in the real environment, Second Life and other multi-user
‘virtual worlds’ offer a more engaging opportunity than two-dimensional spaces
(Warburton, 2009). Blessinger and Wankel (2012) and Coates (2006) also argue that
maximizing the potential for realistic interaction in online education is essential to
encouraging the same active learning and development as exists in campus-based
programming, whether that interaction is through two-dimensional or three-dimensional
interfaces.
The mediated nature of the virtual learning environment itself, and its potential
distraction, is a cause of cognitive load concern, however. Chandler and Sweller (1996)
find that for tasks with high-element interactivity (requiring learners to integrate steps
before information can be considered for further processing), the use of a computer as an
instructional medium added enough cognitive load to interfere with learning (as
compared to a self-contained manual with no computer component). In low-element
interactivity conditions, the medium of delivery (computer-based or print-based) had no
effect, though (Chandler & Sweller, 1996).
The Sense of Presence
The sense of presence is a potentially useful framework for considering the level
of engagement between a learner and her environment. It has been codified chiefly in
light of advances in interactive technologies and computer simulation, but its definition is
not confined to that domain. While it has previously been applied to educational settings
PRESENCE AND ENGAGEMENT 52
in only a limited way, it is a broadly-defined construct that incorporates well-understood
ideas of mental immersion and engagement from a variety of fields.
Defining Presence
While the term presence may not yet enjoy wide recognition in the field of
educational psychology, the concepts that underlie presence are rooted in the very earliest
examinations of philosophy and the functioning of the human mind. The term itself is a
more recent development in a long and ongoing academic discussion of consciousness
and interaction in constructed environments, and aggregates research and philosophy
since the mid-1970s. It has been of particular interest in the field of computer-mediated
communications and virtual reality design, but its implications are of critical importance
in any field concerned with the mediated interaction of individuals and designed
immersive experiences, including entertainment, journalism, education, and healthcare
(Lee, 2004; Lombard & Ditton, 1997). The increasing capabilities and availability of
interaction through technological devices have prompted considerable exploration and
refinement of the term, though it remains a construct easily applied to environments
without artificial interfaces.
In its simplest statement, presence is the phenomenon whereby a virtual
experience or interaction is so involving and natural that its constructed nature or
artificial mediation fades from consciousness (Biocca et al., 2001; Lee, 2004; Lombard &
Ditton, 1997). Presence can refer to a situation in which the entire interaction is
facilitated by an artificial device, such as a computer or television, but it can be equally
applied in more subtly constructed and controlled situations, such as a lecture hall or
classroom. The medium also need not be electronic — books or theatrical performances
PRESENCE AND ENGAGEMENT 53
are also influencers of presence (Gerrig, 1993; Lee, 2004). The degree of presence
corresponds with the thoroughness of the subject’s loss of consciousness of the illusory
components of the experience, but does not require the subject to be entirely unaware of
the mediation or construction, and it may increase or decrease over the course of the
experience (Lee, 2004; Lombard & Ditton, 1997).
The concept of presence is rooted in the earliest explorations of consciousness and
the mind’s ability to distinguish reality from illusion among the physical signals that the
brain receives from the senses (Lee, 2004). The earliest philosophical examples, which
heavily influenced our understanding of the concept that would eventually be known as
presence, are the ideas of truth and belief considered by René Descartes and subsequently
by Baruch Spinoza (Gilbert, 1991).
Descartes first described the mechanism behind the mind’s interpretation of
reality in the 1640s. His conception holds belief and understanding separately, and
requires active choice to believe or not to believe be made every time a new idea enters
our consciousness (Gilbert, 1991). Whether it is a scene observed through vision of
physical objects or a sound received through the ears, each idea passively impresses itself
upon the mind, at which time we make a choice to accept or reject that idea or sensation
as true. In this way, every observation and idea is determined to either be real or illusion
and then acted upon accordingly.
Baruch Spinoza reexamines this idea in his Ethics (1677/1992). Spinoza’s
reconceptualization hinges on an assumption that truth itself is required to reveal
falsehood: “Indeed, just as light makes manifest both itself and darkness, so truth is the
standard both of itself and falsity” (Spinoza, 1677/1992, p. 92). He describes a new order
PRESENCE AND ENGAGEMENT 54
of operations by which one determines what to believe. In Spinoza’s model, all ideas that
arrive in the mind from the senses are assumed to be real (Mason, 2004). The mental
representation of abstract concepts or ideas is similar to the mental representation of
physical objects; therefore, belief in an abstract concept makes it nearly as ‘real’ as a
physical object (Gilbert, 1991). Then, after this initial assumption, the mind considers
evidence and may subsequently reclassify information as illusion.
Blake, Nunez, and Labuschagne’s (2007) study supports Spinoza’s model (which
they term the “Spin model”) when testing against a model where subjects are required to
actively suspend their disbelief. This model is inherently appealing because of the
efficiency that such a sequence implies, which corresponds well with assumptions of the
speed in making such decisions that would be required for survival and function (Gilbert,
1991). Spinoza’s belief-first order of cognition not only accelerates the interpretation of
the environment around a person, but also establishes a foundation that provides for
illusory perceptions, whereby simulated or artificial constructs could be seen, heard, and
felt as if they were real — at least for a time (Gilbert, 1991; Mason, 2004). This notion,
that one could be engaged in an artificial interaction, but perceive it to be real, is the core
concept that enables the idea of presence (Lee, 2004).
Lee (2004) provides one of the most comprehensive overviews of presence and its
foundations. Unlike other conceptions, such as Nakatsu, Rauterberg, and Vorderer’s
(2005), which frame presence as a result of whether an activity is mental, physical, or
mixed, Lee’s model is concerned with the psychological and cognitive interpretation of
the physical senses. He describes human experience of the world outside of one’s own
body to be categorized in one of three distinct realms: real experience, virtual experience,
PRESENCE AND ENGAGEMENT 55
and hallucination. A single person can and will have perceptions and experiences of any
and all of these types, and could experience more than one at a time (Lee, 2004).
Real experiences are sensory perceptions of actual people, objects, or
environments (Lee, 2004). These elements exist as perceived, and the experience of them
could therefore be considered truth. Touching a table, feeling a breeze, or talking with a
person seated beside oneself would all be considered real experiences. Experiences of
actual objects, people, and environments are assumed to be real (Gilbert, 1991; Lombard
& Ditton, 1997).
Hallucination occupies the other end of the experience spectrum, and describes
“nonsensory experience of imaginary objects,” (Lee, 2004, p. 37) as well as imaginary
people, or environments. In a hallucination, the stimuli prompting the experience do not
actually exist outside of the person perceiving them, and are not received through the
body’s actual sensory organs, even though the individual may perceive them as having
been relayed through the usual sensory channels. A person feeling that a ghost may have
tapped one on the shoulder in a haunted house attraction, a fearful person hearing noises
that do not actually exist, or a child interacting with an imaginary friend would all be
considered hallucinations. Hallucinations are often taken to be a sign of mental distress
or dysfunction. This category of experience also includes the kind of “absorption” that
Tellegen and Atkinson (1974) attributes to hypnosis and other artificially-imposed mental
experiences prompted by another person’s verbal direction rather than an individual’s
sensory reception.
Occupying the gap between real experience and hallucination is virtual
experience. Lee (2004) describes virtual experience as the “sensory or nonsensory
PRESENCE AND ENGAGEMENT 56
experience of para-authentic or artificial objects” (p. 38) (though it applies to people and
environments, as well). Virtual experience is distinct from real experience and
hallucination, though it is occasionally confused with either (Lee, 2004). The sense of
presence is concerned entirely with this type of experience, and thus a fuller exploration
of its definition is illuminating.
The first qualifier, “sensory or nonsensory” describes virtual experiences as either
received through the body’s physiological senses or as a product of the individual’s
imagination (Lee, 2004). In this way, virtual experiences may be prompted by factors in
the environment, or may be wholly a product of mental invention.
Distinguishing these perceptions from hallucination and from real experience,
however, is that in a virtual experience, the people, environment, or objects perceived are
“para-authentic or artificial” (Lee, 2004). This distinction allows for virtual experiences
of real stimuli that are simply not in the same space, such as interaction with a person
who is real, but not actually present in the same room, or the imagination and re-
experience of real stimuli from memory. Lee (2004) defines para-authentic to mean
exactly this kind of displaced, yet real, person, object, or environment. Artificial, as in
other contexts, refers to those objects, people, and environments which are manufactured
through human effort and/or technology, but designed to substitute convincingly for real
analogues (Lee, 2004). Lee cautions that artificiality can shift depending on the nature of
the interaction examined:
That is, an object can be actual in one domain, yet artificial in another domain.
For example, it is obvious that computers are actual objects in the physical
domain. In the social domain, however, they become artificial when people start
responding to them as if they were actual social actors (i.e., humans) (Lee, 2004,
p. 35).
PRESENCE AND ENGAGEMENT 57
Assuming the ideas of a virtual experience to be valid, presence then describes the
degree to which a virtual experience is convincing and indistinguishable from a real one
(Lee, 2004). This quality of verisimilitude relies on the accurate mimicry of natural
sensory cues, and it is the number and fidelity of these sensory cues that determine the
strength of the subject’s sense of presence (D. R. Anderson, Lorch, Field, & Sanders,
1981; Biocca, 1997; Lombard & Ditton, 1997). “The senses are the portals to the mind”
(Biocca, 1997, section 1.1 para. 2). When an individual is more present, he or she
perceives a virtual experience to be more like a real experience, and behaves as if it were
real. The most effective systems for creating a sense of presence are those that mediate
the virtual environment most similarly to the way the human body mediates the real
world (Biocca, 1997). Those experiencing presence make conscious and subconscious
judgments as to the authenticity and similarity of the experience to previous interactions
(Biocca et al., 2003). Further, presence can exist in both passive experiences and active
ones (Nakatsu et al., 2005).
The study of presence illuminates the ways in which artificially-constructed
environments and situations can be made to elicit the same cognitive responses as the
real, natural ones they mimic (Lee, 2004; Lombard & Ditton, 1997).
A critically important corollary follows: if presence represents a blurring of the
boundary between the imaginary and real, the failure of presence creates a distracting
refocusing of the subject’s attention on the mediation of the experience itself (Lombard &
Ditton, 1997). Failure of presence can be a jarring experience and introduce large
amounts of extraneous cognitive load for the individual (Lombard & Ditton, 1997).
When the film breaks in a cinema projector, the audience is violently pulled out of the
PRESENCE AND ENGAGEMENT 58
film’s story and characters, and forced to confront the reality of sitting in a room of other
people, watching projected photographs in rapid succession; similarly, a videoconference
interrupted by spinning cursor and a “buffering…” message highlights the artificial
nature of the situation, and interrupts the social flow and cognitive exchange of natural
conversation (Biocca et al., 2003; Lombard & Ditton, 1997). In this way, it may be even
more important to avoid the failure of presence than to cultivate ameliorative effects
through properly nurturing presence (Lombard & Ditton, 1997).
While most, if not all, instructional experiences could be considered artificially
constructed, or mediated by artifice, the utility and increasing popularity of technology-
mediated experiences, including distance education and online communication, make
presence an increasingly important effect to understand and relate to learning goals
(Lombard & Ditton, 1997).
Evolution of the Term Presence
The definition of presence for the purposes of this research closely follows the
well-described Lee (2004) model, which aggregates and subsumes a number of previous
descriptions and definitions that have evolved over the past four decades to describe these
phenomena, largely within the fields of organizational communications,
telecommunications, computer interface design, electronic simulation, and video game
design (T. Kim & Biocca, 1997; Lee, 2004; Lombard & Ditton, 1997).
The term telepresence is often used in the literature as a synonym for presence,
but frequently has a narrower definition that does not include the concepts Lee (2004)
describes as self presence or artificial social presence (International Society for Presence
Research, 2000; Lombard & Ditton, 1997; Lombard & Jones, 2007). For the purposes of
PRESENCE AND ENGAGEMENT 59
this study, the term presence is used and it references the broadest definition of the term,
in line with Lee’s description, with telepresence left to a more specific experience
(mediated para-authentic social and physical presence).
Among the earliest considerations of presence was Short, Williams, and Christie’s
(1976) research on subjective or objective social richness. Short et al.’s term social
richness describes an experience which, either subjectively or objectively, engages with
individuals and makes them feel the warmth and intimacy of a social interaction. The
richness of that experience describes the quality of the experience’s realism and relative
convincingness—“the salience of the other people in the interaction” (Short et al., 1976,
p. 65). Social presence, social richness, and media richness are all similar concepts, and
their study has typically been designed to improve organizational communications
efficiency (Rice, 1992). Rice (1992) advises that information on social or media richness
be used by managers “to appreciate appropriate uses of and opportunities for different
communication media in organizational contexts” (p. 475).
Examining the term telepresence, Kim and Biocca (1997) use realism similarly to
describe the quality of a mediated social interaction, adding further dimension to the idea
of social richness. Social realism has been used to reference portrayals that do or
plausibly could occur in life (Lee, 2004). Perceptual realism refers to sensory-engaging
experiences that give observers believably similar sensory input to those they would
experience in a real event (Lee, 2004). Draper, Kaber, and Usher (1998) subdivide
telepresence domains, and approach the definition of presence with their term
experiential telepresence.
PRESENCE AND ENGAGEMENT 60
In Kim and Biocca (1997), perceptual or social realism is closely tied to the
nature of feeling transported to another environment, and losing consciousness of the
medium that serves as the interface between people that are interacting. This realism is
powerful enough to influence memory as well as changes in attitude that can affect later
decisions, such as purchase intention (T. Kim & Biocca, 1997). This powerful sense of
“transportation” to another environment is a key trait of presence, originally described in
Minsky (1980) as telepresence, and the focus of substantial research in its own right
(Gerrig, 1993; Minsky, 1980; B. Reeves & Nass, 1996; Sheridan, 1992, 1994). Gerrig
(1993) and Reeves (1996), like Kim and Biocca (1997), focus on the moments of
“departure” and of “arrival,” where individuals leave their consciousness of the real
environment and arrive within the virtual one. Frequently, this sense of transportation or
telepresence is held as fully separate from social engagement, sometimes called social
presence or emotional presence (Biocca et al., 2003).
The concepts described as perceptual or psychological immersion by Biocca and
Delaney (1995) and Palmer (1995) are also frequently described as aspects of presence.
Much like the notion of being transported to an entirely different environment, perceptual
or psychological immersion refers to the complete occupation of an individual’s
perception and thought by a virtual experience. Psychological immersion describes the
individuals preoccupation with elements of the virtual experience, whereas perceptual
immersion marks the degree to which an individual’s senses are filled by the cues of the
virtual experience (Biocca & Ben Delaney, 1995; Palmer, 1995).
PRESENCE AND ENGAGEMENT 61
Presence also encompasses previous research on social interaction with a person
or entity within a medium (Lemish, 1982; Lombard, 1995; Picciano, 2002) and social
interaction with a medium itself (Nass & Moon, 2000; B. Reeves & Nass, 1996).
Domains of Presence
While many scholars (Biocca et al., 2001, 2003; IJsselsteijn, de Ridder, Freeman,
& Avons, 2000; Lombard & Ditton, 1997; Moreno & Mayer, 2000; Nakatsu et al., 2005;
Riva & Waterworth, 2003) have offered attempts to systematically map and describe a
number of different specific branches or domains that comprise the broader sense of
presence, Lee (2004) offers one of the most neatly defined such maps, describing three
distinct, simplified domains of virtual experience, each of which can function with either
para-authentic or artificial stimuli. Lee’s domains are physical, social, and self, and they
correspond well to the classifications offered by other scholars, simplifying
categorization, while still offering a richness of description.
Lee (2004)’s physical domain includes the perception of real people, objects, and
environments, but in a mediated fashion (para-authentic) or when the experiences are
fully constructed (artificial) (Lee, 2004). Para-authentic physical presence includes
observation of or interaction with people and events that are real, yet displaced from the
observer’s actual location. These include watching a live television broadcast of a
sporting event, or moving a lunar rover via a remote connection. Artificial physical
presence describes experiences with objects, people, or environments that do not actually
exist, but are believable analogues. Examples include watching a science fiction film or
television show or even reading a detailed fiction account in a novel (Lee, 2004).
PRESENCE AND ENGAGEMENT 62
The social domain defined by Lee (2004) includes the para-authentic interaction
with real people or the real interaction with artificial actors. In the para-authentic social
experience, an individual interacts with another person through a mediated construct,
such as video chat (software and hardware videoconferencing including FaceTime, Skype
and similar technologies), text messaging, or even shouting answers from one’s living
room sofa to a hapless contestant on a televised game show (Lee, 2004; Lombard, 1995;
Lombard & Ditton, 2004; B. Reeves & Nass, 1996). Artificial social presence refers to
real interaction with artificial social actors, such as speaking to a robot or a computer
agent like Apple’s Siri personal assistant, or even verbally coaxing a malfunctioning
automatic teller machine as if it were a stubborn child (Lee, 2004; Lombard & Ditton,
2004). In these cases, even when an individual rationally understands that the object he
or she is interacting with is not a person, the object manifests enough human social
qualities that the individual disregards the artificiality and uses the same protocol that
they would with another human.
The final domain described by Lee (2004), self presence, encompasses the feeling
that an individual himself or herself is located in another location or that an artificial
actor is a manifestation of himself or herself. The para-authentic experience of self
presence includes the remote manipulation of an anthropomorphic surgical robot,
designed to function in real time, just the same as the individual operating it (Lee, 2004).
The artificial experience of self presence includes the close identification with a fictional
character in a drama, such that the individual feels that character is their surrogate, or the
use of an cartoon animal avatar to represent oneself in an online community or
multiplayer game (Lee, 2004).
PRESENCE AND ENGAGEMENT 63
Degrees of Presence
Presence is not a binary property, but exists on a continuum of strength, or degree
of presence experienced by the subject (Lee, 2004). Lombard and Ditton (1997) describe
the sense of presence as dependent on the nature and difficulty of the task, and whether
control belongs to the subject or is automatic (Sheridan, 1992, 1994).
Prior experience or familiarity with a medium can lead to comfort and increase
the sense of presence, although habituation does not necessarily contribute to increased
presence (Lombard & Ditton, 1997). Knowledge of the medium being used in an
experience and how it functions (i.e., how it is engineered) typically decreases the sense
of presence (Lombard & Ditton, 1997).
Stronger presence yields a more immersive, individual experience, therefore high-
presence media (using visual and/or multi-sensory cues) are generally judged more
appropriate for personal tasks (Lombard & Ditton, 1997; Rice, 1992). Conversely,
simple information-transfer tasks generally require less presence (and, in fact, can
sometimes benefit from situational awareness) and therefore call more for low-presence
media (such as audio-only or text-only experiences) (Gilkey, Simpson, & Weisenberger,
2001; Lombard & Ditton, 1997). The highly social nature of presence (in all domains)
would seem to suggest that personal and interpersonal tasks generally favor strong
presence. “Consistent with this idea, the intensity and valence of emotion that occurs
during mediated activities (e.g., as a result of conflict, strong empathy, sexuality, etc.)
seem likely to affect presence” (Lombard & Ditton, 1997, “Nature of task or activity,”
para. 1). A sense of shared space bolsters the authenticity and quality of social
PRESENCE AND ENGAGEMENT 64
interaction, reduces distraction, increases focus, and makes experiences more memorable
(Lombard & Ditton, 1997).
Tasks that require both complex information transfer and social collaboration,
such as classroom learning, should likely require a higher degree of presence. Further,
the use of verbal and non-verbal social cues and gestures (a potential component of multi-
channel instruction) is more effective when there exists a greater degree of presence
(Lombard & Ditton, 1997). For example, results in Moreno and Mayer (2005) indicate
that concurrent guidance (high social presence) in an interactive environment is more
effective for learning than self-reflection exercises.
Presence from Different Perspectives
Presence is at the heart of the human experience and consciousness, and yet our
understanding of it is fragmented and handicapped by a shared literature and terminology
that has only just begun to coalesce. Presence describes a psychological phenomenon
that is universal across many domains, including film theory, business, communications,
computer science, engineering, and fine arts, as well as psychology and sociology,
leading to its study under different terms in different fields (Lombard & Ditton, 1997).
For example, in the critical study of cinematic arts, the phenomenon described as
suture by many scholars (Butte, 2008; Carroll, 1993; Heath, 1981; K.-M. Lin, 2007;
Lurie, 2001; e.g. Oudart, 1978) corresponds with Lee (2004)’s artificial social presence.
Suture describes a cinema audience’s feeling of attachment with a film’s actions and
characters that allows audience members to forget that they are watching projected
images of fictional characters in a darkened auditorium, and instead to become
emotionally and viscerally involved in the outcome of the narrative (Butte, 2008; Oudart,
PRESENCE AND ENGAGEMENT 65
1978). In some cases, audience members may relate so much to a particular character
that the effect is the same as artificial self presence, where one projects oneself onto a
representation that is fictional and not directly faithful to one’s own real traits.
Suture and presence also mirror the definition of transportation or absorption in
other media studies, such as creative writing and narrative studies (Green, Brock, &
Kaufman, 2004). Vorderer, Klimmt, and Ritterfeld (2004) link enjoyment of media
entertainment to similar motivational and engagement values as presence.
Other terms, such as breaking the fourth wall in theatre (Caird, 2013), immersion
and engagement in computer software design (Nakatsu et al., 2005), and telepresence in
business and communications (Rice, 1992) similarly refer to functionally equivalent
phenomena. Additionally, “a vast majority of behavioral science within virtual
environments is predominantly concerned with the construct of presence” (Yee,
Bailenson, Urbanek, Chang, & Merget, 2007, p. 110).
Lombard and Ditton (1997) point out that the concept at the heart of presence—
that an artificial experience be made so convincingly real that a subject treats it the same
as the experience it mimics—makes it one of the most fundamental components of all
experimental design in psychology, sociology, and education, where hypotheses must
regularly be tested in simulated environments for convenience, control and subject
protection. In this way, they argue, the existence of the sense of presence underpins of
most psychology, sociology, and education theory.
Yee, Bailenson, Urbanek, Chang, and Merget (2007) are among those who
specifically leverage this effect, extrapolating the use of online communities to
understand social behavior in the real world, based on an assumption of presence:
PRESENCE AND ENGAGEMENT 66
…if it is the case that behavior online is largely similar to physical behavior, then
it becomes possible to use these online worlds to test behavioral science theories
that are predominantly concerned with physical behavior, both at the micro level
and at the macro level (Yee et al., 2007, p. 116).
Lombard and Ditton (1997) caution that the assumption of presence in research
design, while fundamental, is not necessarily fully proven. Gale, Golledge, Pellegrino,
and Doherty (1990) show that when children are asked to learn navigational information
by watching routes through a neighborhood via a videotape, they were less effective after
five viewings than children who were given one experience in the real neighborhood.
While this illustrates the fallibility of virtual experience, it does not, however, invalidate
the value of presence. Gale et al. could actually be considered as evidence that
understanding the appropriate degree of presence and how to evoke it is critical to the
successful and useful leveraging of mediation technologies.
Flow Theory
Within the realm of educational psychology, the closest concept to presence used
with regularity is Csikszentmihalyi’s (1975, 1990) flow theory. As Novak, Hoffman, and
Yung (1998) caution, there is considerable disagreement on the precise definition of flow.
It is generally accepted, however, that this theory describes a psychological state in which
an individual becomes engrossed in an activity to such a significant extent that he or she
loses perception of the environment around him or her (Csikszentmihalyi, 1990;
Nakamura & Csikszentmihalyi, 2002). The concept of flow describes a mental and
behavioral commitment to a single activity to the exclusion of others, creating a similar
transporting effect to physical and self presence concepts (Csikszentmihalyi, 1975, 1990;
D. J. Shernoff, Csikszentmihalyi, Shneider, & Shernoff, 2003). Flow integrates intrinsic
motivation, personality, and subjective experience. Often, this phenomenon is seen as
PRESENCE AND ENGAGEMENT 67
having ameliorative effects, as it necessarily results in significantly reduced extraneous
cognitive load, and it is often viewed as an explanation for enjoyment of an activity
(Csikszentmihalyi, 1990). The central ideas of mental, emotional, and behavioral
commitment are often linked to the concept of learning engagement (D. J. Shernoff et al.,
2003).
While flow and presence have many similarities, there are important traits that
distinguish the two, and these differences are important to understand when considering
the possible effects of presence on the learning experience. By their definitions, one can
experience flow or presence, or sometimes both simultaneously.
Table 1, below, presents a comparison of the key salient characteristics of flow
and presence, and shows that while the two concepts share a number of characteristics,
there are many other facets which do not correspond well.
In the concept of flow, the subject has a behavioral and mental commitment to an
activity (Csikszentmihalyi, 1990). Similarly, presence is regarded as a state with
significant behavioral, mental, and emotional involvement (Bouchard, Dumoulin,
Michaud, & Gougeon, 2011; Lee, 2004; Lombard & Ditton, 1997). Both concepts
represent subjective experiences that are difficult to measure through outward
observation. When flow is strong, the subject has a perception of interactivity; similarly,
presence is precipitated by both para-authentic and artificial cues and affordances
(Csikszentmihalyi, 1990; Lee, 2004).
PRESENCE AND ENGAGEMENT 68
The two concepts begin to diverge, however, in a few key areas. For example, a
subject must have an intrinsic interest and motivation for participating in an activity for
flow to occur (Csikszentmihalyi, 1990; Liao, 2006). With presence, however, the
experience is often pleasurable, but can also be simply neutral (Lee, 2004). While
watching a sporting event live or manipulating a remote robotic arm might be an exciting
and positive experience, presence can also occur in neutral contexts, such as when simply
using a phone or videoconferencing to discuss an academic issue with a colleague, or
even when frustrated.
Table 1
Comparison of Salient Characteristics of Flow and Presence
Flow Comparison Presence
Behavioral and
Mental Commitment
STRONG
Behavioral, Mental,
and Emotional Involvement
Subjective Experience STRONG Subjective Experience
Perception of Interactivity STRONG
Para-authentic and Artificial
Affordances and Cues
Intrinsic Interest, Motivation WEAK
Often Pleasurable, but
sometimes simply Neutral
Focused Attention POOR Active or Passive
Loss of Self-Consciousness/
Situational Awareness
WEAK
Diminished Ability to
Distinguish
from Real Experience
Self-Interest, Solo Experience POOR
Probable Social Interaction,
Emotional engagement
PRESENCE AND ENGAGEMENT 69
Flow requires focused attention in order to experience it, whereas an experience
with a high degree of presence can be either active or passive (Csikszentmihalyi, 1990;
Lombard & Ditton, 1997; Nakatsu et al., 2005; Nakamura, & Csikszentmihalyi, 2002).
Given adequate sensory stimulation, presence can occur without the consciousness of the
subject, and even without focused attention (Biocca, 1997).
The hallmark of flow is a loss of self-consciousness and situational awareness,
which is similar to presence’s defining feature of diminished ability to distinguish virtual
experience from real experience (Csikszentmihalyi, 1990; Lee, 2004). Presence,
however, doesn’t necessarily require a loss of self-consciousness — para-authentic and
artificial self presence actually transfer self-consciousness to remote environments or
project one’s self-image onto an artificial construction (Lee, 2004).
Critically, flow occurs as a part of a individual, solo activity, where the subject is
engaged only with himself or herself (Csikszentmihalyi, 1990; Liao, 2006; Novak et al.,
1998). Presence, on the other hand, can be self-oriented (self presence), socially-engaged
(social presence), or neutrally-associated (physical presence), and nearly always involves
interaction with another person, object, or environment, even when self-directed and self-
interested (Lee, 2004; Lombard & Ditton, 1997). In all domains, presence has inherent
social and emotional effects, focused either inwardly or outwardly (Lombard & Ditton,
1997).
Despite these clear distinctions, flow is often used by educational psychology
scholars as a stand-in for the concepts recognized in presence literature. Some scholars
operationally define components of flow in ways that would be recognized as presence
by scholars in that field.
PRESENCE AND ENGAGEMENT 70
Novak, Hoffman, and Yung (1998) define telepresence as a component of flow in
a way most similar to Csikszentmihalyi’s (1990) interactivity or the idea of social
presence as defined by Lee (2004). Novak, Hoffman, and Yung (2000) conceptualize the
experience of the World Wide Web in terms congruent with both presence and flow.
Liao (2006) uses flow theory as a useful as a means of measuring engagement
with a computer-mediated distance education environment, defining flow in a similar
way to para-authentic social presence. Liao posits that flow is dependent on perceived
skills, a feeling of challenge, a feeling of control, and perception of interactivity, which
map well to social and self presence models.
Hoffman and Novak (1996) define flow in terms that echo presence: “the state
occurring during network navigation, which is (1) characterized by a seamless sequence
of responses facilitated by machine interactivity, (2) intrinsically enjoyable, (3)
accompanied by a loss of self-consciousness, and (4) self-reinforcing” (p. 57), though
points 2 and 4 are not required by the definition of presence used in this study. Finneran
and Zhang (2003) also begins to approach the definition of presence, recasting flow in
computer-mediated environments with person, artifact, and task components that bear
similarities to Lee’s (2004) physical, social, and self categorizations.
Joo, Joung, and Kim (2013) offer an extensive study of the effects of flow and
presence in a corporate e-learning environment. While Joo, Joung and Kim avoid the
broader term presence, they do examine teaching presence and cognitive presence, which
could be considered as sub-categories of para-authentic social presence relating to the
instructor’s sense of presence and the student’s sense of presence. Flow, in the Joo,
Joung, and Kim model, emphasizes the motivational aspects of the common definition,
PRESENCE AND ENGAGEMENT 71
and is a close analogue for learning engagement. Their findings indicate an unexpectedly
complex relationship between presence, flow, satisfaction, and persistence, where
teaching presence directly influences persistence and cognitive presence directly affects
flow, which in turn, affects satisfaction, which affects persistence (Joo et al., 2013).
Cognitive Theory for Multimedia Learning and the Multimedia Principle
The other concept in educational psychology worth examining in the context of
presence is the cognitive theory for multimedia learning (CTML). Cognitive theory for
multimedia learning (CTML) uses cognitive load theory to inform the principles of
multimedia-based pedagogy, addressing what elements of media contribute to or mitigate
extraneous or germane load and what factors can improve the learning process (Mayer,
2005, 2009; Mayer & Moreno, 1998, 2003). Both presence and CTML address core
cognitive functioning, and posit ways in which an experience can be molded to
manipulate the mind’s processing of ideas.
CTML is predicated on three core principles: (a) the human information
processing system is dual-channel (visual and auditory), (b) each channel has a limited
capacity, and (c) learning requires active processing in a coordinated process of selecting,
organizing, and integrating (Mayer, 2005; Mayer & Moreno, 1998). Presence is also
concerned with the impact of sensory input on cognition, though it considers all senses,
rather than strictly vision and hearing, and takes into account the existence and quality of
social cues and affordances contained within sensory inputs (Biocca, 1997; Lombard &
Ditton, 1997). For these reasons, presence is more highly sensitive to the authenticity of
an experience, and driven by social and emotional factors that aren’t specified in the
CTML frame. While CTML principles hold that an animated display of a water engine
PRESENCE AND ENGAGEMENT 72
might be more effective than a textual description, even with concurrent narration, it
would not be a high-presence experience (though it would be an improvement in
presence over presenting the text alone).
Mayer’s (2005) multimedia principle, that people learn more deeply with words
and images than with either alone, does have a resonance with the concept of presence.
In a review of 40 studies, Mayer (1999) affirms that focused, mixed-mode, integrated
visual and aural presentations are superior for knowledge transfer. Ditton (1997)
suggests that memory of an experience or task is improved when there is a high degree of
presence, which, like multimedia experiences, benefits from enriched and broad sensory
input.
Mayer’s (2005, 1998) conception of CTML specifies a series of five processes
that take place in the conscious mind: (a) selecting relevant words from presented text or
narration, (b) selecting relevant images, (c) organizing the words into a coherent
representation, (d) organizing the images into a coherent representation, and (e)
integrating both the verbal and pictorial representations with each other and with prior
knowledge. Spinoza’s conception of belief, which underpins presence, holds that all
sensory input is accepted and then processed subsequently, allowing for a more passive
sequence of mental processes, in which ‘feeling’ takes a greater role, and mental
representations can be more easily influenced by passive acceptance of sensory inputs
(Gilbert, 1991).
Another key area of divergence is that, similar to flow theory, CTML specifies
that active cognitive processing requires attention, and consists of building mental models
or schema (Mayer et al., 2001; Mayer & Moreno, 1998; Moreno & Mayer, 2000).
PRESENCE AND ENGAGEMENT 73
Presence can often be a passive experience, though the mind does call on established
models and schema when interpreting sensory input, suggesting that high-presence
experiences may have ameliorative effects similar to multimedia instruction grounded in
CTML principles (Biocca, 1997; Lee, 2004; Lombard & Ditton, 1997). Ditton’s (1997)
observed memory improvements in high-presence situations may hinge on schema
construction activities just as specified by CTML. In Mayer’s (2005) conception,
presented material must have a coherent structure and should provide guidance to the
learner for how to construct a mental model/schema. High-presence experiences rely on
familiar schema in the subject’s past, cued by current stimuli (Biocca, 1997; Lee, 2004).
As such, it is not unreasonable to imagine that such recalled schema may then be primed
for additional modification and construction.
Overload of the senses is a key concern for CTML. Though increased sensory
stimulation can often enhance presence, it is the richness of the input, rather than the
quantity that matters (Biocca, 1997; IJsselsteijn et al., 2000). Moreno and Mayer (2000)
find that extraneous music and sound effects overload auditory working memory
capacity. In presence, increased sound is not as valuable as increased appropriateness
and realism of sound (Biocca, 1997). Similarly, Bartsch and Cobern (2003) find that
unrelated or irrelevant images in a PowerPoint presentation correlated with poorer
performance on recognition and recall tasks.
Similar findings have been established regarding the coherence effect and
seductive details. The coherence effect holds that interesting but irrelevant content
(seductive details) presented concurrently or prior to relevant (‘target’) content will
negatively affect the learner’s ability to recall the target content later (Garner, Brown,
PRESENCE AND ENGAGEMENT 74
Sanders, & Menke, 1992; Garner, Gillingham, & White, 1989; Lehman, Schraw,
McCrudden, & Hartley, 2007; Mayer et al., 2001; Sanchez & Wiley, 2006). Mayer,
Heiser, and Lonn (2001) confirm the coherence effect with three experiments among
college students, in which seductive details added to animation and narration both before
and during a presentation led to poorer performance on knowledge transfer instruments.
Other research (Garner et al., 1989, 1992; Lehman et al., 2007; Sanchez & Wiley, 2006)
consistently confirms this interference with memory recall and knowledge transfer at
various age levels and media types, especially with complex material.
Seductive details appear to interfere with learning by priming learners for a
schema that is unrelated to the actual schema intended for learners to acquire (Harp &
Mayer, 1998). It should be noted, though, that Park, Moreno, Seufert, and Brünken
(2011) approach the dangers of seductive details from a perspective of the limits on
working memory, and their study suggests that in low-load conditions, seductive details
may actually enhance learning performance. Even with their 2 x 2 experiment in a high
school setting (n = 100), however, high-load conditions showed the same interference
with recall and knowledge transfer as in other studies.
In two experiments, Mayer (2001) finds that college students tested more poorly
for knowledge retention and transfer when presented with on-screen text simultaneous to
animation and audible narration, suggesting an overload of the visual channel.
Considering this same evidence through a lens of presence, the text and animation would
represent an awkward mixing of modes: presence achieved through text calls for the
reader to engage their imagination and depends on the believability of situations, people,
and actions as compared to the reader’s past experiences, whereas presence achieved
PRESENCE AND ENGAGEMENT 75
through the visual ‘channel’ alone is based on the realism of what is observed directly
(Lee, 2004; Nakatsu et al., 2005). In this way, a text-based experience, though
employing the visual faculties of the reader to receive it, hinges on imagined sensory
input, rather than the actual visual input of printed words (Lee, 2004). Mixing the two
uses of vision therefore distracts from the realistic replication of an experience, and
becomes a factor in reducing the overall presence of the experience.
The modality effect suggests that using both visual and auditory modes
simultaneously improves knowledge retention (Brünken, Plass, & Leutner, 2004; Moreno
& Mayer, 2000; Reinwein, 2012). Mousavi, Low, and Sweller (1995) find evidence in
six trials that using mixed auditory and visual modes as a part of a presentation improves
the capacity for processing information through reduction of cognitive demands.
Through a presence lens, though, this type of improvement is explained through a more
holistic stimulation of the senses and coordination between audio and visual input, rather
than simply mixing the two channels of reception (Biocca, 1997; IJsselsteijn et al., 2000).
While too much cognitive load may interfere with the ability to process new
information, some studies suggest that a positive emotional or motivational effect may
exist for extraneous details such as background music, functioning in parallel to cognition
(Dove, 2009; Huk, Bieger, Ohrmann, & Weigel, 2004; Kampfe, Sedlmeier, & Renkewitz,
2011; Linek, Marte, & Albert, 2011). Linek, Marte, and Albert (2011) show no
significant learning interference from the use of background music in a video game
environment, but do note a significantly positive motivational effect. Both Huk, Bieger,
Ohrmann, and Weigel (2004) and Thompson, Schellenberg, and Letnic (2012) suggest
that prior knowledge is the salient factor affecting whether any given multimedia element
PRESENCE AND ENGAGEMENT 76
offers more affective and motivational benefit than cognitive detriment (presumably
through overload).
The highly emotional and often social nature of presence experiences is an aspect
not well mirrored by CTML, but is better reflected in the extended multimedia learning
theory proposed by Huk et al. (2004). Extended multimedia learning theory builds on the
concepts of CTML, but integrates the elements of affect, mood, and emotion as factors to
weigh in consideration of the effectiveness of a multimedia presentation (Huk et al.,
2004). Huk et al. apply arousal theory to posit that increased affective and emotional
responses contribute to stronger attention and focus, and therefore contribute to the
learning process. High-presence experiences often produce such affective and emotional
arousal (Lombard & Ditton, 1997). Bouchard, Dumoulin, Michaud, and Gougeon (2011)
also find that presence is stronger when emotions are higher.
Measures of Presence
Because of the internal, cognitive and affective nature of the sense of presence,
measurement of its strength is fraught with challenges, including the disadvantages of
both offline measurement, such as pencil and paper self-reports, and online measurement,
such as the real-time monitoring of physiological responses. IJsselsteijn, de Ridder,
Freeman, and Avons (2000) refer to these two categories as subjective measures and
objective corroborative measures.
Retrospective self-reports are common instruments for measuring the sense of
presence, owing to the highly-subjective nature of presence and the relative ease and
inexpensiveness of self-reports (IJsselsteijn et al., 2000; Lombard & Ditton, 2004). Self-
reporting of presence is often accurate, though it does not capture subtle or subconscious
PRESENCE AND ENGAGEMENT 77
effects of presence well (Lombard & Ditton, 1997). Though there is some potential for
real-time self reporting, most are retrospective, and therefore incur a response bias that is
a threat to construct validity (Early, 2006). In the case of interpersonal tests of social
presence, measuring homophilly—the degree to which a subject feels the other person in
the mediated interaction is similar in attitudes, behaviors, or emotions—can provide
additional insight about the subject’s degree of presence (Biocca et al., 2003).
Accessible retrospective self-report instruments include the ITC-Sense of
Presence Inventory (ITC-SOPI) (Lessiter, Freeman, Keogh, & Davidoff, 2001), the
Temple Presence Inventory (TPI) (Lombard & Ditton, 2004), and the Early Sense of
Presence Inventory (ESoPI) (Early, 2008). All three comprise Likert-scale items (5-point
scales in the ITC-SOPI; 7-point scales in the TPI and ESoPI) addressing a broad range of
presence factors. The ITC-SOPI uses 44 items in four sub-scales: (a) sense of physical
space, (b) engagement, (c) ecological validity, and (d) negative effects (Lessiter et al.,
2001). The authors recommend analyzing each separately, as they expect each to be
differentially influenced by environmental manipulations (Lessiter et al., 2001). The TPI
uses 42 Likert scale items in eight sub-scales (Lombard & Ditton, 2004). The ESoPI,
which is based on the TPI, comprises 48 Likert scale items, one open-ended question, and
10 demographic questions (Early, 2013). The ESoPI items are not explicitly divided into
sub-scales, but 42 of the items match those of the TPI and its eight sub-scales.
Some measures use direct observation of subjects’ behaviors when placed in
experimental conditions. “To understand how large the potential a medium has to change
an individual, researchers have typically measured how realistically a user behaves while
inside of that medium” (Yee et al., 2007, p. 116). Such observations are an excellent
PRESENCE AND ENGAGEMENT 78
gauge of some types of presence, especially physical presence and some aspects of social
presence, where there is an expectation of outward reaction or social interaction through a
mediated interface, and observers can compare the quality and quantity of behavioral
interaction or synchronization with that in non-mediated analogues (Biocca et al., 2003).
Involvement, identification, and coordination of behavior could be used as indicators in
these cases (Biocca et al., 2003). Such observations, however, do not track the finer
detail of subjects’ emotional and affective experience, or the details of their
psychological state. Because of this, social presence lends itself to subjective
measurement in the areas of richness, involvement, immediacy, intimacy, social
judgment, and behavior (Biocca et al., 2001).
Physiological measurement of minor changes in the subject’s physical state offers
the potential of finer detail and reduction of response bias, but direct correlations to the
sense of presence are difficult to draw, since presence is most often considered to be a
psychological/mental state and the true causes of physiological indicators are difficult to
pin down (Biocca et al., 2003; Lombard & Ditton, 1997). Additionally, given the
sensitivity of presence to environmental conditions, the instrumentation used to collect
physiological data itself is likely to affect the responses, though technological advances
may offer opportunities for less-intrusive data gathering (Early, 2006).
As a latent construct, presence favors triangulation through multiple measures and
multiple instruments, using both objective and subjective approaches (Early, 2006;
IJsselsteijn et al., 2000). Biocca, Harms, and Burgoon (Biocca et al., 2003) further
emphasize that presence should always be measured and represented as a scalar value, as
it exists on a continuum, rather than as simply a Boolean.
PRESENCE AND ENGAGEMENT 79
Presence in the Context of Learning Engagement
Given the universality of presence across different domains of the human
experience, strategies for heightening presence would have cross-applicability in various
fields of study — i.e., what increases presence in one field would similarly increase it in
another field. Biocca et al. (2003) emphasize the possibilities inherent in the study of
presence. Describing social presence, they argue that
by addressing issues of what essential attributes are needed to establish
connection with others, we may arrive at a better understanding of how humans
arrive at that sense of mutuality that underpins all communication between people
and that is a prerequisite to establishing common ground (Biocca et al., 2003, p.
10).
Thus, an understanding of strategies for increasing presence in other disciplines
may inform strategies for learning engagement if the two are interrelated (Dickey, 2005;
Lombard & Ditton, 1997). For example, Dickey (2005) directly links virtual
environment and computer/video game design with learning engagement, offering that
the central purpose in game design is to engage players, and thus the same strategies that
work in game design may also work in classrooms where the goal is to engage learners.
In gaming and simulations, the quality of a user’s experience plays a critical role in
engagement and the feeling of social interaction (Harteveld et al., 2011).
Presence may have an ameliorative effect on learning engagement through
integration of sensations and corresponding reduction of cognitive load. In high-presence
experiences, the mediation of reality fades from consciousness, potentially freeing
capacity in working memory. The emotional and affective stimulation associated with
some presence experiences could also influence learning engagement. The Vivaldi effect
finds that playing familiar music concurrent with activities improves affect as well as
PRESENCE AND ENGAGEMENT 80
learning and even physical ability (Kampfe et al., 2011; Mammarella, Fairfield, &
Cornoldi, 2007).
In the immersion hypothesis, the feeling of presence leads to active cognitive
processing, which, in turn, leads to meaningful learning outcomes (Clark, Early, & Yates,
2006). Lombard and Ditton (1997) also emphasize that higher degrees of presence
correspond with deeper engagement and better learning and memory outcomes.
It is also possible that presence and learning engagement are simply ‘sister
phenomena,’ and describe the same, or highly-related conditions (Dickey, 2005). Table
2, below, presents a comparison of the salient characteristics of presence and learning
engagement. The comparison highlights the many similarities that exist between the key
elements of both concepts.
Both terms describe a state of behavioral, cognitive, and emotional or affective
involvement in a situation, and both are highly subjective (Biocca, 1997; Fredricks et al.,
2004; Lee, 2004; Lombard & Ditton, 1997; Skinner et al., 2008). Both concepts rely on
cues and signals within the environment, many of which are subconsciously noted
(Biocca et al., 2001; Coates, 2006; Fredricks et al., 2004; Kearsley & Shneiderman, 1998;
Lee, 2004; Lombard & Ditton, 1997). Both learning engagement and strong sense of
presence can elicit pleasurable, memorable feelings, but both can also be neutral
conditions (Biocca, 1997; Dickey, 2005; Fredricks et al., 2004; Jimerson et al., 2003;
Lee, 2004).
Even with factors that do not match exactly, there exists a fair comparison
between the two concepts. Presence can be experienced by a passive subject or an
actively engaged one, whereas learning engagement is a necessarily active form (Dickey,
PRESENCE AND ENGAGEMENT 81
2005; Fredricks et al., 2004; Kearsley & Shneiderman, 1998; Lee, 2004). Presence is
gauged by the fidelity of a simulated or replicated experience, whereas learning
engagement is itself a real experience, though it is similarly heightened by the
authenticity of the learning task (Biocca et al., 2003; Kearsley & Shneiderman, 1998;
Lee, 2004; Lombard & Ditton, 1997). Learning engagement is heightened by
social interaction, which is, of course, the hallmark of social presence, but not
necessarily required for physical or self presence (Fredricks et al., 2004; Lee, 2004;
Skinner et al., 2008).
Table 2
Comparison of Salient Characteristics of Presence and Learning Engagement
Presence Comparison Learning Engagement
Behavioral, Mental,
and Emotional Involvement
STRONG
Behavioral, Cognitive,
and Affective Involvement
Subjective Experience STRONG Subjective Experience
Para-authentic and Artificial
Affordances and Cues
STRONG
Social Cues and
Affordances
Often Pleasurable, but
sometimes simply Neutral
STRONG
Often Pleasurable, but
sometimes simply Neutral
Active or Passive FAIR Active
Diminished Ability to Distinguish
from Real Experience
FAIR Authentic Experience
Probable Social Interaction,
Emotional Engagement
STRONG Social Interaction
PRESENCE AND ENGAGEMENT 82
It is useful to examine three particular subdomains of presence closely, as they
are especially applicable to learning: social presence, mental immersion, and richness
or realism.
Social presence and co-presence. The domain social presence is particularly
important for educators, as the importance of collaboration and of social interaction, both
among peers and with the instructor, are ingredients for strong learning engagement.
Though often conceived of as very different, Vygotsky, Piaget, and Bandura all
emphasize some common themes important to effective teaching and learning, among
which is the value of social interaction (Tudge & Winterhoff, 2010). Furrer and Skinner
(2003) observe that in third through sixth graders, social “relatedness” among peers and
between students and teachers is critical to learning engagement, motivation, and
academic performance. Sulbaran and Baker (2001) look at the use of distributed virtual
reality in higher education and find that it not only allows for social interaction between
engineering students, but also improves learning outcomes.
Biocca et al. (2003) describe social presence as the sense of being with another
(through or in a medium). “Social presence research may be means of exploring the
larger issues in theories of mind, social cognition, and interpersonal communication”
(Biocca et al., 2003, p. 463). Sensory presence of the other defines social presence in
both real and virtual spaces (Biocca et al., 2003).
The term co-presence extends the concept of social presence with the idea of
mutual awareness: both the subject and the other are aware of each other’s presence
(Biocca et al., 2003). The body of the other is not as important to perceive as cues of
another intelligence that animates the entity — Biocca et al. (2003) term this
PRESENCE AND ENGAGEMENT 83
psychological involvement. “In a definitional approach that seeks to connect both
mediated and un-mediated approaches, the body—be it virtual or physical—is
conceptualized as a medium that provides cues to the intentional states of another”
(Biocca et al., 2003, p. 472).
Sproull’s (1973) study of prekindergarten children exposed to television
illuminates the power of social presence. The study found that viewing Sesame Street
increased attention and social interaction behaviors among the children, as if characters
performed on the show were real actors in the children’s environment. Lombard and
Ditton (1997) note that simplistic cues of ‘human-ness,’ such as speech or the use of
interactive language are enough to create social presence around inanimate objects.
Webster, Trivino, and Ryan (1994) report that computer interaction is most effective
when cues of humanity stimulate a sense of “playfulness” that suggests a personality
rather than mechanical function. Such social cues are instrumental in establishing social
presence between real people separated by an artificial medium, as when a television
news anchor simulates eye contact and personal space with the viewer (Lombard &
Ditton, 1997).
Picciano’s (2002) results show that the degree of social presence felt by learners
predicts performance on a written assignment (though performance on the final
examination was not statistically significant). He notes that interaction is necessary for
social presence, but cautions that it is not equivalent to presence and that presence cannot
be assumed just because there is interaction. Moreno and Mayer (2004) point to
personalization and individualized consideration as factors that are also influential in
PRESENCE AND ENGAGEMENT 84
increasing students’ feeling of social presence and correlate with better learning
outcomes.
In a study of the massively-multiplayer online role-playing game Second Life,
Yee, Bailenson, Urbanek, Chang, and Merget (2007) find that social norms regarding
person-to-person interaction (the relationship between gender, interpersonal distance
(IPD), and eye gaze) function in the same way in the virtual environment as they would
in the real world. “In order to understand the impact that a certain medium has upon an
individual, it is important to understand how well he or she differentiates the media from
actual reality” (Yee et al., 2007, p. 115). Their evidence supports the idea that social
interaction in a mediated environment can be as effective as in an unmediated
environment—the heart of social presence. Their study also supports the notion that
expected psychological and sociological principles function in an equivalent manner in
virtual environments.
Mental immersion. Another key category of presence with important
educational implications is the sense of mental immersion, given that it can focus
attention, promote full learning engagement, and reduce extraneous cognitive load.
Lombard and Ditton (1997) assert that mental immersion is often the hallmark of
artificial environments with high degrees of presence. Mental immersion most closely
corresponds to Lee’s (2004) self presence subdomain. It describes a level of cognitive
focus on the task and simulation at hand, and a conception of one’s own place within an
experience.
Riva and Waterworth (2003) argue that the sense of presence is directly related to
the integration of the three layers of consciousness—the more integrated they are
PRESENCE AND ENGAGEMENT 85
(focused), the more present the person is. They assert that presence can be classified
based on its neurological function, matching the levels of consciousness: proto presence,
core presence, and extended presence. According to Riva and Waterworth, presence can
be described in terms of focus (integration of the selves), locus (real or mediated world),
and sensus (state of arousal). In their conception, media attempt to engage one or more
layers of presence, with extended presence being the easiest to invoke, and proto
presence being the hardest—nearly impossible, but approachable with highly-
sophisticated physical simulations.
Mental immersion is linked with the concept of transportation, which Green,
Brock, and Kaufman (2004) describe as the mechanism by which narratives can affect
beliefs. This power of transportation that can affect beliefs aligns well with the
educational goals of impressing new materials on students.
Additionally, Vorderer, Klimmt, and Ritterfeld (Vorderer et al., 2004) also note
that enjoyment increases as the experience of a media-based environment becomes more
fully immersive and interactive. In the classroom, this effect could not only lead to more
positive affect and learning engagement in the short term, but also to satisfaction and
retention/matriculation in the long term.
Richness, realism. Richness and realism are also particularly important aspects
of presence for the educator to consider. Realism and richness can affect the degree of
social presence and mental immersion. The more realistic the experience, the more likely
the viewer is to suspend disbelief and experience presence (Lombard & Ditton, 1997;
Norman, 1990).
PRESENCE AND ENGAGEMENT 86
Green and Brock (2000) assert that beliefs are shaped only when those
experiencing a narrative believe it as if it were real. Li, Daugherty, and Biocca (2002)
find that more realistic, three-dimensional representations increase both presence and
positive advertising-related behaviors that mirror educational goals: retention,
recognition, and intent/motivation.
Moreno and Mayer (2002) find in two experiments that realism of the media can
have a positive impact on the subject’s overall sense of presence, with more realistic and
immersive presentations evoking greater degrees of presence. While their study failed to
support either positive or negative impacts on learning, the authors caution that this result
is likely due to the lack of realism in their simulation.
Summary and Conclusion
Whether physical or virtual, the design of the learning environment has an impact
on the success and quality of knowledge and skill acquisition. Learning engagement is
the broadly-defined description for mental and behavioral commitment to the learning
process. Optimizing learning engagement is or should be a primary goal for educators.
The sense of presence describes a cognitive and affective response to an artificial
or mediated experience that blurs the distinction between simulation and reality. Given
the ameliorative qualities of high-presence experiences and the mental engagement they
engender, presence is a useful framework with which to consider, measure, and improve
the conduciveness of a given environment to the social and intellectual interaction that is
necessary for learning. The literature suggests that high-presence experiences improve
memory, engagement, positive affect, and focus (Lombard & Ditton, 1997). When the
PRESENCE AND ENGAGEMENT 87
sense of presence is weak or disrupted, extraneous load is dramatically increased, and
positive social and motivational effects are reduced (Lombard & Ditton, 1997).
The function of the learning process is largely internal and difficult to measure;
likewise, the sense of presence is subjective and challenging to measure. Given the
conceptualization of presence and the conceptualization of learning engagement, there is
a likelihood that presence is an influencer of learning engagement. Thus, measurement
of presence, and our understanding of its impact, may shed light on new strategies for
improving the measurement and encouragement of learning engagement. The body of
presence theory and research offers new and potentially valuable ways to conceptualize,
influence, and measure the quality of a learning environment.
PRESENCE AND ENGAGEMENT 88
CHAPTER THREE: METHODOLOGY
The present study examined the sense of presence and its relationship to learning
engagement in the physical classroom and the virtual learning environment. The quasi-
experimental study employed quantitative and qualitative methods to assess master’s
degree students using convenience sampling. Utilizing a self-report instrument designed
to investigate participants’ sense of presence combined with a self-report instrument
designed to investigate participants’ learning engagement, the study examined the ability
of presence to predict learning engagement, and explored the relationship of the learning
environment (including specific changes within that environment) to the sense of
presence.
The research questions enumerated in Chapter 1 were each considered using
statistical analysis. The study also employed structural equation modeling to examine the
relationship of presence to learning engagement. Since both of these concepts are latent
factors (unable to be measured or observed directly), structural equation modeling
provides a useful way to understand the possible relationship between the two factors. A
hypothesized model for the relationship of presence to learning engagement was tested to
understand the degree of fit, and modified until the best possible fit was achieved (see
further detail in Chapter 4). Follow-up qualitative analysis was also conducted, and is
discussed separately in Chapter 4.
Research Design
The Institutional Review Board (IRB) of the participating university approved the
study, and all necessary clearances were obtained from the program in which the study
was conducted. Students attending the same class but in different locations—through
PRESENCE AND ENGAGEMENT 89
distance education systems and in a traditional classroom setting—were surveyed to
understand their level of presence and their degree of learning engagement. The study
employed a sense of presence measurement instrument and an accepted learning
engagement inventory, administered in four higher education scenarios: a physical
classroom in a traditional format, a physical classroom augmented with real-time
computer-based text chat, a virtual learning environment with both audiovisual and text
chat elements, and a streamlined virtual learning environment with no text chat interface.
The physical classrooms and virtual learning environments each used the same course
material as the other, and in some cases, the same instructor. Both instruments were
combined and edited into a single instrument, so as not to overwhelm participants, and
were coded in different sub-categories to facilitate detailed modeling.
The study assessment was quantitative, and was conducted in two class sessions,
or phases. The self-report instrument measuring sense of presence and learning
engagement was administered following a typical online learning session in the virtual
learning environment and an on-campus classroom session of the same material with the
availability of real-time text chat (Phase 1). To assess the effects of alternate physical
classroom and virtual interface designs, the classroom and VLE conditions were altered
for a second class session and subsequent quantitative measurement (Phase 2). A
diagram of the general research design is provided in Figure 1.
PRESENCE AND ENGAGEMENT 90
Participants and Setting
Participants were drawn from 12 sections of two courses in a master of arts in
teaching program at large private research university in southern California. The class
size for online class sections in the studied program averaged 13.875 students, while in-
person sections averaged 17 students. The average number of participants per class in the
Figure 1. Participant flow chart showing overall research design. *Enrollment
was established prior to Phase 1 administration, but participants were also
reminded of informed consent prior to Phase 2 administration.
PRESENCE AND ENGAGEMENT 91
sample was 11.33 students in online classes and 15.33 students in on-campus classes,
with total sample sizes of n = 46 in the classroom environment and n = 102 in the virtual
learning environment. While the course material was identical across the sections of each
class, some class sessions were conducted in a physical classroom with students in a
traditional in-person format, while other sections utilized the school’s existing virtual
learning environment (LMS software solution) as cohorts of distance learners. Eight
instructors taught the sections, with three teaching more than one section. Where
appropriate, data have been separated by course material and instructor, but are presented
in aggregate unless otherwise noted.
The VLE software system employed was Adobe’s Connect platform, which offers
both audiovisual and text elements, including an audiovisual chat that allows users to see
and hear all other participants (arrayed in a grid on the screen), text chat that allows for
real-time text messaging viewable by all participants, and multimedia slides presented by
Figure 2. A representation of the appearance of the VLE in use.
PRESENCE AND ENGAGEMENT 92
the teacher and viewable to all participants. See Figure 2 for a representation of the
general appearance of the software solution as deployed at the participating institution.
The overall sample included 148 students, with 46 on-campus and 102 online.
The sample was heavily female, with 102 (29 on-campus; 73 online) self-reporting as
female and 37 male (16 on-campus; 21 online). Race or ethnicity reported by the
subjects (each could choose more than one) were 73 White, 31 Hispanic/Latino, 34
Asian/Pacific Islander, 3 Black/African American, 2 Native American/Indian, and 5
other/mixed. Of the 138 participants that reported their ages, the average age was 28.36
years old (24.08 years old on campus; 30.44 years old online), with the most common age
being 22 (22 on-campus; 26 online).
Instrumentation
Learning engagement was measured by administration of the Student Course
Engagement Questionnaire (SCEQ, described in Chapter 2) (Handelsman et al., 2005).
This survey uses 23 five-point Likert scale items to explore participants’ skills, emotions,
participation, interaction, and performance in a specific course, along with a few more
global characteristics such as orientation toward an incremental theory of learning and
goal orientation. The entire form was used for this study.
The Early Sense of Presence Inventory (ESoPI; described in Chapter 2) (Early,
2008) was selected as a means of assessing participants’ level of presence during the
instructional experience. The inventory was adapted from the Temple Presence
Inventory questionnaire with permission. The instrument comprises 47 Likert scale items
and 12 demographic inquiries.
PRESENCE AND ENGAGEMENT 93
For the purposes of this study, only the first 35 items were selected, and the
demographic questions were replaced by three targeted demographic items (which were
optional). One additional item was added to assess participants’ overall affective
impressions.
Both the original ESoPI and the modified version used in this study are provided
in their entirety as Appendix A and B, respectively. The modifications were made under
the supervision of the original instrument designer, and were not expected to
meaningfully alter the internal validity of the instrument.
The two instruments were combined into one form so as not to overwhelm
participants. The combined inventory remains a self-report instrument, conducted via a
paper form for the classroom participants and via web-based survey form for the virtual
learning environment participants.
Variables
Learning Engagement
Learning engagement is a dependent variable, as measured by the Student Course
Engagement Questionnaire (SCEQ, described in Chapter 2) (Handelsman et al., 2005).
Overall learning engagement is expressed as an average of all responses on the SCEQ
items on the combined instrument, each of which is a Likert scale item with responses
ranging from 1 (very characteristic of me) to 5 (not at all characteristic of me). All items
were reverse coded to indicate the degree of learning engagement, after which, 1
indicates low engagement, and 5 indicates full learning engagement.
The SCEQ is also subdivided by the instrument authors into four subscales: (a)
skills engagement, (b) emotional engagement, (c) participation/interaction engagement,
PRESENCE AND ENGAGEMENT 94
and (d) performance engagement. The average score of the items in each subscale
represents the participant’s score for that subscale.
Both the total score of all items and scores for each sub scale were calculated and
used in the statistical analysis that follows in Chapter 4.
Sense of Presence
The independent variable studied was the participants’ sense of presence, as
measured by the Early Sense of Presence Inventory (ESoPI) (Early, 2008). Overall
presence is expressed as an average score on the 47 ESoPI items on the combined
instrument, each of which is a Likert scale item with responses ranging from 1 (not at all
/ never / not well) to 7 (very much / often / very well). Some items are reverse coded to
ensure proper agreement with other items in indicating strength of presence sensed, with
1 indicating virtually no sense of presence, and 7 indicating a complete immersion and
total sense of presence within the environment.
The ESoPI also designates six sub scales, and items within the ESoPI are further
coded for their indication of loadings on these factors: (a) spatial presence (corresponding
to physical presence - artificial/para-authentic), (b) social presence/parasocial interaction,
(c) social presence - passive interpersonal (corresponding with para-authentic social
presence), (d) social presence - active interpersonal (corresponding with para-authentic
social presence), (e) engagement (mental immersion) (corresponding with para-authentic
physical presence), and (f) social richness. The participant’s score for each sub scale
consists of an average score of that subscale’s designated items. The items that
correspond with each subscale are provided in Appendix C.
PRESENCE AND ENGAGEMENT 95
Both the total Sense of Presence score and scores for each ESoPI subscale were
calculated and used in the analysis of the results.
Social presence. The average score of items coded for social presence provide an
indication of how much the participant felt that he or she interacted with other people in
an authentic social exchange, whether the experience was mediated by electronic
hardware and software in the online condition or by the physical space configuration and
protocol of the in-person classroom condition. In the ESoPI, these items correspond with
the social presence - passive interpersonal and social presence - active interpersonal sub
scales. Social presence was expected to be the most useful domain of presence to
consider in the educational process. The score of this variable reflects the degree to
which the participant felt she or he was engaged in an authentic and personal interaction
with other real people. When interacting via a virtual environment, the degree of social
presence can vary from a keen awareness of and distraction by the artificial nature of the
interaction (at the low extreme) to a feeling indistinguishable from interacting with a live
human being in the same space, the interface having faded from consciousness (at the
high extreme). Even when one is in the presence of a real person, the sense of whether
one is engaged in a genuine, personal, focused interaction can vary along this spectrum of
quality. The variable measures the participant’s perception of social cues in the scenario,
and how well those cues simulated those in a genuine and personalized social interaction.
Physical presence. The average score of the items coded for physical presence
indicate the degree to which a participant felt that he or she had been transported to a
different environment. In the ESoPI, these items correspond with the spatial presence,
engagement (mental immersion) and social richness sub scales. In the virtual learning
PRESENCE AND ENGAGEMENT 96
environment, this functions as the strength of the participant’s impression that she or he is
sharing a real space with the instructor and peers. In the classroom space, it similarly
reflects the strength of the participant’s sense of being in a well-defined space shared by
the instructor and peers, free of distracting influences. Low scores in both environments
reflect a feeling of distance from the ‘action’ of the educational process.
Virtual Environment Design
One independent variable in the proposed study was the design of the virtual
learning environment interface. The learning management software used by the program
features flexibility in what interaction elements are made available to participants.
Typically, the environment is presented to participants with live video and audio
representations of the other participants in the class session, arrayed in a matrix of
windows. Accompanying this video and audio representation element, a window
provides a continuously presented live stream of text messages issued by participants to
the entire group (participants and instructor). This stream of text information exists
alongside the windows showing the video of the faces of the other participants.
In Phase 1, the virtual learning environment was presented as described, with both
the audiovisual interaction element and the text interaction element functioning and
available. In Phase 2, the text interaction element was disabled, and only the audiovisual
interaction element was presented.
Classroom Design
Another independent variable in the study is the design of the physical classroom
used in each phase of the scenario.
PRESENCE AND ENGAGEMENT 97
In Phase 1, the classroom was organized in a traditional ‘lecture hall’ style, with
rows of student desks facing a designated side of the classroom, where the instructor
stood and faced the students as they sat at their desks. The classroom had the added
element of real-time text chat available concurrently with the instructor’s presentation via
the students’ personal laptops.
In Phase 2, the classroom was organized in the same fashion, but without the
added text chat availability.
Demographics
In addition to the main study variables, age, gender, and race or ethnicity were
requested. These items were all marked as optional, though 93% of the participants did
self-report data for these items. The results are presented in Chapter 4.
Procedure
The study was conducted in four discrete sessions, with two conducted in each of
the proposed environments (see Figure 1). The first session in each environment,
designated Phase 1, was replicated in the second session in each environment, designated
Phase 2, except for the change in the design variables, as described above.
The procedure began with Phase 1 in the classroom environment. The instructor
conducted the class session according to her or his preferred pedagogy, using the ‘lecture
hall’ style classroom design with added text chat. At the conclusion of the class session,
the investigator introduced himself and provided information about the study and the
voluntary nature of participation. Then, he distributed the combined SCEQ-ESoPI
instrument on paper, to be completed by each participant. The instructor collected
PRESENCE AND ENGAGEMENT 98
completed surveys from all students who elected to participate. Those who did not elect
to participate did not complete a survey.
During the same week and same course material unit, the Phase 1 investigation of
the virtual learning environment was also conducted. In this session, the instructor
conducted the class session using the program’s virtual learning interface, with her or his
preferred pedagogical methods. In Phase 1, both the audiovisual interface and the text
stream interface was displayed and enabled for all viewers. At the conclusion of the class
session, the investigator was introduced, and described information about the study and
the voluntary nature of participation as in the classroom session. Students were then
invited to visit and complete a web-based version of the combined SCEQ-ESoPI
instrument using a web browser on their personal computer. Those who elected not to
participate were free to leave the virtual learning environment without completing the
survey.
Phase 2 was conducted the following week, when the classes were discussing the
next unit of the same course material, using the same classroom and virtual learning
environment sections (the same students). The classroom session was conducted in a
similar manner as in Phase 1, but without the added text chat availability. Following the
conduct of the class, the same SCEQ-ESoPI instrument distributed in Phase 1 was again
distributed to participants on paper for immediate completion.
In the same week, using the same course material unit, the Phase 2 investigation
of the virtual learning environment was also conducted. The instructor conducted the
class in her or his preferred style, utilizing the established virtual learning interface. In
Phase 2, however, only the audiovisual interface was displayed and enabled, with the text
PRESENCE AND ENGAGEMENT 99
stream interface disabled and hidden. Following the completion of the class session,
participants were invited to complete the web-based SCEQ-ESoPI form once again.
An additional optional survey was made available to members of the VLE class at
the conclusion of Phase 2, with the goal of gathering additional information about the
students’ affective response to the VLE class session with no text chat availability. This
follow-up survey is discussed in greater detail in Chapter 4.
PRESENCE AND ENGAGEMENT 100
CHAPTER FOUR: RESULTS
The analysis of the data followed a four step mixed-methods process, designed to
investigate the behavior from several perspectives, helping to develop the richest possible
understanding of the behaviors at work in the study. The first step was to examine the
basic demographics and descriptive statistics of the sample. Then, the psychometric
validity of the instruments was tested to confirm appropriate performance with this
particular sample.
Then, the results were analyzed using statistical strategies to understand the basic
relationship between variables as posed by the research questions. These analyses
included completion of multiple regression and ordinary least squares regression analysis,
independent-samples and paired-samples t-tests, and analysis of variance (ANOVA).
The results of these tests were compared against the research questions for the study.
A basic structural equation model was proposed from consideration of applicable
theory and investigated using AMOS software. This initial model was subsequently
modified to improve goodness of fit, producing two viable models that describe the
behavior of the latent factors and observed variables of interest.
Finally, qualitative data was gathered from a follow-up survey, and analyzed for
common themes and insight into the behaviors observed in the initial statistical analyses
and structural equation modeling.
This chapter will review the results of these analyses in detail, beginning with the
descriptive statistics, continuing with the t-tests, regressions, and ANOVA results,
grouped by research question, following with the results of structural equation modeling,
and concluding with a brief look at qualitative data gathered in the follow-up survey.
PRESENCE AND ENGAGEMENT 101
Descriptive Statistics & Demographic Variables
As discussed in Chapter 3, the sample consisted of students engaged in a master
of arts in teaching program, and included both students enrolled in a traditional on-
campus setting as well as students enrolled in a distance education program, conducted
via real-time virtual learning environment (VLE).
Demographic information was provided by 93% of the participants, and is
summarized in Table 3.
The full sample had 148 participants, all of whom were master’s degree students.
The sample demographics were similar in both online and on-campus groups, though the
online group skewed older (mean = 24.08 years of age on campus; mean = 30.44 years of
age online). The reported age ranged from 21 to 49 years old (21 to 35 on campus; 22 to
49 online). Six participants reported having a birthdate between occasions of
measurement, and their reported ages in Phase 1 and Phase 2 were averaged for
calculation of demographic statistics. Of those responding to the gender question, 74%
were female (64% female on campus; 78% female online).
Overall, race was reported to be majority White (53% overall; 37% on campus;
60% online), with a fairly even split between Asian/Pacific Islander (25% overall; 36%
on campus; 19% online) and Hispanic/Latino (23% overall; 25% on campus; 22%
online). Asian/Pacific Islander students were much better represented in the on-campus
sample, where they made up nearly the same percentage as White students; White
students made up 60% of the online group.
PRESENCE AND ENGAGEMENT 102
Table 3
Summary of Reported Demographic Information
Statistic Gender Age (in years) Race/Ethnicity*
On-Campus Classroom
Frequency/
Median
Female=29
Male=16
24.08 White=17
Hispanic/Latino=11
Asian/Pacific Islander=16
Black/African American=0
Native American/Indian=0
Other=1
Mode Female 22 White
n** 45 45 44
Online Virtual Learning Environment
Frequency/
Median
Female=73
Male=21
30.44 White=56
Hispanic/Latino=20
Asian/Pacific Islander=18
Black/African American=3
Native American/Indian=2
Other=4
Mode Female 26 White
n** 93 93 93
Total Sample (both environments)
Frequency/
Median
Female=102
Male=37
28.36 White=73
Hispanic/Latino=31
Asian/Pacific Islander=34
Black/African American=3
Native American/Indian=2
Other=5
Mode Female 22 White
n** 138 138 137
*Participants were allowed to choose more than one ethnicity. **n represents total
participants reporting demographic information of the specified type.
PRESENCE AND ENGAGEMENT 103
The mode for the on-campus participants was a 22-year-old White female, while
the mode for the online participants was a 26-year-old White female.
Due to the small sample size, results were aggregated, and no age-, race-, or
gender-specific statistics were rendered.
Analysis of Statistical Consistency
All variables were scored for both Phase 1 and Phase 2, designated W1 and W2 in
the dataset, using the items from the appropriate administrations of the combined
instrument.
Learning Engagement
The learning engagement component was composed of the items from the Student
Course Engagement Questionnaire (SCEQ; (Handelsman et al., 2005). The total score
for the learning engagement component in Phase 1, with n = 136, had a mean of 4.1, a
median of 4.2, a range of 3.5, and SD = 0.55. In Phase 2, learning engagement, with n =
126, had a mean of 3.9, a median of 4.1, a range of 3.5, and SD = 0.73.
To examine the reliability and validity of the learning engagement scale as a
whole, Cronbach’s alpha was calculated for both W1 and W2. The Cronbach’s alpha
coefficient is a measure of the degree to which the instrument’s items are correlated, as a
way of understanding the internal consistency and general construct validity of the scale
(DeVellis, 2003). DeVellis (2003) recommends the use of instruments with a
demonstrated Cronbach’s alpha coefficient of .70 or greater. Standardized Cronbach’s
alpha was used, with the assumption of equal variances when all items are self-reported
Likert scale items with the same size scale (Falk & Savalei, 2011, p. 451).
PRESENCE AND ENGAGEMENT 104
Handelsman, et al. (2005) report that the SCEQ scale has good internal
consistency (values below in Table 4), and it has good consistency in the current study.
For W1, the overall learning engagement scale standardized Cronbach’s alpha coefficient
was .91; for W2, the standardized Cronbach’s alpha coefficient was .95. Deleting items
would not have a significant effect. The coefficients for each subscale are provided in
Table 4.
The skills engagement subscale, with eight items, had n = 136 in W1 and n = 128
in W2. The scores on this subscale had a mean of 4.0 in W1 and 3.9 in W2, with SD =
.72 in W1 and SD = .84 in W2. The standardized Cronbach’s alpha coefficient was .83 in
W1 and .89 in W2, indicating a high degree of internal consistency in both phases.
The emotional engagement subscale comprises five items, and had n = 138 in W1
and n = 129 in W2. The scores on this subscale had a mean of 4.1 in W1 and 4.0 in W2,
with SD = .76 in W1 and SD = .88 in W2. The standardized Cronbach’s alpha coefficient
Table 4
Psychometrics for Learning Engagement Subscales in the Current Study
Subscale
W1
standardized
Cronbach’s
alpha
W2
standardized
Cronbach’s
alpha
Instrument as
reported by
Handelsman,
et al. (2005)
Skills engagement .83 .89 .82
Emotional engagement .87 .90 .82
Participation/
Interaction
.73 .73 .79
Performance .72 .87 .76
PRESENCE AND ENGAGEMENT 105
was .87 in W1 and .90 in W2, indicating a high degree of internal consistency in both
phases.
The participation/interaction engagement subscale, with five items, had n = 138 in
W1 and n = 131 in W2. The scores on this subscale had a mean of 3.7 in W1 and 3.6 in
W2, with SD = .70 in W1 and SD = .73 in W2. The standardized Cronbach’s alpha
coefficient was .73 in W1 and .73 in W2, indicating a good degree of internal consistency
in both phases.
The performance engagement subscale, with only three items, had n = 138 in W1
and n = 132 in W2. The scores on this subscale had a mean of 4.5 in W1 and 4.3 in W2,
Table 5
Summary Statistics for Learning Engagement
Scale n standardized α mean SD range
Skills W1 136 .83 4.0 0.72 3.5
Skills W2 128 .89 3.9 0.84 3.9
Emotional W1 138 .87 4.1 0.76 4.0
Emotional W2 129 .90 4.0 0.88 3.8
Participation/
Interaction W1
138 .73 3.7 0.70 3.4
Participation/
Interaction W2
131 .73 3.6 0.73 3.4
Performance W1 138 .72 4.5 0.65 4.0
Performance W2 132 .87 4.3 0.88 4.0
Overall Learning
Engagement W1
136 .91 4.1 0.55 3.5
Overall Learning
Engagement W2
126 .95 3.9 0.73 3.5
PRESENCE AND ENGAGEMENT 106
with SD = .65 in W1 and SD = .88 in W2. The standardized Cronbach’s alpha coefficient
was .72 in W1 and .87 in W2, indicating a good degree of internal consistency in Phase 1
and a high degree of consistency in Phase 2 (this volatility is likely due to the smaller
number of items on this subscale).
Sense of Presence
The sense of presence component of the combined instrument comprises 33 items
from the Early Sense of Presence Inventory (ESoPI) (Early, 2008), each a 7-point Likert
scale. The summary score for the sense of presence component in Phase 1, with n = 122,
had a mean of 5.1, a median of 5.2, a range of 5.1, and SD = 0.88. In Phase 2, sense of
presence, with n = 117, had a mean of 4.8, a median of 4.9, a range of 4.9, and SD = 1.20.
To examine the reliability of the sense of presence scale as a whole, standardized
Cronbach’s alpha was again calculated for both W1 and W2. For W1, the overall
standardized Cronbach’s alpha coefficient for the sense of presence was .95; for W2, the
Cronbach’s alpha coefficient was .97, indicating very strong inter-item correlation and a
strong internal consistency. Deleting any one item would not significantly effect the
value. The coefficients for each of the six subscales are provided in Table 6.
PRESENCE AND ENGAGEMENT 107
The spatial presence subscale, with six items (one of which was added to this
instrument and is not original to the ESoPI), had n = 129 in W1 and n = 125 in W2. The
scores on this subscale had a mean of 4.9 in W1 and 4.6 in W2, with SD = 1.08 in W1
and SD = 1.44 in W2. The standardized Cronbach’s alpha coefficient was .70 in W1 and
.89 in W2, indicating a good degree of internal consistency in Phase 1 and a high degree
of internal consistency in Phase 2.
The social presence/parasocial interaction subscale, with seven items, had n = 129
in W1 and n = 123 in W2. The scores on this subscale had a mean of 5.2 in W1 and 4.8
in W2, with SD = 1.06 in W1 and SD = 1.47 in W2. The standardized Cronbach’s alpha
coefficient was .86 in W1 and .94 in W2, indicating a high degree of internal consistency
in Phase 1 and an excellent degree of internal consistency in Phase 2.
Table 6
Psychometrics for Presence Subscales in the Current Study
Subscale
W1 standardized
Cronbach’s alpha
W2 standardized
Cronbach’s alpha
Spatial presence .70 .89
Social presence/
parasocial interaction
.86 .94
Social presence -
passive interpersonal
.82 .88
Social presence -
active interpersonal
.83 .84
Engagement
(mental immersion)
.85 .86
Social richness .91 .95
PRESENCE AND ENGAGEMENT 108
The social presence - passive interpersonal subscale, with four items, had n = 132
in W1 and n = 125 in W2. The scores on this subscale had a mean of 5.0 in W1 and 4.9
in W2, with SD = 1.35 in W1 and SD = 1.46 in W2. The standardized Cronbach’s alpha
coefficient was .82 in W1 and .88 in W2, indicating a high degree of internal consistency
in both phases.
The social presence - active interpersonal subscale, with just three items, had n =
131 in W1 and n = 126 in W2. The scores on this subscale had a mean of 5.2 in W1 and
5.0 in W2, with SD = 1.33 in W1 and SD = 1.43 in W2. The standardized Cronbach’s
alpha coefficient was .83 in W1 and .84 in W2, indicating a high degree of internal
consistency in both phases.
The engagement or mental immersion subscale, with six items, had n = 130 in W1
and n = 123 in W2. The scores on this subscale had a mean of 5.2 in W1 and 4.8 in W2,
with SD = 1.02 in W1 and SD = 1.19 in W2. The standardized Cronbach’s alpha
coefficient was .85 in W1 and .86 in W2, indicating a high degree of internal consistency
in both phases.
The social richness subscale, with seven items, had n = 133 in W1 and n = 126 in
W2. The scores on this subscale had a mean of 5.3 in W1 and 4.9 in W2, with SD = 1.09
in W1 and SD = 1.44 in W2. The standardized Cronbach’s alpha coefficient was .91 in
W1 and .95 in W2, indicating an excellent degree of internal consistency in both phases.
PRESENCE AND ENGAGEMENT 109
Table 7
Summary Statistics for Sense of Presence
Scale n standardized α mean SD range
Spatial Presence W1 129 .70 4.9 1.08 6.0
Spatial Presence W2 125 .89 4.6 1.44 5.7
Social Presence/
Parasocial Interaction W1
129 .86 5.2 1.06 5.3
Social Presence/
Parasocial Interaction W2
123 .94 4.8 1.47 6.0
Social Presence -
Passive Interpersonal W1
132 .82 5.0 1.35 5.3
Social Presence -
Passive Interpersonal W2
125 .88 4.9 1.46 5.8
Social Presence -
Active Interpersonal W1
131 .83 5.2 1.33 5.3
Social Presence -
Active Interpersonal W2
126 .84 5.0 1.43 6.0
Engagement
(Mental Immersion) W1
130 .85 5.2 1.02 4.8
Engagement
(Mental Immersion) W2
123 .86 4.8 1.19 5.0
Social Richness W1 133 .91 5.3 1.09 5.4
Social Richness W2 126 .95 4.9 1.44 6.0
Overall Sense of
Presence W1
122 .95 5.1 0.88 5.1
Overall Sense of
Presence W2
117 .97 4.8 1.20 4.9
PRESENCE AND ENGAGEMENT 110
Analysis of Research Questions
Research Question 1: Does the degree of presence sensed by a student predict the
degree of learning engagement?
To understand the most fundamental research question, whether the degree of
presence sensed by a student predicts the degree of learning engagement, multiple
regression was chosen to examine the relationship between the component scales of
presence as the independent variables and learning engagement as the dependent variable.
Phase 1 was chosen as the occasion of measurement for this finding.
Regression analysis allows the researcher to determine the likelihood that changes
in one or more variables of interest (the independent variables) may predict the change in
another variable (the dependent variable). The regression provides an indication of the
relationship of the form:
Yi = β0 + β1x1 + β2x2 + … + 𝜀
where Yi is the predicted value of the dependent variable for a future sample, β0 indicates
the intercept of the axis, βixi represents the slope coefficient of the regression line for
each variable (describing the relationship) and 𝜀 represents measurement error.
Using the six presence subscales as independent variables and the summary value
for learning engagement as the dependent variable, Pearson product-moment correlation
coefficient indicated a statistically significant (p < .005) correlation for every subscale
with learning engagement, with 42% of the variance explained by the six subscales of
sense of presence (adjusted r
2
= .42). ANOVA indicates significance (p < .005) and
thus the null hypothesis is rejected. Detailed output is provided in Table 8 and 9.
PRESENCE AND ENGAGEMENT 111
Table 8
Selected Output of Research Question 1 Multiple Regression Analysis
Model Summary
b
r r
2
Adjusted r
2
Std. Error
of the
Estimate
.669
a
.447 .420 .41954
ANOVA
b
Model
Sum of
Squares df
Mean
Square F Sig.
Regression 16.960 6 2.827 16.059 < .001
a
*
Residual 20.946 119 .176
Total 37.906 125
Coefficients
b
Model
Standardized
Beta t Sig. 95% CI for ß
(Constant) 9.817 < .001* [1.737, 2.615]
Spatial Presence W1 .089 .887 .377 [-.056, .146]
Social Presence/
Parasocial Interaction
W1
-.064 -.581 .562 [-.147, .081]
Social Presence -
Passive Interpersonal
W1
-.046 -.586 .559 [-.082, .045]
Social Presence -
Active Interpersonal
W1
.187 1.985 .045** [.000, .155]
Engagement (Mental
Immersion) W1
.457 3.923 < .001* [.122, .372]
Social Richness W1 .098 .865 .389 [-.064, .164]
Note. CI = Confidence Interval. a. Predictors: (Constant), SoP-Social Richness W1,
SoP-Social Presence - Passive Interpersonal W1, SoP-Spatial Presence W1, SoP-
Social Presence - Active Interpersonal W1, SoP-Social Presence/Parasocial
Interaction W1, SoP-Engagement (Mental Immersion) W1. b. Dependent Variable:
Learning Engagement Summary W1. * p < .001 ** p < .05
(continued)
PRESENCE AND ENGAGEMENT 112
Table 8
Selected Output of Research Question 1 Multiple Regression Analysis
(continued)
Coefficients
a
(continued)
Correlations Collinearity Statistics
Model Zero-order Partial Part Tolerance VIF
(Constant)
Spatial Presence W1 .422 .081 .060 .464 2.154
Social Presence/
Parasocial Interaction
W1
.421 -.053 -.040 .379 2.639
Social Presence -
Passive Interpersonal
W1
.236 -.054 -.040 .747 1.339
Social Presence -
Active Interpersonal
W1
.524 .179 .135 .521 1.919
Engagement (Mental
Immersion) W1
.646 .338 .267 .342 2.921
Social Richness W1 .555 .079 .059 .358 2.790
a. Dependent Variable: Learning Engagement Summary W1.
PRESENCE AND ENGAGEMENT 113
Correlations
LE_Sum SoP_Spat SoP_ParaInt SoP_SocPas SoP_SocAct SoP_MI SoP_Rich
Pearson
Correlation LE_Sum 1.000 .422 .421 .236 .524 .646 .555
SoP_Spat .422 1.000 .706 .419 .461 .563 .549
SoP_ParaInt .421 .706 1.000 .474 .574 .604 .621
SoP_SocPas .236 .419 .474 1.000 .374 .371 .369
SoP_SocAct .524 .461 .574 .374 1.000 .637 .603
SoP_MI .646 .563 .604 .371 .637 1.000 .770
SoP_Rich .555 .549 .621 .369 .603 .770 1.000
Sig. (1-tailed) LE_Sum .000 .000 .003 .000 .000 .000
SoP_Spat .000 .000 .000 .000 .000 .000
SoP_ParaInt .000 .000 .000 .000 .000 .000
SoP_SocPas .003 .000 .000 .000 .000 .000
SoP_SocAct .000 .000 .000 .000 .000 .000
SoP_MI .000 .000 .000 .000 .000 .000
SoP_Rich .000 .000 .000 .000 .000 .000
n LE_Sum 136 127 127 130 129 128 131
SoP_Spat 127 129 126 127 127 127 128
SoP_ParaInt 127 126 129 127 127 126 128
SoP_SocPas 130 127 127 132 130 129 132
SoP_SocAct 129 127 127 130 131 129 131
SoP_MI 128 127 126 129 129 130 130
SoP_Rich 131 128 128 132 131 130 133
Note: LE_Sum = Learning Engagement Summary W1, SoP_Spat = SoP-Spatial Presence W1, SoP_ParaInt = SoP-Social Presence/Parasocial Interaction W1,
SoP_SocPas = SoP-Social Presence - Passive Interpersonal W1, SoP_SocAct = SoP-Social Presence - Active Interpersonal W1, SoP_MI = SoP-Engagement (Mental
Immersion) W1, SoP_Rich = SoP-Social Richness W1.
Table 9
Research Question 1 Correlations
PRESENCE AND ENGAGEMENT 114
Examining the beta values for each factor, only two presence subscales made a
statistically significant unique contribution to the model, when controlling for the
variance explained by all other variables in the model. The first is engagement or mental
immersion (standardized beta = .46, p < .001). There is a degree of face validity to this
finding, since mental immersion itself is a form of engagement with the present
environment, though learning engagement is a significantly broader concept than mental
immersion alone. The other subscale that made a statistically significant unique
contribution at the p < .05 level is the social presence - active interpersonal scale
(standardized beta = .19, p = .049). None of the other beta values were significant,
indicating that there may be a high degree of overlap in the effects of the other subscales.
The part correlation coefficients indicate that approximately 7% of the variance in
learning engagement is uniquely explained by engagement / mental immersion. Nearly
2% is uniquely explained by the social presence - active interpersonal scale, while each
of the other independent variables uniquely explain less than 1% of the variance in
learning engagement.
Research Question 1a: Does the degree of presence sensed by a student predict the
degree of learning engagement when controlling for the type of classroom
environment (on-campus versus distance learning)?
To answer this research question, ordinary least squares regression was used to
examine the data. The learning engagement summary score was used for the dependent
variable, and the sense of presence summary score was used as an independent variable.
A dummy variable for environment was calculated to be used in the regression, where the
PRESENCE AND ENGAGEMENT 115
on-campus classroom was assigned the value of 0 and the online VLE was assigned the
value of 1. This dummy variable was used for a categorical linear regression analysis.
Inputting the environment variable followed by the summary sense of presence
score, ANOVA indicates that the relationship to learning engagement is significant
(p < .001). The Pearson product-moment correlation coefficient indicated a statistically
significant (p < .005) correlation for both environment (p = .002) and sense of presence
(p < .001), with 40% of the variance in learning engagement explained (adjusted r
2
=
.40). The Part correlation coefficients indicated that sense of presence uniquely explains
approximately 34% of the variance, while environment uniquely explains approximately
7% of the variance.
To investigate the potential of an interaction effect, the regression was also run
with both of the previous variables and an interaction variable computed by multiplying
sense of presence by location. This enabled a calculation of the significance of the
difference between the slope of the regression of sense of presence on learning
engagement with only on-campus cases, and the slope of the regression of the sense of
presence on learning engagement with only the online cases. The results indicate that the
interaction between sense of presence and location is not significant (difference in slope
= 0.108, p = .236).
Research Question 2: Does the sense of presence for learners attending distance
education and on-campus classes differ when measured by the Early Sense of
Presence Inventory?
This question was addressed by examining the relationship between a two-level
independent variable, location, and the continuous dependent variable sense of presence
PRESENCE AND ENGAGEMENT 116
as represented by the summary sense of presence score (all items on the instrument from
the ESoPI). An independent-samples t-test was performed and indicated that in Phase 1
(when both locations included text chat), there was no significant difference in scores
between on-campus (M = 5.1, SD = 0.93) and online (M = 5.1, SD = 0.86);
t (120) = -0.110, p = .913 (two-tailed). The magnitude of the difference in the means
(mean difference = -0.02, 95% CI: -.36 to .33) was too small to be of interest (eta squared
< .001).
In Phase 2 (when both environments were without text chat), however, an
independent-samples t-test indicated that there was a significant difference in scores
between on-campus (M = 5.2, SD = 1.03) and online (M = 4.5, SD = 1.22); t (115) =
3.105, p = .002 (two-tailed). The magnitude of the difference in the means (mean
difference = 0.69, 95% CI: .25 to 1.13) was moderate (eta squared = .077).
Because neither Phase 1 nor Phase 2 represented the typical configurations of
both classrooms, a third model was also tested, using the Phase 2 summary score for on-
campus (no text chat) and the Phase 1 summary score for online (with text chat), to
compare the two environments in their most typical state. An independent-samples t-test
of this model (referred to as Model 3 in the data set) revealed no significant difference in
the scores for on-campus without text chat (M = 5.2, SD = 1.03) and online with text chat
(M = 5.1, SD = 0.86); t (126) = 0.641, p = .522 (two-tailed). The magnitude of the
difference in the means (mean difference = 0.11, 95% CI: -.23 to .45) was very small (eta
squared = .003).
PRESENCE AND ENGAGEMENT 117
Research Question 2a: How does the sense of presence for learners attending
distance education and on-campus classes differ when measured by the Early Sense
of Presence Inventory, accounting for instructor and content?
The technology undergirding the virtual learning environment often limits the
number of people that can participate in most synchronous VLE sessions. The VLE
system in use during the present study is reported by faculty to support a maximum of
about 15-18 students simultaneously. Therefore, courses are typically divided into many
small class sections with different instructors in different sections to reduce the workload
on any one faculty member. This makes controlling for instructor and course content
while obtaining appropriate sample sizes extremely challenging. This is further
complicated by the low number of courses taught by the same instructor in an on-campus
setting as online. In the present study, only three instructors taught in both environments
with the same material covered in both environments. The same analysis conducted for
research question 2 was repeated with each of these more homogeneous groups to answer
research question 2a.
For instructor 1, independent-samples t-tests were conducted for Phase 1, Phase 2,
and Model 3, and the results are presented in Table 10.
PRESENCE AND ENGAGEMENT 118
The results indicate that none of the three comparisons represent a significant
difference (p < .05) in scores.
For instructor 2, independent-samples t-tests were conducted for Phase 1, Phase 2,
and Model 3, and the results are presented in Table 11:
Table 10
Results of Independent-Samples T-Tests for Instructor 1
Location n M SD t (df) p (two-tailed)
On-campus W1 7 5.5 0.89 .788 (15) .443
Online W1 10 5.1 0.88
On-campus W2 8 5.3 0.98 1.679 (14) .115
Online W2 8 4.5 1.04
On-campus Model 3 8 5.3 0.98 .387 (16) .704
Online Model 3 10 5.1 0.88
Table 11
Results of Independent-Samples T-Tests for Instructor 2
Location n M SD t (df) p (two-tailed)
On-campus W1 10 5.1 1.01 -.518 (12.711) .614
†
Online W1 16 5.3 0.62
On-campus W2 14 5.6 1.10 2.145 (29) .040*
Online W2 17 4.7 1.08
On-campus Model 3 14 5.6 1.10 .774 (19.880) .448
†
Online Model 3 16 5.3 0.62
* p < .05;
†
Equal variances not assumed
PRESENCE AND ENGAGEMENT 119
The results indicate that there is a significant difference (p < .05) between the
scores of the on-campus class section and the online class section in Phase 2: t (19) =
2.145; p = .040. Both of the other pairings show no significant difference.
For instructor 3, independent-samples t-tests were conducted for Phase 1, Phase 2,
and Model 3, and the results are presented in Table 12.
The results indicate that there is a significant difference (p < .05) between the
scores of the on-campus class section and the online class section in Phase 2: t (11.878) =
3.965; p = .002. Both of the other pairings show no significant difference.
In the analysis of all of the instructors, the extremely small sample sizes make the
results much less meaningful.
Table 12
Results of Independent-Samples T-Tests for Instructor 3
Location n M SD t (df) p (two-tailed)
On-campus W1 20 5.0 0.88 .025 (24) .980
Online W1 6 5.0 0.99
On-campus W2 21 5.0 0.98 3.965 (11.878) .002*
†
Online W2 4 3.8 0.39
On-campus Model 3 21 5.0 0.98 .062 (25) .951
Online Model 3 6 5.0 0.99
* p < .01;
†
Equal variances not assumed
PRESENCE AND ENGAGEMENT 120
Research Question 3: Does on-campus learners’ sense of presence differ under
different classroom configurations when accounting for instructor and content?
To answer this question, a paired-samples t-test was conducted to compare the
sense of presence scores for the on-campus samples in Phase 1 (the experimental
condition—with text chat) with Phase 2 (the control condition—without text chat).
While the mean summary score for sense of presence decreased in the experimental
condition (with text chat) (M = 5.09, SD = 0.94) from the control condition (without text)
(M = 5.14, SD = 0.96), this increase was not statistically significant: t (36) = -.368,
p = .715 (two-tailed).
Research Question 4: Does distance education learners’ sense of presence differ
under different screen interface configurations when accounting for instructor and
content?
To answer this question, a paired-samples t-test was conducted to compare the
sense of presence scores for the online samples in Phase 1 (the control condition—with
text chat) with Phase 2 (the experimental condition—without text chat). There was a
statistically significant (p < .001) decrease in the sense of presence score between the
control condition (with text chat) (M = 5.04, SD = 0.82) and the experimental condition
(without text chat) (M = 4.50, SD = 1.15): t (65) = 4.739, p < .001 (two-tailed). The
mean decrease in SoP scores was 0.54 with a 95% confidence interval ranging from 0.31
to 0.77. The effect size was large (eta squared = 0.257).
PRESENCE AND ENGAGEMENT 121
Structural Equation Modeling
Introduction
Structural equation modeling, or SEM, is an advanced statistical methodology for
the multivariate causal analysis of a theory, and particularly useful in exploring variables,
such as sense of presence and learning engagement, that cannot be directly observed. At
its most basic, structural equation modeling is an expansion of multiple regression and
analysis of variance techniques (Bentler, 1980; Lei & Wu, 2007). It takes a hypothesis-
testing approach, confirming a theory that is described in advance of analyzing the data,
rather than the exploratory approach of traditional factor analysis (Byrne, 1994, 2010).
SEM is grounded in two key qualities: “(a) that the causal processes under study are
represented by a series of structural (i.e., regression) equations, and (b) that these
structural relations can be modeled pictorially to enable a clearer conceptualization of the
theory under study” (Byrne, 1994, p. 3). In structural equation modeling, the researcher
constructs a hypothetical model describing the causal interactions between observed
variables and latent (unobserved) constructs. The data from the observed variables is
then tested to determine how accurately the model describes the interrelationships of the
elements—the model’s ‘goodness of fit’ (Byrne, 1994).
The confirmatory nature of structural equation modeling sets it apart from other
multivariate analyses, which are usually descriptive in nature, and less well-suited to
hypothesis testing (Byrne, 2010). SEM makes possible a broad statistical conception of a
theory in action that is difficult or impossible using other techniques. A structural
equation model incorporates not only observed (or manifest) variables, but also the
PRESENCE AND ENGAGEMENT 122
unobserved (or unobservable) latent factors
1
that may play key roles in a related system
of variables. In this way, structural equation modeling provides one of the most
convenient ways to describe the latent structure that underlies observed behavior (Byrne,
2010).
Path analysis and confirmatory factor analysis are both techniques related to
structural equation modeling. Path analysis predates SEM, and could be considered its
forebear. SEM builds on path analysis’s use of graphical analysis of hypothetical
relationships. In path analysis, however, latent variables and error terms are generally not
represented (Lei & Wu, 2007). Confirmatory factor analysis could be considered a
building block of SEM. In confirmatory factor analysis, a group of observations are
tested to confirm that a single latent factor influences their value. Structural equation
modeling expands that exploration by connecting latent factors to each other (and
sometimes observed variables to each other or to multiple factors) and testing more than
one factor at a time, in a unified system of multivariate relationships (Byrne, 1994, 2010).
SEM uses a graphic notation to visualize the interrelationships between variables
and factors, making it particularly accessible and intuitive. In a structural equation
model, latent factors are represented by ovals or circles, and observed variables are
represented by rectangles or squares. Arrows are drawn between each related latent
factor and observed variable, with the direction of the arrow representing one factor or
variable’s influence on another. This is can be confusing, as it is not necessarily the
direction of some other visualization schemes such as path analysis for CFA. In SEM,
the arrows can be thought of as a notation that one factor ‘causes’ change in another
1
Latent variables are generally termed factors in SEM.
PRESENCE AND ENGAGEMENT 123
(Byrne, 2010). Structural equation modeling also explicitly accounts for error,
representing residual factors as circles (since error is, after all, an unobserved variable),
thus, error is appropriately shown as another latent influence on both observed and non-
observed variables. Every relationship in the model can be represented by a basic
structural equation (regression) of the form:
Variable = Factor(s) + Error
Elements in a structural model can serve as both dependent and independent
variables, depending on the other variables being examined. Each arrow points from an
independent variable to the dependent variable it influences. However, the independent
variable may simultaneously serve as a dependent variable in another relationship. When
factors are fully independent variables, they are termed exogenous, and when they are
dependent in any relationship, they are considered endogenous. When it occurs,
structural models also represent covariance between variables/factors as double-headed
arrows.
Given the large number of configurations possible to specify in SEM, it is
important to consider whether a model is identified, or defined in such a way that there is
an appropriate ratio of unknown elements to known elements and therefore, a unique
solution for the values of the parameters can be found (Bentler, 1980, 1988). If too many
elements are unknown, there is no mathematical way to verify the model, because
parameters would then be subject to arbitrariness, and more than one solution would be
possible. The number of data points sampled (also called “sample moments”) is the
number of unique elements in the sample covariance matrix and any sample means. For
k observed variables, this number is k(k+1)/2 + l sample means (Arbuckle, 2013; Byrne,
PRESENCE AND ENGAGEMENT 124
2010). When the number of unique parameters to be estimated (also called free
parameters) exceeds the number of data points sampled, the model is under-identified. If
the number of unique parameters to be estimated is exactly the same as the number of
data points, the model is considered just-identified. In this case, it means that the model
is a perfect fit, but this is actually unfavorable, because such a model has no room for
error and is scientifically uninteresting because it is unable to be rejected (Bentler, 1980;
Byrne, 2010; Lei & Wu, 2007). The desired specification for a model is for it to be over-
identified, where the number of unique parameters to be estimated is less than the number
of data points. If a model is just-identified or under-identified, the results are not
interpretable or reportable and the model must be respecified (Bentler, 1980).
The difference between the number of data points available and the number of
unique parameters to be estimated is known as the model’s degrees of freedom. Thus, a
testable, over-identified model has positive degrees of freedom, whereas a just-identified
model has 0 degrees of freedom and an under-identified model has negative degrees of
freedom.
Preliminary Model
In the current study, the AMOS software package, version 22.0.0 (Arbuckle,
2013), was used to create and analyze a hypothetical structural model for the relationship
between sense of presence and learning engagement (both latent constructs). A
preliminary model was crafted by applying existing theory in the literature. This model
was then tested for goodness of fit using the data gathered as a part of the study.
Subsequently, the model was strategically modified until a model with the best possible
fit was discovered.
PRESENCE AND ENGAGEMENT 125
The preliminary model was created to represent the primary hypothesis for the
study, which follows from existing presence and learning engagement theory in the
literature: the sense of presence is an influence on learning engagement. To describe this
concept graphically in its simplest form, sense of presence is represented as an oval, and
learning engagement as an oval as well, with an arrow pointing from sense of presence to
learning engagement (Figure 3):
Because sense of presence is not hypothesized to be the only influence on
learning engagement, and SEM explicitly accounts for error, the residual (sometimes
called disturbance for latent variables) is added as another circle influencing learning
engagement (Figure 4):
Figure 3. Concept representation in its most basic form.
Figure 4. Concept representation with disturbance added.
PRESENCE AND ENGAGEMENT 126
Error terms are frequently labeled with a single letter and number, typically with
an e for error factors on observed variables and a d or r for the disturbance or residual on
latent factors, so here, it is labeled as D1.
Following from this most basic relationship, the observed variables are added as
rectangles, with arrows to indicate that they are dependent variables (i.e., the latent
variables ‘cause’ these observations to register as a particular value). In the present
study, each of the two instruments utilized (ESoPI for sense of presence and SCEQ for
learning engagement) comprise multiple subscales, each of which is designed to examine
a particular latent concept that is itself a component of the two primary latent factors,
sense of presence and learning engagement. Because of this, the most thorough model
would include a second level of latent variables, each influenced by their overall concept
and influencing the observed scores on each survey item associated with that concept.
This would, however, introduce enough statistical complexity to the model that a sample
as small as the one in the present study would be inadequate to accurately test it. In order
to facilitate realistic testing of the model, the summary score from each subscale was
therefore accepted as an observed variable in and of itself. In the interest of parsimony,
the location variable was not added to the model, since the preliminary analysis showed
no significant influence on sense of presence.
These steps resulted in the preliminary hypothesized model, shown in Figure 5.
PRESENCE AND ENGAGEMENT 127
This model is recursive, and has 23 variables (including 11 residual terms, 2 latent
factors, and 10 observed variables). In the model, 12 elements are exogenous—
functioning as independent factors; 11 are endogenous—functioning, at least in part, as
dependent variables. With the relationships shown in Figure 5 as “1.0” fixed at 1.0, there
are 31 distinct parameters to be estimated (the values of each undefined regression and
each variable and factor, including residuals). Given 65 sample data points (sample
moments)—calculated as 10(10+1)/2 unique elements in the covariance matrix + 10
means—the model has 34 degrees of freedom (65-31=34). This indicates that the model
is sufficiently over-identified.
The next step, therefore, is the hypothesis testing of the model. In general, a test
statistic is generated to reflect the magnitude of the discrepancy between the covariance
matrix of the actual observed data and the covariance matrix that is described by the
proposed structural model. There exists, however, a variety of different test statistics,
Figure 5. Preliminary hypothesized model.
PRESENCE AND ENGAGEMENT 128
and an ongoing debate among researchers as to the best methods and statistics to use in
testing (Bollen & Long, 1992; Hu & Bentler, 1999; Lei & Wu, 2007).
The most basic test of goodness of fit, the chi-square (𝜒
2
) likelihood ratio test,
compares the model as a whole to the alternative that the observed variables are “simply
correlated to an arbitrary extent” (Bentler, 1980, p. 428). The chi-square test considers
the null hypothesis Σ = Σ(𝛩) where Σ represents the covariance matrix of observed
variables, Σ(𝛩) is the proposed model’s covariance matrix, and 𝛩 represents the number
of free parameters to be estimated in the model. It follows, then, that the null hypothesis
could also be expressed as Σ - Σ(𝛩) = 0, implying that the covariances match exactly,
leaving no error in Σ - Σ(𝛩) (Bollen, 1989; Bollen & Long, 1992; Byrne, 2010). In this
way, the 𝜒
2
statistic reflects the discrepancy between the hypothesized model and the true
relationships observed, and the associated probability statistic represents the likelihood
that the model represents an exact reproduction of the data and relationships observed.
Because an exact match is the null hypothesis, the higher the p value of the chi-square
test, the closer the proposed model is to a perfect fit (Byrne, 2010).
When the 𝜒
2
statistic is large compared to the degrees of freedom, it indicates a
poor fit (Bentler, 1980). Because of this, it is customary to report the ratio of 𝜒
2
to df (the
degrees of freedom). Likewise, if the test of statistical significance is passed (p < .05),
the model is generally rejected, as it means that the likelihood of an exact match is too
low to be of use (Bollen & Long, 1992).
The chi-square test for this preliminary hypothesized model shows that the model
has poor fit. For this preliminary model, 𝜒
2
= 112.78, p < .001, 𝜒
2
/df = 3.32. Since the
PRESENCE AND ENGAGEMENT 129
ratio of 𝜒
2
to df is quite large, and the statistic is significant, this model should be
rejected.
The chi-square test, while still the primary test for goodness of fit, often shows an
inappropriate sensitivity with larger sample sizes (Hu & Bentler, 1999; Lei & Wu, 2007).
While researchers such as Kenny and McCoach (2003) have also expressed misgivings
about the accuracy of the figure with small sample size and low numbers of variables, the
chi-square test may be a more reasonable measure for a small sample size and low model
complexity such as the current study. Because larger sample sizes (generally more than
200 samples) are preferred for structural model testing, and chi-square begins to become
unreliable with larger sample sizes, a variety of other goodness of fit standards have been
developed (Hu & Bentler, 1999; Kenny & McCoach, 2003). Debate about which
measures are best abounds, but Hu and Bentler (1999) and Bollen and Long (Bollen &
Long, 1992, 1993) recommend reporting two statistics to combat potential Type I error.
One useful and popular alternative test is the root mean squared error of
approximation (RMSEA), an absolute fit index (Browne & Cudeck, 1993; Hu & Bentler,
1999; Kenny & McCoach, 2003; Steiger & Lind, 1980). RMSEA takes into account the
complexity of the model, has a known distribution, and is non centrality based (Hu &
Bentler, 1999). The RMSEA statistic reflects a good fit when it has a very low value,
generally between 0 and 0.1. MacCallum, Browne, and Sugawara (1996) recommend the
cutoff values of .01 for excellent fit, .05 for good fit, and .08 as mediocre fit. AMOS can
also calculate a 90% confidence interval for the RMSEA statistic. Paired with RMSEA is
p of close fit, or pClose: if this statistic is not significant (p > .05), the model fit is
considered to be ‘close.’
PRESENCE AND ENGAGEMENT 130
The RMSEA value for this model is .126 (default model), with a .100-.152 90%
CI, and pClose < .001, indicating a poor fit, which is not surprising given the 𝜒
2
value for
this model.
Model Revisions
Given the poor fit of the initial model, additional refinements were made and
tested. Removing some of the subscales on the two instruments that had the lowest
Cronbach’s alpha values produced the following model, Model A, which passes standard
reference points for goodness of fit on the chi-square and RMSEA tests (Figure 6):
For this model, 𝜒
2
= 9.21, df = 8, 𝜒
2
/df = 1.15, p = .325, indicating a good fit
(Bentler, 1980; Hu & Bentler, 1999). This model also meets the ‘close’ standard with
RMSEA: RMSEA = .032 (.000-.105 90% CI), pClose = .574, indicating a good fit (Hu &
Bentler, 1999; MacCallum et al., 1996).
Detailed results are presented below in Table 13 and Table 14.
Figure 6. Structural Equation Model A.
PRESENCE AND ENGAGEMENT 131
Table 14
Selected AMOS Output for Model A: Goodness-of-Fit Statistics
CMIN (Chi-square)
Model NPAR CMIN df p CMIN/df
Default Model 19 9.213 8 0.325 1.152
Saturated Model 27 0.000 0
Independence Model 6 392.553 21 0.000 18.693
RMSEA
Model RMSEA LO 90 HI 90 pClose
Default Model .032 .000 .105 .574
Independence Model .347 .317 .377 .000
Note: NPAR = Number of parameters to be estimated; CMIN = minimum discrepancy (𝜒
2
);
RMSEA = root mean squared error of approximation
Table 13
Selected AMOS Output for Model A: Summary Notes
Number of distinct sample moments: 27
Number of distinct parameters to be estimated: 19
Degrees of freedom (27 - 19): 8
Result
Minimum was achieved.
Chi-square = 9.213
Degrees of freedom = 8
Probability level = .325
Variables
Number of variables: 15
Number of observed variables: 6
Number of unobserved variables: 9
Number of exogenous variables: 8
Number of endogenous variables: 7
PRESENCE AND ENGAGEMENT 132
Model A expresses sense of presence as a latent factor that causes measurable
results on four particular subscales of the ESoPI: spatial presence, social presence - active
interpersonal, engagement (mental immersion), and social richness. Sense of presence +
error causes learning engagement, another latent construct which causes results on two
particular subscales of the SCEQ: skills engagement and emotional engagement. This
model is a good fit for the data gathered in the current study, and suggests these subscales
are the most useful in measuring the two latent constructs.
A second model, Model B, was created by removing the spatial presence score.
This model was also tested and found to be a close fit (Figure 7):
For this model, 𝜒
2
= 4.12, df = 4, 𝜒
2
/df = 1.03, p = .391, also indicating a good fit
(Bentler, 1980; Hu & Bentler, 1999). This model performs even better than the previous
model with the RMSEA test: RMSEA = .014 (.000-.126 90% CI), pClose = .567,
indicating a good fit (Hu & Bentler, 1999; MacCallum et al., 1996).
Details of the output from AMOS are provided in Table 15 and Table 16 below:
Figure 7. Structural Equation Model B.
PRESENCE AND ENGAGEMENT 133
Table 16
Selected AMOS Output for Model B: Goodness-of-Fit Statistics
CMIN (Chi-square)
Model NPAR CMIN df p CMIN/df
Default Model 16 4.115 4 .391 1.029
Saturated Model 20 .000 0
Independence Model 5 333.957 15 .000 22.264
RMSEA
Model RMSEA LO 90 HI 90 pClose
Default Model .014 .000 .126 .567
Independence Model .380 .345 .416 .000
Note: NPAR = Number of parameters to be estimated; CMIN = minimum discrepancy (𝜒
2
);
RMSEA = root mean squared error of approximation
Table 15
Selected AMOS Output for Model B: Summary Notes
Number of distinct sample moments: 20
Number of distinct parameters to be estimated: 16
Degrees of freedom (20 - 16): 4
Result
Minimum was achieved.
Chi-square = 4.115
Degrees of freedom = 4
Probability level = .391
Variables
Number of variables: 13
Number of observed variables: 5
Number of unobserved variables: 8
Number of exogenous variables: 7
Number of endogenous variables: 6
PRESENCE AND ENGAGEMENT 134
While Model B represents a slightly closer fit than Model A, both are likely, and
are presented here, as the fuller Model A provides more direction for possible practice,
while Model B presents a more parsimonious description of the behavior, though its
improvement is slight. A more thorough exploration of the implications of these two
structural equation models is presented in the following chapter.
Additional Survey Information
During Phase 2 of the procedure, participants in the online class sections had a
particularly strong reaction to the test condition, removing the text chat function from the
VLE. The participants’ displeasure prompted the creation of a follow-up qualitative
survey, designed to garner additional information and color about the students’ views on
the text chat element of the VLE. Participation was extremely limited (n = 12),
composed of volunteers from the online class sections of the main study.
Because of the very low sample size, the results must be treated as anecdotal.
Nevertheless, selected relevant findings are presented here, and these findings both
support the results of structural equation modeling and other statistical tests, as well as
offer insight into the potential causes of the observed behavior.
The follow-up survey asked participants 11 questions, 4 of which were free text
responses, and the rest 5-point Likert scales asking participants to rate their agreement
with various statements about the value and usage of text chat in the online learning
environment. Because of the small sample size, all responses were treated as qualitative
responses, and no overall statistics were calculated.
Several common themes emerged in the responses, and they are discussed in this
section. These themes included initial hesitation and disappointment at the lack of text
PRESENCE AND ENGAGEMENT 135
chat availability, and an overall negative view of class without text chat. They responded
that the text chat provided for a ‘side channel’ with greater immediacy, and reported
many themes congruent with mental immersion and social richness presence concepts.
Participants also commented on text chat’s utility as a technological backup and
recordkeeping device. Speculation reported by participants about the role of age in the
results and the possibilities of text chat in an on-campus setting are also presented below
and discussed at greater length in Chapter 5.
Overall Reactions
Students reported some trepidation immediately on discovery of the text chat’s
absence, with seven marking “slightly negative” and another three “substantially
negative” as their initial reaction on learning that the class would be held without text
chat. Nearly all of the respondents agreed (3) or strongly agreed (8) with the statement,
“I feel that class was more difficult without the text chat box.” Surprisingly, students
noted that the absence of the text chat increased, rather than decreased their level of
distraction.
Text Chat as a Second Channel
Participants cited text chat as a favored element that established a multichannel
experience. “The side channel is valuable to asking the small questions that relate to the
task at hand, without interrupting the flow of the speaker's thoughts,” explained
Respondent H. Respondent B commented, “Listening to one person talk at a time is
really inefficient.…Without the chat, I feel more likely to ‘tune out’ once I recognize
someone is on a rant, because I know there is no way for me to have a chance to engage
PRESENCE AND ENGAGEMENT 136
any time soon.” Immediacy was frequently cited as a desirable advantage of text chat:
“the chat pod allows for instant input/support/questioning” (Respondent E).
Several students noted that they felt more comfortable expressing themselves in
the text chat, as they perceived it to be a less intrusive or less public venue: “Sometimes I
don't want to interrupt the professor with what I perceive as menial thoughts on the
subject so I usually type it in the chat” (Respondent K). Respondent I noted, “the text
chat gives an option to validate someone’s verbal comments/sharing without obnoxiously
interrupting.” Respondent C commented, “I like the immediacy of being able to respond
to classmates and affirm their comments, give feedback and such without interrupting the
whole flow of the discussion.”
Text Chat as Presence Enrichment
Students frequently found text chat to be an enrichment of the online learning
environment in a way that suggests presence in the form of mental immersion and social
richness, as well as learning engagement, offering responses such as respondent H, who
comments, “text chat keeps the wheels greased; the text chat allows for the small details
that put the lesson in context. It allows for more engagement in the lesson.”
Respondent C also highlighted the fact that text chat prompted more active mental
engagement: “[It] keeps my attention during longer all class discussions.” Respondent F
cited the richness of the “free flow of ideas and comments” as a positive result of the text
chat.
Respondent B highlighted the unique quality of text communication as “persistent
whereas speech is not, making it easier to digest information. Chat can be digested at the
reader’s pace, whereas speech must be processed or lost at whatever speed it is given.”
PRESENCE AND ENGAGEMENT 137
This student noted that while the VLE provides a method for students to open one-to-one
audiovisual chats between students, doing so creates an “embarrassing” alert message
that is shown to the entire class. “One professor even explicitly asked students to stop
because of the disruptive nature of the notifications.”
There was strong evidence that even though the text chat would be seen as an
artificial mediation of the person-to-person interactions, its function in the VLE actually
increased presence: “Without the chat box, I felt less connected with the subject being
presented” (Respondent K). Respondent H suggested that text chat helped to compensate
for the physical presence lost in the online environment: “As this is an online class, we
miss out on the informal opportunities to talk and share while waiting for a professor to
unlock a door, or bump into each other on a physical campus.”
Text Chat as a Utility
Text chat was also seen as an important backup for the audiovisual interface. One
participant expressed frustration that when her camera malfunctioned and she had no easy
way to communicate that fact to the rest of the class. It was also seen as a tool for
communal thinking and record-keeping by Respondent H:
Text chat can serve as a communal journal. Often as I am listening in on a
discussion; I am reminded of a link to a website, or some other resource that I find
particularly valuable and either would like to share, or review at a later time. Text
chat is the perfect medium for this task.
Age and Text Chat
Respondent I singled out age as a potential confound for the function of text chat
in an online learning environment: “The professors that seem to get annoyed with the
chat also seem to not be able to ‘keep up’ as quickly with technology or are older (this
sounds extremely ageist but is just based on experience). Many of us are able to keep up
PRESENCE AND ENGAGEMENT 138
with the chat as well as the class discussion and use both to our advantage.” While the
limited sample size and age range of the main data set makes testing this premise
difficult, there was no statistically significant effect when age was regressed with sense of
presence (p = .612) in Phase 1 but there was a correlation at the p < .5 level in Phase 2
(p = .071; adjusted r
2
= .02).
While the evidence may not be clear that age has an influence on presence, the
students’ interest in having the text chat available as a part of a multichannel sensory
experience, as well as their facility with such an environment might indeed be
generational. The mean and mode of age in the study suggest that most participants were
a part of the ‘millennial’ generation that Schwalbe (2009) says, “might well be reading e-
mail while holding a cell phone to one ear and listening to an MP3 player with the other”
(p. 53). In Schwalbe’s study, 81% of the 18- to 24-year-old college students used text
messaging on their phones and other devices “frequently or occasionally,” with email and
instant messaging on the computer at 95%. Schwalbe also found the same desire for
instant feedback as seen in the present study (Schwalbe, 2009, p. 60). Myers and
Sadaghiani (2010) also found in the growing academic literature and popular press
substantial evidence of this inclination toward multichannel, immediate communication
in this age group. Six of the 12 participants in the current survey reported that they used
text chat while they themselves were speaking.
Text Chat in an On-Campus Setting
When the online student participants were asked to speculate on the potential
utility of text chat in an in-person class, the respondents were mixed in their opinions.
While many felt that the same positive experience with text chat in the online
PRESENCE AND ENGAGEMENT 139
environment would translate to the on-campus classroom, several were concerned that in
that particular context it would be a distraction, responding that it would be valuable
“only if it is visible at the front of the room in a way that does not detract from a focus on
the speaker/faciliator[sic] at the front or from other classmates” (Respondent C).
Respondent E commented, “a back-channel in an in-person classroom MAY be more of a
distraction than support.” Respondent I’s comment highlights the already-strong sense of
presence in an in-person classroom, implying that text chat is important for building
presence in a VLE, but may be problematic in person: “In-person format is a richer
experience and I think the chat could be a little distracting there.”
Summary
Overall, the results of the analysis for the present study indicate that presence is
an influence on learning engagement, with 42% of the variance explained by the six
subscales of the ESoPI instrument. While the learning environment (online or in-person)
did have a statistically significant effect on learning engagement, its interaction with
presence was not statistically significant. Comparing the classroom environment to the
online environment, the sense of presence was only significantly different in Phase 2,
when the difference in the configuration of the VLE may have made the comparison
especially dramatic. When both environments were compared in their most typical state
(Model 3), no statistically significant difference in sense of presence was found. In
examining the changes that were made to each of the two environments, no significant
effect on the sense of presence was found in the on-campus classroom when text chat was
added, but a statistically significant change in presence was found in the online
environment when text chat was removed.
PRESENCE AND ENGAGEMENT 140
Through structural equation modeling, a better understanding of the interaction
between specific presence elements and learning engagement qualities was obtained
through hypothesis, testing, and revision. In the end, two models achieved reasonable
goodness of fit, and suggest a more specific understanding of the relationship between
not only sense of presence and learning engagement, but also the most effective
measurement scales of each of these latent concepts.
Qualitative data obtained from a follow-up survey of online class participants also
supports the finding that text chat makes a significant difference in the sense of presence
in that environment, and suggests avenues for further exploration of these effects.
A fuller discussion of these results, their implications, limitations, and
ramifications for future research is presented in the following chapter.
PRESENCE AND ENGAGEMENT 141
CHAPTER FIVE: DISCUSSION
The purpose of the present study was to examine the relationship between the
sense of presence and learning engagement in both the traditional on-campus classroom
and a synchronous, real-time online classroom environment. Research was conducted
and analyzed in three major ways: (a) through traditional statistical analysis in an effort to
answer the six primary research questions, (b) through structural equation modeling to
understand the overall interaction between observed variables and latent factors, and (c)
through a qualitative follow-up survey. In this chapter, further discussion of the results
and their implications is provided. General discussion and interpretation of each of these
analytical facets is provided, followed by a consideration of the study’s limitations. The
chapter then offers suggestions of this study’s implications for practice, and concludes
with recommendations for future research directions.
Internal Consistency
Before conducting broader statistical analysis, the internal consistency of the
instruments was examined through the use of factor analysis. In light of relevant theory,
an inspection of the questions on both the SCEQ (measuring learning engagement) and
the ESoPI (measuring sense of presence) indicates that both have face validity through a
logical adherence to our understanding of these two concepts and their components in the
literature. In both cases, the authors of the instruments also provided limited data on the
consistency of their scales from their own pilot studies with the instruments.
For the most accurate understanding of the internal consistency of the scales as
they pertain to this particular sample, however, factor analysis was used to inspect the
consistency of the instruments using the data obtained in the current study. The
PRESENCE AND ENGAGEMENT 142
Cronbach’s alpha values calculated for each subscale of each instrument indicate how
well the questions ‘hang together’ and describe the same latent construct.
Through these analyses, both measures appear internally consistent, and all
subscales of both measures are also consistent. This finding undergirds the other
statistical analyses, so its assessment and positive result increase the confidence in the
correctness of the study’s other findings.
Research Questions
Six research questions were posed at the beginning of the study. While the
statistical analysis for each question is included in the previous chapter, this section will
briefly discuss the interpretation and meaningfulness of the findings related to each
question.
Research Question 1
The central inquiry of the present study was posed in research question 1, namely,
does the degree of presence sensed by a student predict the degree of learning
engagement? The question was deliberately simple, and called for multiple regression of
the component sense of presence concepts (as independent variables) on learning
engagement (the dependent variable).
The results definitively show that the components of sense of presence, as
measured by the ESoPI, predict the degree of learning engagement, as measured by the
SCEQ, explaining 42% of the variance. This result reinforces the value of presence
research to educational psychology, and animates further examination of the ways in
which the sense of presence influences learning engagement.
PRESENCE AND ENGAGEMENT 143
While this finding represents a somewhat simplistic view of the relationship,
further understanding of the relationships at work was obtained through structural
equation modeling (discussed later in this chapter).
Research Question 1a
This research question, does the degree of presence sensed by a student predict
the degree of learning engagement when controlling for the type of classroom
environment (on-campus versus distance learning), follows on research question 1 to
determine whether the relationship in that question is affected by the location itself. The
results indicate that while the difference in venue does have a statistically significant
effect on the degree of learning engagement, there is no significant interaction between
presence and location—each has a unique effect on learning engagement, but the overlap
between the two is not significant. This interpretation is further supported by the findings
in the following two questions.
Research Question 2
This question sought to understand the influence of the learning environment
type, on-campus or online, on sense of presence. It asks: Does the sense of presence for
learners attending distance education and on-campus classes differ when measured by
the Early Sense of Presence Inventory? Overall, the findings indicated that the difference
between the on-campus environment and the online one does not affect the sense of
presence, bolstering the findings in research question 1a.
In Phase 1, when both environments made text chat available to participants, and
in Model 3, when both environments were in their ‘typical’ configuration (on-campus
without text chat and online with text chat), the difference in the mean scores for sense of
PRESENCE AND ENGAGEMENT 144
presence between the two environments was not statistically significant. This indicates
that the environment itself does not significantly influence the sense of presence—a
surprising outcome, given the intuitive expectation that presence would be highly
sensitive to the difference between the wholly-designed qualities of the online learning
environment and the minimally-mediated conditions of an on-campus classroom.
It is the opinion of the author that the statistically significant result in Phase 2,
where both environments were without text chat, should not be taken as refutation of this
conclusion, but instead suggests that the experimental condition (absence of text chat) in
the online environment was dramatic enough to highlight the difference in presence not
only between experiment and control condition, but also between online and on-campus.
Because the on-campus location compared in Phase 2 represented a typical condition,
whereas the online location in Phase 2 represented an atypical one, Model 3 was
developed to pair both typical conditions. The result for Model 3 matches the Phase 1
result, supporting the conclusion that location itself is not a significant influence on sense
of presence. The results suggest the possibility that the VLE in its standard configuration
is well-tuned to approximate the same level of presence as in an on-campus setting. The
result in Phase 2 is further examined in later discussion.
Research Question 2a
This question magnifies the specificity of research question 2, asking if the same
relationship between location and sense of presence is true when accounting for instructor
and content. As stated in the previous chapter, the difficulties of obtaining sample
conditions where instructor and content are adequately controlled for resulted in only
three subsamples appropriate to answer this question, each of which had a smaller-than-
PRESENCE AND ENGAGEMENT 145
desirable sample size. These three samples largely agree with the findings in question 2,
but their small sample sizes make for weak statistical power, and undermine their
usefulness.
It should be noted that when analyzing each set of data controlled for instructor
and content in Phase 1, Phase 2, and the typical-condition-paired Model 3, only two of
the nine comparisons showed statistical significance (Phase 2 with Instructor 2 and Phase
2 with Instructor 3), which is subject to the same caution regarding the test conditions as
research question 2. (Also worth noting: the significance in Phase 2 with Instructor 2 was
at the p < .05 level—a coarser alpha value than used elsewhere in the present study.)
Therefore, the best interpretation is that the relationship between the variables matches
that of the broader research question 2—there is no significant difference in sense of
presence between the two environments studied when controlling for instructor and class
content.
Research Question 3
Both this question, Does on-campus learners’ sense of presence differ under
different classroom configurations when accounting for instructor and content? and the
following one call for an examination of whether the experimental alterations to the class
environment affected students’ sense of presence.
In research question 3, addressing the on-campus sample, the outcome showed no
statistically significant difference in the sense of presence scores recorded in the typical
on-campus classroom condition (Phase 2) from those in the experimental on-campus
condition (Phase 1), where text chat was introduced. During the experimental procedure
in the on-campus location, it was noted that the students did not use the text chat that was
PRESENCE AND ENGAGEMENT 146
made available to a very large degree (especially as compared to the students in the
online sample). The low usage may indicate that the availability of text chat was not
noticeable enough to participants to meaningfully alter their experience from that of the
typical condition (Phase 2), without text chat. This phenomenon is further discussed in
the limitations section below.
It should be noted that while the instructor was the same in both phases for each
sample, the content was only controlled for in a broad sense—the subject of the course in
each occasion of measurement was the same, but the particular lesson was not. This
limitation is further discussed later in this chapter.
Research Question 4
The final research question represents the same inquiry as research question 3, but
in the online setting: Does distance education learners’ sense of presence differ under
different screen interface configurations when accounting for instructor and content? As
with the previous research question, while the instructor was the same in both phases for
each sample, the content was only controlled for in a broad sense—the subject of the
course in each occasion of measurement was the same, but the particular lesson was not.
In the online setting, the difference between the experimental and control
conditions was statistically significant, and unexpectedly negative. Sense of presence
scores in the online session without text chat (Phase 2) were significantly lower than
when text chat was available (Phase 1).
The literature (e.g., Lee, 2004; Lombard & Ditton, 1997) suggests that artificial
elements with limited realism in mimicking natural interactions, such as text chat, will
have a deleterious effect on social presence. Intuitively, one might expect that the most
PRESENCE AND ENGAGEMENT 147
natural social interaction would come from the most realistic representation of a face-to-
face conversation (Lombard & Ditton, 1997). Even the split-attention assumption within
CTML (Chandler & Sweller, 1991; Mayer & Moreno, 1998; Mousavi et al., 1995;
Reinwein, 2012) would suggest that mixing the modes between text and audiovisual
conversation might reduce engagement. The opposite effect was found, however: text
chat improves sense of presence in the VLE.
This is a particularly important and interesting result, also reflected in and refined
by the results of the qualitative responses to the follow-up survey. The follow-up survey
and SEM results (discussed below) seem to suggest that despite the intuitive importance
placed on social presence, text chat may be strengthening mental immersion and social
richness, and these factors may be substantially responsible for participants’ improved
overall sense of presence when text chat is available.
The implications of this finding for practice and future research will be discussed
at length later in this chapter.
Structural Equation Models
Structural equation modeling allows for statistical examination of an entire system
of relationships among both observed variables and latent factors. In the present study, a
hypothesized model of the relationship between the latent factors sense of presence and
learning engagement, and the observed scores on the SCEQ and ESoPI was constructed.
This initial model was tested for goodness of fit, and rejected based on its poor fit.
Further refinements were made to the model, removing observed subscale
variables with the least internal consistency (ordering them by their Cronbach’s alpha
values). This cycle of model construction, testing, refinement, and further testing
PRESENCE AND ENGAGEMENT 148
resulted in the construction of two structural equation models with adequate goodness of
fit.
These final two models (Model A and Model B) represent statistically sound
representations of the multivariate relationships at work between presence and learning
engagement.
The increased specificity of these models over the initially-hypothesized model
suggests that when attempting to bolster the sense of presence in the classroom, extra
emphasis should be placed on the following specific presence concepts, which are the
strongest indicators of the sense of presence that influences learning engagement in
Model A and Model B:
• active interpersonal social presence (Model A and B)—feeling like one can and
does interact directly with another human being in an expected reciprocal social
relationship, and general likelihood to take actual physical actions in response
(International Society for Presence Research, 2000; Lombard & Ditton, 1997).
• engagement/mental immersion (Model A and B)—the degree to which the subject
is cognitively focused on the content of the experience and that the mediated
circumstance dominates the subject’s senses and “submerges the perceptual
system” (Biocca & Ben Delaney, 1995, p. 57).
• social richness (Model A and B)—feeling that the mediated environment is
populated with other live social actors and natural objects and that interaction is
fully responsive, vivid, warm, natural and generally unmediated and unobstructed
(Lombard & Ditton, 1997; Lombard et al., 2000).
PRESENCE AND ENGAGEMENT 149
• spatial presence (Model A only)—feeling as if one has been transported or is
physically located in another realistic location, with other actors and a degree of
control over that environment (Lombard & Ditton, 1997).
The complete list of items contained in each ESoPI subscale is provided in
Appendix C.
These two models also indicate that two particular subscales (of four) on the
SCEQ are the best indicators of the learning engagement influenced by sense of presence:
skills and emotional engagement.
The skills subscale includes metacognitive strategy usage, commitment of effort,
and the practice of appropriate learning strategies in the classroom (items include “taking
good notes in class,” “looking over class notes between classes to make sure I understand
the material,” “putting forth effort,” and “being organized”) (Handelsman et al., 2005).
The emotional learning engagement subscale items represent the student’s emotional
involvement with the material, desirable translation and application of concepts, and in-
and out- of class interest in material (items include “applying course material to my life,”
“thinking about the course between class meetings,” and “really desiring to learn the
material”) (Handelsman et al., 2005).
Follow-up Survey
Prompted by observation of participants’ strong emotional reactions in Phase 2
(without text chat) in the online sample, an additional follow-up survey was conducted to
obtain qualitative information about the experimental condition. While the full results
are presented in Chapter 4 with discussion, a summary interpretation is provided in this
section.
PRESENCE AND ENGAGEMENT 150
The limited scope of the follow-up survey restricts its generalizability, but the
results support the quantitative findings in the study, and provide insight into the finding
that the availability of text chat as a part of the VLE positively influences the sense of
presence in the online environment. Given that presence influences learning engagement,
and has no statistically significant interaction with the location (on-campus vs. online)
itself, the relationship of text chat to presence in the online environment is of particular
interest.
In their follow-up survey responses, the students express a preference for text chat
availability, describing its function in ways that suggest it is strengthening their mental
immersion by combatting boredom and the frustration of waiting to interact with the
audiovisual proceedings, where only one person can be heard at a time. Based on
participants’ comments, the text chat boosts social richness through its immediacy and
responsiveness, while also increasing mental immersion by occupying more of the
participants’ attention and shifting their cognitive focus to the content of the experience
rather than the structure of its presentation.
Participants reported suspicions that older individuals may struggle with text chat
as a distraction, and age may indeed play a role in this unanticipated result. Since active
interpersonal social presence remains an important influence, there is a possibility that
students in this general age group are so familiar with text-based interaction in their daily
lives and facile with its form, that it may represent a fully realistic means of typical
human interaction—i.e., the moderation is natural for this age group, reduced to no more
significant interference than the air between individuals when talking face-to-face.
PRESENCE AND ENGAGEMENT 151
The consensus among participants was that the chat offered more positive
immersion and interaction possibilities than the audiovisual module alone, moderating
potential deficits in the VLE experience. One participant suggested that this more
artificial interface helped to compensate for the lack of natural, organic interaction
possibilities that would be offered by a physical classroom, such as bumping into a
classmate in the hallway outside the room, or casually approaching a professor after
class. Participants’ key phrases not only indicate text chat bolstering social richness but
also support the sense of spatial presence, as they suggest students have a mental
conception of the VLE as a place, even as they suggest ways that it must compensate for
its inherent weaknesses as compared to a physical space.
Overall, the follow-up survey supports the idea that while, on the surface, text
chat may seem to counter the goal of immersive presence, it actually increases the
immersion and richness of the experience, leading to better overall sense of presence and
learning engagement.
Limitations of the Present Study
While every effort was made to conduct the study in a manner that was as
scientifically sound as possible, several important limitations should be noted.
Instrumentation
The Student Course Engagement Questionnaire (SCEQ) was chosen to test
learning engagement, and the Early Sense of Presence Inventory (ESoPI) was chosen to
test the sense of presence. Both instruments are self-report surveys, and have the
potential for inaccurate reporting, because of misinterpretation of questions, faulty
memory, or unintended responses (Salkind, 2012). Though the tests were administered
PRESENCE AND ENGAGEMENT 152
anonymously, there is always the possibility that self-report answers are biased by social
desirability (respondents may have attempted to intuit favorable or appropriate answers)
(Kimberlin & Winterstein, 2008; Salkind, 2012). Self-report tests are more corruptible in
general than physiological observation, but are more reliable in gaining insight about
latent, internal phenomena, such as the sense of presence (Salkind, 2012; Wylie, 1974).
The combination of the two instruments into a single survey may have been
responsible for unintended responses, as both were presented with their original Likert
scales, which differ in the number of scale points and the reversal of coding (Very
Characteristic of Me to Not at all Characteristic of Me versus Not at all to Very Much).
Based on analysis of responses, however, significant measurement error is not indicated.
Inconsistency in the definition of the terms presence and learning engagement
limits the instruments’ generalizability as each represents a specific interpretation of
these terms. With both concepts, broad consensus is somewhat elusive, and so the
composition of each instrument must be clearly understood when generalizing from the
data.
Sample Size
With an overall sample size of 148, the study is adequate for the statistical tests
conducted, but a larger sample (n > 200) would provide the ability to test larger and more
complex structural equation models with confidence. As discussed previously, the
sample sizes for the instructor-controlled groups within the main sample are all too small
for meaningful analysis. As mentioned previously, the current limits of real-time
audiovisual virtual learning interfaces make large samples with the same instructor and
PRESENCE AND ENGAGEMENT 153
content difficult to obtain, especially when attempting to pair VLE sessions with on-
campus sessions with the same instructor and material.
Sample Composition
The sample does not have a wide range of ages, and may be subject to unseen
error if the phenomena observed are significantly influenced by age. Further study with a
wider age range is discussed in the following section. The education level was chosen as
master’s degree students, and further study is recommended on other education levels.
The sample was also heavily female, and so gender bias may be apparent in the data
obtained.
The study used a convenient sample, rather than one where subjects were
randomly assigned to conditions, and therefore maybe subject to selection bias
(Kimberlin & Winterstein, 2008; Salkind, 2012). True randomization is often difficult to
obtain in educational settings, but administrative support may allow for such sampling in
future studies (Salkind, 2012).
Consistency of Experimental Conditions
In the current study, the overall subject matter was held constant and class
material was constant at each occasion of measurement, but eight instructors taught
sections, with three instructors teaching in both locations. While instructors follow the
same syllabus and were asked to standardize their instruction as much as possible, style
variations between instructors and general differences between class sessions may have
introduced inconsistency to the data. The high degree of internal consistency on both
instruments, however, indicates that this was not likely to have been a confound in this
study.
PRESENCE AND ENGAGEMENT 154
A slight threat of order effects exists because the same measure was administered
to the same subjects twice (Kimberlin & Winterstein, 2008; Wylie, 1974). Given the
length of the measurement tool, its call for reaction assessment rather than skills
performance, and the time between occasions of measurement (one week), practice
effects are not expected to have significantly biased the results. The length of the
measurement could potentially cause boredom effects on the second occasion of
measurement, but sample sizes in Phase 1 and Phase 2 were similar.
Model Complexity
The structural equation models tested were parsimonious for the sample obtained.
More thorough structural equation models may be constructed that describe the
component concepts of presence and learning engagement as a second level of latent
variables, with more specific data points as the observed variables. Doing so would
likely require a larger sample size for strong analysis and care would need to be taken to
create a properly over-identified model.
On-Campus Intervention Fidelity
As mentioned in the previous discussion section, during the experimental
condition in the on-campus location (Phase 1 - with text), students did not use the text
chat very much when it was made available. A lack of experience using text chat in this
setting may be responsible for the low volume of text messages. The smaller
participation in text chat in the on-campus setting may bias the results in that they may
fail to reflect a comparable text chat element to its heavily-used online analogue.
PRESENCE AND ENGAGEMENT 155
Implications for Practice
The results of the study are specific enough to suggest several practical actions
and strategies. This section presents a number of key recommendations for educational
practice based on the results obtained.
1. Educators should familiarize themselves with the concept of presence and utilize
presence enhancement strategies in instructional settings.
Presence is a statistically significant predictor of learning engagement, and the
strength of the relationship (42% of the variance explained) should encourage educators
to investigate methods for enhancing presence in the learning environment. This primary
finding of the study is important in that it links the canon of presence research to learning
engagement, a core interest in educational psychology and instruction. Educators should
avail themselves of existing presence literature, understand the concept of presence, and
consider implementing strategies that improve presence as examined in other fields, such
as computer science, entertainment, and telecommunications. Improving the sense of
presence should also boost the level of learning engagement, so it follows that any
effective strategies for improving presence should also meet the goal of improving
learning engagement.
The findings in the current study indicate that there is no statistically significant
interaction between location and sense of presence, so presence strategies are equally
important to pursue in the physical classroom as in the virtual environment.
PRESENCE AND ENGAGEMENT 156
2. Text chat should be considered integral to real-time synchronous virtual learning
environments.
The results of the current study indicate that the current, typical VLE, offering
real-time, synchronous audiovisual and text-based interaction is well-optimized for
presence, presenting no statistically significant difference from the typical on-campus
environment. While some instructors may be tempted to remove the text chat element for
fear of distraction or decreased level of presence, the results indicate that such an action
would have the opposite effect of what is intended. The removal of text chat decreases
presence, and qualitative responses suggest that it leads to distraction and disengagement
among students. Therefore, instructors should utilize the text chat whenever possible in
online environments.
The presence of text chat in the on-campus environment, however, did not show a
statistically significant impact on presence. The current results do not support its usage
in the classroom, but they do not discourage its use, either.
Recommendations based on this finding for research on VLE design, as well as
text chat in the classroom, are presented in the next section.
3. Educators should focus efforts on class elements that boost mental immersion,
social richness, active interpersonal social presence, and spatial presence.
These four subscales of sense of presence—mental immersion, social richness,
active interpersonal social presence, and spatial presence—are most clearly shown to be
the components of the presence that influence learning engagement as seen in the highly-
desirable vectors of skills and emotional engagement. The structural equation models
indicate that one can expect these presence components to have the greatest effect on the
PRESENCE AND ENGAGEMENT 157
skills and emotional engagement of students. The first three presence components are the
observed presence variables in Model B (the closest fit), and all four comprise the
observed presence variables of Model A.
4. Online learning can evoke the same level of presence as in-person learning.
Bias against online learning as being unable to replicate the experience of the
classroom seems unfounded in light of the results in the present study. Since the sense of
presence in the online environment can serve as a theoretical proxy for the ability of the
experience to faithfully emulate that of an in-person classroom, and the results of the
current study show no significant difference in presence when both environments are in
their typical conditions (Model 3), a compelling case is made that VLEs can be (and
indeed, are currently) made adequately to emulate non-mediated experiences.
The results do indicate a statistically significant environmental influence on
learning engagement, however (7% of the variance explained in Research Question 1a),
meaning that other factors tied to the online environment beyond presence are also at
work influencing learning engagement. Continued investigation of these and other
educational factors and comparison of their relative strengths in each environment are
indicated. The current study should be seen as a component of a larger body of research
that may eventually provide a complete picture of online learning and how it compares to
the traditional classroom-based model.
Recommendations for Future Research
The results of the current study inspire a wide array of directions for further
investigation. A number of recommendations for future research are suggested below,
including larger and more complex studies, examination of classroom design and
PRESENCE AND ENGAGEMENT 158
components, and expanded inquiry into the observed effects in other venues and with
other demographics.
Larger Studies
Foremost among other recommendations is the encouragement for additional and
more extensive testing and research of presence-influencing elements in educational
contexts. A survey of existing literature finds a dearth of such studies, and yet obvious
interest among many education researchers who seem to reach toward presence concepts
(e.g. Joo et al., 2013; Liao, 2006; van Merriënboer, Kirschner, & Kester, 2003).
Studies with larger sample sizes would be helpful in confirming the relationships
indicated in the current study, as well as providing the opportunity for more extensive
and complex modeling. Such models could include additional latent factors as
components of presence and learning engagement.
Wider SEM Exploration
Structural equation modeling has proven helpful in the analysis of behavioral
analysis, where latent factors are suspected to be important (Lei & Wu, 2007). Its usage
in the current study allowed for more detailed insight into the unseen mechanism
underlying presence and learning engagement.
Further research utilizing structural equation modeling should be conducted,
exploring a broader array of latent variables and interactions, including flow, motivation,
experience, and observed demographic variables. Such SEM development would
necessitate the gathering of larger and more diverse samples, and more extensive
instrumentation, but offers the possibility of a more complete picture of the elements of
classroom design in both in-person and online contexts. Such wider structural models
PRESENCE AND ENGAGEMENT 159
would also help to place sense of presence within the broader framework of other factors
with more extensive previous research and understanding in educational settings.
New Directions for VLE Design
The finding that removing text chat from the virtual learning environment
decreases presence, rather than supporting it, suggests that the design philosophy for
virtual learning environments should be reconsidered. The evolution of VLE software
has been toward more and more realistic representations of class participants. Indeed,
presence discussion in Lee (2004), Lombard and Ditton (1997) and others would seem to
suggest that more naturalistic experiences would produce higher levels of social presence
and social richness. The current study indicates, however, that seemingly more artificial
elements such as text chat can provide deeper mental immersion and richness, even
without emulating a natural interaction with high fidelity. Indeed, for millennial students,
text interaction may well be very familiar and as comfortable and natural a social
interaction as face-to-face conversation. Even when setting aside the social cues of the
text interaction, the qualitative comments in the current study suggest that text chat
combats boredom and increases mental immersion and engagement to a degree that
compensates for any artificiality overall.
By looking to encourage these particular facets of presence, a broader and more
imaginative approach to VLE construction might be explored. The term skeuomorphism
refers to new objects or materials that retain or imitate the ornamentation and affordances
of their predecessors, such as a computer music program that imitates the volume dials of
a hi-fi stereo, or a recorded shutter sound broadcast by a digital camera when taking a
photo (see Gross, Bardzell, & Bardzell, 2014; Hayles, 1999; Pogue, 2013). Skeuomorphs
PRESENCE AND ENGAGEMENT 160
allow users of new objects to feel that they are old and familiar, rather than new and off-
putting. In this way, efforts to faithfully represent a face-to-face conversation in VLEs
represent a kind of skeuomorphism that should, in theory, help to maintain the sense of
presence by blurring the artificiality of the experience. The negative results of the current
study when removing the text chat interface, which should streamline and strengthen the
realistic representation of natural human interaction, suggest that immersive experience
does not necessarily require faithful representation of low-technology interaction.
Further research should concentrate on this aspect of the findings in an effort to
determine the essential elements in the VLE, and what the limits of reduced
skeuomorphism might be (if there are any). For example, research should examine
whether eliminating the video component (or restricting it to representing only the current
speaker), while maintaining audio from all participants (and text chat) might achieve
similar levels of presence sensed. If so, it would suggest the possibility that more
students might be accommodated in a single VLE session, since it is largely the real-time
display of every participant’s audiovisual representation that limits the number of
concurrent users in present VLE software. This could also have a practical effect of
reducing the financial and technological resources required to implement a successful
VLE.
Technology Enhancement of the On-Campus Classroom
In the present study, when text chat was introduced to the on-campus classroom,
no significant difference was made to the degree of presence sensed. As previously
discussed, this, coupled with the low volume of text messages in the on-campus
environment as compared to the online environment, indicates a weak importance of text
PRESENCE AND ENGAGEMENT 161
chat in a traditional classroom. Further research, however, should examine additional
examples of technology use in physical environments. Electronic whiteboards, mobile
device usage, and even projected slide presentations would all be reasonable elements to
study in regards to their influence on presence in an on-campus classroom. Such research
would strive to identify technological elements or technology use cases that enhance the
immersion or social richness of the classroom environment. This type of research offers
the key to realizing practical means to implement the findings of the present study.
Demographic-Related Research
Logically, the results of the present study can only be confirmed for the
population represented by the sample. As discussed previously, the sample used in the
present study has a narrow range of ages represented, and is limited to the context of
students at an American institution of higher education, engaged in master’s studies.
Expanding the scope of the sample would be an important step in future research into
presence and learning engagement.
Based on the results obtained in the follow-up survey, inquiry focusing on the
relationship of age to presence and learning engagement would be beneficial.
Researchers should strive to obtain participants from both other and younger generations
of learners, with an aim toward understanding whether the presence sensed by students in
the ‘millennial’ generation, well-represented in the current sample, is significantly
different from that sensed by other ages.
Education level is another demographic variable that future research should
examine in the context of presence and learning engagement. Do the relationships
observed in the current study hold for students at an elementary or secondary level, or in
PRESENCE AND ENGAGEMENT 162
a continuing education or corporate training context? These questions, like the effect of
age, are important to answer in aiming to broaden the applicability of these results and
refine the overall understanding of presence in the learning process.
Finally, demographic-focused research should be conducted related to the effect
of culture on sense of presence and learning engagement. With online learning seen as a
primary option in the expansion of international higher education offerings (Altbach &
Knight, 2007), the effect of culture is especially important to understand when designing
VLEs and programs. Culture is especially important to consider in aspects of presence
such as proxemics, the study of perceived interpersonal distance and its effects (E. T.
Hall, 1992; Kristoffersson, Eklundh, & Loutfi, 2013; Yee et al., 2007). Education
research on the impact of cultural difference (e.g. P. J. Smith & Smith, 1999) also
indicates important considerations. A systematic exploration of the interrelationship of
culture, presence, and learning engagement would be informative in the process of
reaching practical conclusions regarding the effective design of cross-cultural classrooms.
Development of New Instruments
While acceptable instruments for measuring presence exist, none of the most
widely used instruments, including the Temple Presence Inventory upon which the Early
Sense of Presence Inventory is based, use a broadly-accessible model of presence such as
that described in Lee (2004). As Lombard and Jones (2007) note, “despite many
discussions and explications, there is still substantial disagreement about the meanings
and scope of the telepresence and presence concepts” (p. 198). While the concept of
presence is still the subject of a wide range of academic opinions, easily-understood
PRESENCE AND ENGAGEMENT 163
aggregate definitions such as Lee’s model offer the potential for more widespread
understanding among practitioners.
Development of new presence instruments, especially ones tuned to Lee’s model
and others with broad, easily-accessible conceptions of presence, will help to validate
results such as those in the current study, as well as make practical applications clearer.
Additional study by educational psychologists and others in the field of education will
also help to hone the currently accepted conceptions of presence both within the field and
in other disciplines. As with most latent constructs, the definition of presence will surely
continue to evolve, but the effective application of presence theory will be greatly
enhanced by pursuing the clearest, most accessible, and most widely-held definition
possible.
Physiological and Qualitative Research of Presence in the Classroom and VLE
In addition to improved self-report instruments, further inquiry into presence in
educational environments should include physiological measures as well as more in-
depth qualitative study.
As discussed in Chapter 2, self-report instruments are affordable, relatively easy
to implement, and helpful in the study of internal behaviors or reactions. Nevertheless,
self-reports carry with them inherent confounds from subjects that may incorrectly self-
assess or report their behavior, due to confusion about questions, mistakes in memory, or
social conformity. While physiological measures may not be easy to link with latent
behaviors with confidence, observations of physical changes, such as eye movement
tracking, can provide additional insight about how learners interact with different types of
mediated environments, even when they may not be fully conscious of their own
PRESENCE AND ENGAGEMENT 164
behaviors. Such specific observation can also provide more precise and quantifiable data
regarding students’ usage of elements within learning environments, especially online
and electronic ones.
In the present study, a follow-up survey added qualitative data to the
understanding of the main study’s more quantified results, helping to guide their
interpretation. Especially when seeking to understand the most highly influential
elements of both online and in-person educational environments, qualitative inquiry can
provide important insight about students’ experiences. Further qualitative study is
recommended in these pursuits.
Expanded Venues
The present study addresses only on-campus and online classroom environments,
but educational experiences can and do take place in other structured venues. Education
in two particular types of mediated environment would be logical concentrations for
future study of presence and its influence on learning engagement: application
development for computers and mobile devices, and themed environments such as
museums. Additionally, research on asynchronous online learning environments should
be pursued.
Rapidly expanding availability and widespread use of mobile devices and
computer software has led not only to a blossoming of educational software applications
(“apps”) aimed at all age groups, but also to considerable interest from instructors and
school administrators in incorporating tablets and laptop computers into traditional on-
campus curricula at all education levels (e.g., Culén & Gasparini, 2011; Foote, 2010;
Lennon & Girard, 2012; Melhuish & Falloon, 2010; Mennenga & Hendrickx, 2008;
PRESENCE AND ENGAGEMENT 165
Rush, 2008; Russell & Higgins, 2004). The types of experiences provided by apps on
tablets, laptops, and other electronic platforms are fully mediated, designed experiences,
and therefore, an understanding and application of presence theory has the potential for
significant effects. Further study should be conducted to understand any differences in
the influence of presence on learning engagement in app designs, the overall effect of the
app environment on presence, and the effects of individual app elements on the sense of
presence.
Nontraditional education venues such as immersive museum and theme park
experiences should also be the subject of additional inquiry. Museum exhibits and theme
park attractions are becoming increasingly sophisticated, and, like digital apps, represent
fully mediated environments, whose goals often involve convincing visitors they have
been transported to another time, place, or environment such as having been shrunken to
the size of an ant, or transported to Germany at the time of the fall of the Berlin wall, in
the hopes of creating an illuminating and memorable educational experience (e.g.,
Jeffers, 2004; Seymour, Ashton, & Edwards, 1986; Wood & Wolf, 2008). The increased
theming of museums and zoos, and the growth of ‘edutainment’ at theme parks, is not
without controversy (e.g., Beardsworth & Bryman, 2001; Francaviglia, 1995), but
certainly represents another venue where a fuller understanding of the nature of presence
and learning engagement should be sought, not only for the improvement of such
experiences, but for the potential to expand our understanding of presence-influencing
elements which might be adapted for use in traditional classroom venues.
Finally, further inquiry should also be conducted in the realm of asynchronous
online learning. The present study examined synchronous, live virtual learning
PRESENCE AND ENGAGEMENT 166
environments, where all participants are portrayed in real time and may interact with each
other. Also popular, however, among higher education programs and ‘massive open
online courses’ (MOOCs) is the asynchronous online design, which can often more
closely resemble an app than a truly interactive classroom (Tao & Zhang, 2013). These
online experiences typically pair recorded lecture content or one-way live broadcast with
text-based discussion forums. Given the wide adoption and significant structural
differences from real-time, synchronous VLEs, an understanding of the behavior of
presence in asynchronous online platforms, its influence on learning engagement, and the
environmental elements that augment or detract from it, is another important avenue for
future research.
Summary
Understanding presence, and its influences, offers a wealth of new useful tools
and strategies for strengthening and enhancing learning engagement. The current study
has shown clear evidence that presence—especially mental immersion, social richness,
active interpersonal social presence, and spatial presence—is a significant predictor of
learning engagement. Given the paramount importance of learning engagement to the
educational experience, it would be highly advisable to pursue further study of presence
in educational contexts, as well as inquiry regarding the most effective tactics for
enhancing the sense of presence in these contexts.
The results indicate that educators, educational researchers, and designers of
educational interfaces should avail themselves of presence concepts, and make efforts to
increase their students’ sense of presence in both classrooms and virtual learning
environments. As pedagogy and curricula continue to evolve and embrace new
PRESENCE AND ENGAGEMENT 167
technologies and new possibilities for interaction, it is vital to consider existing literature
related to presence, and contribute to a fuller understanding of its influences and
applications in education.
The results also reinforce the value of text-based interaction in the online learning
environment, pointing to its value as an enhancement to mental immersion and richness.
While some educators may be suspicious of its potential for distraction, the
counterintuitive results in this study show that it should be considered an essential
element for the maintenance of a sense of presence that is statistically similar to that of an
on-campus classroom.
The present study affirms and reinforces that the sense of presence should
continue to be a focus of scholarly research in education and educational psychology.
This study represents only a first step toward a complete understanding of presence in an
educational context, but it offers tantalizing evidence that presence represents an
important avenue for future inquiry and practical application. A clearer understanding of
its properties and its effects on learning engagement could be of significant and
continuing benefit to educators and students, making the learning process more effective
in both traditional and emerging venues of instruction.
PRESENCE AND ENGAGEMENT 168
References
Ahlfeldt, S., Mehta, S., & Sellnow, T. (2005). Measurement and analysis of student
engagement in university classes where varying levels of PBL methods of
instruction are in use. Higher Education Research & Development, 24(1), 5–20.
doi:10.1080/0729436052000318541
Alexander, P. A., Schallert, D. L., & Reynolds, R. E. (2009). What is learning anyway? A
topographical perspective considered. Educational Psychologist, 44(3), 176–192.
doi:10.1080/00461520903029006
Altbach, P. G., & Knight, J. (2007). The internationalization of higher education:
Motivations and realities. Journal of Studies in International Education, 11(3/4),
290–305. doi:10.1177/1028315307303542
Anderson, D. R., Lorch, E. P., Field, D. E., & Sanders, J. (1981). The effects of TV
program comprehensibility on preschool children's visual attention to television.
Child Development, 52(1), 151-157.
Appleton, J. J., Christenson, S. L., & Furlong, M. J. (2008). Student engagement with
school: Critical conceptual and methodological issues of the construct.
Psychology in the Schools, 45(5), 369–386. doi:10.1002/pits.20303
Appleton, J. J., Christenson, S. L., Kim, D., & Reschly, A. L. (2006). Measuring
cognitive and psychological engagement: Validation of the Student Engagement
Instrument. Journal of School Psychology, 44(5), 427–445.
doi:10.1016/j.jsp.2006.04.002
Arbuckle, J. L. (2013). IBM SPSS Amos 22 User’s Guide (pp. 1–700). IBM.
PRESENCE AND ENGAGEMENT 169
Armstrong, L. (2007). Competing in the global higher education marketplace:
Outsourcing, twinning, and franchising. New Directions for Higher Education,
140, 131–138. doi:10.1002/he.287
Astin, A. (1984). Student involvement: A developmental theory for higher education.
Journal of College Student Personnel, 25(4), 297–308.
Baddeley, A. (1992). Working memory. Science, 255(5044), 556–559.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84(2), 191–215.
Bandura, A. (1991). Social cognitive theory of self-regulation. Organizational Behavior
and Human Decision Processes, 50(2), 248–287. doi:10.1016/0749-
5978(91)90022-L
Bandura, A. (1993). Perceived self-efficacy in cognitive development and functioning.
Educational Psychologist, 28(2), 117–148. doi:10.1207/s15326985ep2802_3
Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of
Psychology, 52, 1-26.
Bandura, A., & Schunk, D. H. (1981). Cultivating competence, self-efficacy, and intrinsic
interest through proximal self-motivation. Journal of Personality and Social
Psychology, 41(3), 586–598.
Bartsch, R. A., & Cobern, K. M. (2003). Effectiveness of PowerPoint presentations in
lectures. Computers & Education, 41(1), 77–86. doi:10.1016/S0360-
1315(03)00027-7
PRESENCE AND ENGAGEMENT 170
Beardsworth, A., & Bryman, A. (2001). The wild animal in late modernity: The case of
the Disneyization of zoos. Tourist Studies, 1(1), 83–104.
doi:10.1177/146879760100100105
Beery, T. A., Shell, D., Gillespie, G., & Werdman, E. (2013). The impact of learning
space on teaching behaviors. Nurse Education in Practice, 13(5), 382–387.
doi:10.1016/j.nepr.2012.11.001
Bentler, P. M. (1980). Multivariate-analysis with latent-variables - causal-modeling.
Annual Review of Psychology, 31, 419–456.
Bentler, P. M. (1988). Causal Modeling via Structural Equation Systems. In Handbook of
multivariate experimental psychology (pp. 317–335). Boston, MA: Springer US.
doi:10.1007/978-1-4613-0893-5_9
Biggs, J. B. (1987). Learning process questionnaire manual. Student approaches to
learning and studying. ERIC.
Billinghurst, M., Weghorst, S., & Furness, T., III. (1998). Shared space: An augmented
reality approach for computer supported collaborative work. Virtual Reality, 3(1),
25–36.
Biocca, F. (1997). The cyborg's dilemma: Progressive embodiment in virtual
environments. Journal of Computer-Mediated Communication, 3(2), 113–144.
Biocca, F., & Delaney, B. (1995). Immersive virtual reality. In F. Biocca & M. R. Levy
(Eds.), Communication in the age of virtual reality (pp. 57–126). Hillsdale, NJ:
Lawrence Erlbaum Associates.
PRESENCE AND ENGAGEMENT 171
Biocca, F., Burgoon, J., Harms, C., & Stoner, M. (2001, May). Criteria and scope
conditions for a theory and measure of social presence. Paper presented at
Presence 2001 conference, Philadelphia, PA.
Biocca, F., Harms, C., & Burgoon, J. K. (2003). Toward a more robust theory and
measure of social presence: Review and suggested criteria. Presence:
Teleoperators and Virtual Environments, 12(5), 456–480.
Blake, E., Nunez, D., & Labuschagne, B. (2007). Longitudinal effects on presence:
Suspension of disbelief or distrust of naive belief? (pp. 291–295). Presented at the
Proceedings of the 10th Annual International Workshop on Presence.
Blessinger, P., & Wankel, C. (2012). Innovative approaches in higher education: An
introduction to using immersive interfaces. In C. Wankel, P. Blessinger, J.
Stanaityte, & N. Washington (Eds.), Increasing student engagement and retention
using immersive interfaces (pp. 3–14). Bingley, UK: Emerald Group Publishing.
Bollen, K. A. (1989). Structural Equations with Latent Variables. New York, NY: Wiley-
Interscience.
Bollen, K. A., & Long, J. S. (1992). Tests for Structural Equation Models: Introduction.
Sociological Methods & Research, 21(2), 123–131.
doi:10.1177/0049124192021002001
Bollen, K. A., & Long, J. S. (1993). Testing structural equation models. Newbury Park,
NJ: Sage Publications.
PRESENCE AND ENGAGEMENT 172
Bouchard, S., Dumoulin, S., Michaud, M., & Gougeon, V. (2011). Telepresence
experienced in videoconference varies according to emotions involved in
videoconference sessions. Studies in Health Technology and Informatics, 167,
128–132.
Brand, J., & Kinash, S. (2010). Pad-agogy: A quasi-experimental and ethnographic pilot
test of the iPad in a blended mobile learning environment (pp. 1–6). Presented at
the 27th Annual Conference of the Australian Society for Computers in Learning
in Tertiary Education (ASCILITE), Sydney, Australia. Retrieved from
http://epublications.bond.edu.au/tls/16
Browne, M. W., & Cudeck, R. (1993). Alternative Ways of Assessing Model Fit. In K.
A. Bollen & J. S. Long (Eds.), Testing Structural Equation Models. Newbury
Park, CA: Sage.
Brünken, R., Plass, J. L., & Leutner, D. (2004). Assessment of cognitive load in
multimedia learning with dual-task methodology: Auditory load and modality
effects. Instructional Science, 32(1), 115–132.
doi:10.1023/B:TRUC.0000021812.96911.c5
Burgess, J. W., & Fordyce, W. K. (1989). Effects of preschool environments on
nonverbal social behavior: toddlers' interpersonal distances to teachers and
classmates change with environmental density, classroom design, and parent-child
interactions. Journal of Child Psychology and Psychiatry, and Allied Disciplines,
30(2), 261–276.
Butte, G. (2008). Suture and the Narration of Subjectivity in Film. Poetics Today, 29(2),
277–308. doi:10.1215/03335372-2007-026
PRESENCE AND ENGAGEMENT 173
Byrne, B. M. (1994). Structural Equation Modeling with EQS and EQS/WINDOWS.
Newbury Park, CA: Sage.
Byrne, B. M. (2010). Structural Equation Modeling With AMOS (2nd ed.). New York,
NY: Routledge.
Caird, J. (2013). Theatre Craft: A Director's Practical Companion from A to Z. New
York, NY: Faber & Faber.
Caldwell, B. S. (1994). The Learning-Friendly Classroom: The educational process is
enhanced when human factors principles are integrated into classroom design.
Ergonomics in Design: the Quarterly of Human Factors Applications, 2(1), 30–
35. doi:10.1177/106480469400200110
Carroll, N. (1993). Toward a Theory of Point-of-View Editing: Communication,
Emotion, and the Movies. Poetics Today, 14(1), 123–141.
Chandler, P., & Sweller, J. (1991). Cognitive load theory and the format of instruction.
Cognition and Instruction, 8(4), 293–332.
Chandler, P., & Sweller, J. (1996). Cognitive load while learning to use a computer
program. Applied Cognitive Psychology, 10(2), 151–170.
Clark, R. E., Early, S., & Yates, K. (2006). The suspension of disbelief: Effects of
mediated “presence” on motivation and learning. Presented at the ICT Seminar
Series.
Coates, H. (2006). Student Engagement in Campus-Based and Online Education. New
York, NY: Routledge.
Csikszentmihalyi, M. (1975). Beyond boredom and anxiety. San Francisco, CA: Jossey-
Bass.
PRESENCE AND ENGAGEMENT 174
Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York,
NY: Harper & Row.
Csikszentmihalyi, M., & Larson, R. (1987). Validity and reliability of the experience-
sampling method. The Journal of Nervous and Mental Disease, 175(9), 526–536.
Culén, A. L., & Gasparini, A. (2011). iPad: A new classroom technology? A report from
two pilot studies. INFuture2011: “Information Sciences and E-Society,” 1–10.
Retrieved from http://hnk.ffzg.hr/
da Graça, V. A. C., Kowaltowski, D. C. C. K., & Petreche, J. R. D. (2007). An evaluation
method for school building design at the preliminary phase with optimisation of
aspects of environmental comfort for the school system of the State São Paulo in
Brazil. Building and Environment, 42(2), 984–999.
doi:10.1016/j.buildenv.2005.10.020
DeVellis, R. F. (2003). Scale development: Theory and applications. Thousand Oaks,
CA: Sage Publications, Inc.
Dickey, M. D. (2005). Engaging by design: How engagement strategies in popular
computer and video games can inform instructional design. Educational
Technology Research and Development, 53(2), 67–83.
Ditton, T. B. (1997). The Unintentional Blending of Direct Experience and Mediated
Experience. Temple University.
Dove, M. K. (2009). The relationship of rhythmic and melodic perception with
background music distraction in college level students. Manhattan, KS:
University of Kansas.
PRESENCE AND ENGAGEMENT 175
Draper, J. V. J., Kaber, D. B. D., & Usher, J. M. J. (1998). Telepresence. Human Factors:
Journal of the Human Factors and Ergonomics Society, 40(3), 354–375.
Eadie, G. M. (2001). The impact of ICT on schools: Classroom design and curriculum
delivery. Schools for the Future. Retrieved from: http://colos1.fri.uni-
lj.si/infor/DOKUMENTI/A_POROCILO/REFERENCE/ORGANIZIRANOST/m
obileclassroom.pdf
Early, S. (2006). Measurement of presence in mediated environments: obstacles and
opportunities, 1–14. Center for Cognitive Technology, University of Southern
California Rossier School of Education.
Early, S. F. (2008). Early Sense of Presence Inventory (Media Questionnaire).
Unpublished instrument.
Enyedy, N. (2003). Knowledge construction and collective practice: At the intersection of
learning, talk, and social configurations in a computer-mediated mathematics
classroom. The Journal of the Learning Sciences, 12(3), 361–407.
Espey, M. (2008). Does space matter? Classroom design and team-based learning.
Review of Agricultural Economics, 30(4), 764–775. doi:10.1111/j.1467-
9353.2008.00445.x
Falk, C. F., & Savalei, V. (2011). The relationship between unstandardized and
standardized alpha, true reliability, and the underlying measurement model.
Journal of Personality Assessment, 93(5), 445–453.
doi:10.1080/00223891.2011.594129
Finn, J. D. (1993). School engagement & students at risk. National Center for Education
Statistics.
PRESENCE AND ENGAGEMENT 176
Finneran, C. M., & Zhang, P. (2003). A person-artifact-task (PAT) model of flow
antecedents in computer-mediated environments. International Journal of
Human-Computer Studies, 59(4), 475–496.
Foote, C. (2010). Checking out the iPad. MultiMedia & Internet@Schools, 17(6), 17–19.
Francaviglia, R. (1995). History after Disney: The significance of “imagineered”
historical places. The Public Historian, 17(4), 69–74.
doi:10.2307/3378386?ref=search-gateway:77212ef85768e08183e608b3e00d9aec
Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engagement: Potential
of the concept, state of the evidence. Review of Educational Research, 74(1), 59–
109.
Fredricks, J., McColskey, W., Meli, J., Mordica, J., Montrosse, B., & Mooney, K. (2011).
Measuring student engagement in upper elementary through high school: A
description of 21 instruments (No. 098) (pp. 1–88). Greensboro, NC: Regional
Educational Laboratory Southeast.
Furrer, C., & Skinner, E. (2003). Sense of relatedness as a factor in children's academic
engagement and performance. Journal of Educational Psychology, 95(1), 148–
162. doi:10.1037/0022-0663.95.1.148
Gale, N., Golledge, R. G., Pellegrino, J. W., & Doherty, S. (1990). The acquisition and
integration of route knowledge in an unfamiliar neighborhood. Journal of
Environmental Psychology, 10(1), 3–25. doi:10.1016/S0272-4944(05)80021-0
PRESENCE AND ENGAGEMENT 177
Garner, R., Brown, R., Sanders, S., & Menke, D. J. (1992). “Seductive details” and
learning from text. In K. A. Renninger, S. Hidi, & A. Krapp (Eds.), The role of
interest in learning and development (pp. 239–254). Hillsdale, NJ: Lawrence
Erlbaum Associates, Inc.
Garner, R., Gillingham, M. G., & White, C. S. (1989). Effects of “seductive details” on
macroprocessing and microprocessing in adults and children. Cognition and
Instruction, 6(1), 41–57.
Garrison, D. R., & Anderson, T. (2003). E-Learning in the 21st century: A framework for
research and practice. London: Routledge Falmer.
Garrison, D. R., & Cleveland-Innes, M. (2005). Facilitating cognitive presence in online
learning: Interaction is not enough. The American Journal of Distance Education,
19(3), 133–148.
Gerrig, R. J. (1993). Experiencing Narrative Worlds. New Haven, CT: Yale University
Press.
Ghani, J. A., & Deshpande, S. P. (1994). Task characteristics and the experience of
optimal flow in human—computer interaction. The Journal of Psychology,
128(4), 381–391.
Gibson, E. J. (1988). Exploratory-behavior in the development of perceiving, acting, and
the acquiring of knowledge. Annual Review of Psychology, 39, 1–41.
Gilbert, D. T. (1991). How mental systems believe. American Psychologist, 46(2), 107–
119.
PRESENCE AND ENGAGEMENT 178
Gilkey, R. H., Simpson, B. D., & Weisenberger, J. M. (2001). Creating auditory
presence. In M. J. Smith, G. Salvendy, D. Harris, & R. J. Koubek (Eds.), Usability
Evaluation and Interface Design (pp. 609–613). Mahwah, NJ: CRC Press.
Green, M. C., & Brock, T. C. (2000). The role of transportation in the persuasiveness of
public narratives. Journal of Personality and Social Psychology, 79(5), 701–721.
doi:10.1037//0022-3514.79.5.701
Green, M. C., Brock, T. C., & Kaufman, G. F. (2004). Understanding media enjoyment:
The role of transportation into narrative worlds. Communication Theory, 14(4),
311–327.
Gross, S., Bardzell, J., & Bardzell, S. (2014). Skeu the evolution: skeuomorphs, style, and
the material of tangible interactions (pp. 53–60). Presented at the 8th International
Conference on Tangible, Embedded and Embodied Interaction.
Guardino, C., & Fullerton, E. (2010). Changing behaviors by changing the classroom
environment. Teaching Exceptional Children, 42(6), 8–13.
Hall, E. T. (1992). The Hidden Dimension. Gloucester, MA: Peter Smith Pub
Incorporated.
Hall, J. L. (2007). Distance learning through synchronous interactive television. Journal
of Veterinary Medical Education, 34(3), 263–268.
Hall, T. E., Meyer, A., & Rose, D. H. (2012). Universal Design for Learning in the
Classroom. New York, NY: Guilford Press.
Handelsman, M. M., Briggs, W. L., Sullivan, N., & Towler, A. (2005). A measure of
college student course engagement. Journal of Educational Administration, 98(3),
184–191.
PRESENCE AND ENGAGEMENT 179
Harp, S. E., & Mayer, R. E. (1998). How seductive details do their damage: A theory of
cognitive interest in science learning. Journal of Educational Psychology, 90,
414–434.
Harteveld, C., Thij, ten, E., & Copier, M. (2011). Design for engaging experience and
social interaction. Simulation & Gaming, 42(5), 590–595.
doi:10.1177/1046878111426960
Hayles, N. K. (1999). How We Became Posthuman. Chicago: University of Chicago
Press.
Heath, S. (1981). Questions of cinema. Bloomington, IN: Indiana University Press.
Herrington, J., Oliver, R., & Reeves, T. C. (2003). Patterns of engagement in authentic
online learning environments. Australian Journal of Educational Technology,
19(1), 59–71.
Hoffman, D. L., & Novak, T. P. (1996). Marketing in hypermedia computer-mediated
environments: Conceptual foundations. Journal of Marketing, 60(3), 50–68.
Hu, L.-T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Structural Equation
Modeling, 6(1), 1–55. doi:10.1080/10705519909540118
Huk, T., Bieger, S., Ohrmann, S., & Weigel, B. (2004). Computer animations in science
education: is background music beneficial or detrimental? (pp. 4227–4234).
Presented at ED-MEDIA.
IJsselsteijn, W. A., de Ridder, H., Freeman, J., & Avons, S. E. (2000). Presence: concept,
determinants, and measurement (pp. 520–529). Presented at Electronic Imaging,
International Society for Optics and Photonics.
PRESENCE AND ENGAGEMENT 180
International Society for Presence Research. (2000, January 1). The Concept of Presence:
Explication Statement. Retrieved March 7, 2014, from http://ispr.info
Jeffers, C. S. (2004). In a cultural vortex: Theme parks, experience, and opportunities for
art education. Studies in Art Education, 45(3), 221–233. doi:10.2307/1320970
Jimerson, S. R., Campos, E., & Greif, J. L. (2003). Toward an understanding of
definitions and measures of school engagement and related terms. The California
School Psychologist, 8(1), 7–27.
Joo, Y. J., Joung, S., & Kim, E. K. (2013). Structural relationships among e-learners'
sense of presence, usage, flow, satisfaction, and persistence. Educational
Technology & Society, (16), 310–324.
Kalyuga, S. (2011). Cognitive load theory: How many types of load does it really need?
Educational Psychology Review, 23(1), 1–19. doi:10.1007/s10648-010-9150-7
Kampfe, J., Sedlmeier, P., & Renkewitz, F. (2011). The impact of background music on
adult listeners: A meta-analysis. Psychology of Music, 39(4), 424–448.
doi:10.1177/0305735610376261
Kearsley, G., & Shneiderman, B. (1998). Engagement theory: A framework for
technology-based teaching and learning. Educational Technology, 38(5), 20–23.
Kenny, D. A., & McCoach, D. B. (2003). Effect of the number of variables on measures
of fit in structural equation modeling. Structural Equation Modeling, 10(3), 333–
351. doi:10.1207/S15328007SEM1003_1
Kim, T., & Biocca, F. (1997). Telepresence via television: Two dimensions of
telepresence may have different connections to memory and persuasion. Journal
of Computer-Mediated Communication, 3(2), 0–0.
PRESENCE AND ENGAGEMENT 181
Kimberlin, C. L., & Winterstein, A. G. (2008). Validity and reliability of measurement
instruments used in research. American Journal of Health-System Pharmacy,
65(23), 2276–2284. doi:10.2146/ajhp070364
Kirschner, F., Kester, L., & Corbalan, G. (2011). Cognitive load theory and multimedia
learning, task characteristics and learning engagement: The current state of the
art. Computers in Human Behavior, 27(1), 1–4. doi:10.1016/j.chb.2010.05.003
Kirschner, P. A. (2002). Cognitive load theory: Implications of cognitive load theory on
the design of learning. Learning and Instruction, 12(1), 1–10.
Kirschner, P. A., Ayres, P., & Chandler, P. (2011a). Contemporary cognitive load theory
research: The good, the bad and the ugly. Computers in Human Behavior, 27(1),
99–105. doi:10.1016/j.chb.2010.06.025
Kirschner, P., Kirschner, F., & Paas, F. (2011b, July 12). Cognitive load theory. The Gale
Group. Retrieved September 20, 2011, from https://blackboard.usc.edu/
Klein, D. C. D., O'Neil, H. F., Jr, & Baker, E. L. (1998). A Cognitive Demands Analysis
of Innovative Technologies (No. 454) (pp. 1–30). Los Angeles, CA: National
Center for Research on Evaluation, Standards, and Student Testing (CRESST),
Center for the Study of Evaluation (CSE).
Kolb, A. Y., & Kolb, D. A. (2005). Learning styles and learning spaces: Enhancing
experiential learning in higher education. Academy of Management, Learning &
Education, 4(2), 193–212.
Kotovsky, K., Hayes, J. R., & Simon, H. A. (1985). Why are some problems hard?
Evidence from Tower of Hanoi. Cognitive Psychology, 17(2), 248–294.
doi:10.1016/0010-0285(85)90009-X
PRESENCE AND ENGAGEMENT 182
Kreijns, K., Kirschner, P. A., & Jochems, W. (2003). Identifying the pitfalls for social
interaction in computer-supported collaborative learning environments: A review
of the research. Computers in Human Behavior, 19(3), 335–353.
Kristoffersson, A., Eklundh, K. S., & Loutfi, A. (2013). Measuring the quality of
interaction in mobile robotic telepresence: A pilot’s perspective. International
Journal of Social Robotics, 5(1), 89–101. doi:10.1007/s12369-012-0166-7
Lee, K. M. (2004). Presence, explicated. Communication Theory, 14(1), 27–50.
Lehman, S., Schraw, G., McCrudden, M. T., & Hartley, K. (2007). Processing and recall
of seductive details in scientific text. Contemporary Educational Psychology,
32(4), 569–587. doi:10.1016/j.cedpsych.2006.07.002
Lei, P.-W., & Wu, Q. (2007). Introduction to structural equation modeling: Issues and
practical considerations. Educational Measurement Issues and Practice, 26(3),
33–43. doi:10.1111/j.1745-3992.2007.00099.x
Lemish, D. (1982). The rules of viewing television in public places. Journal of
Broadcasting, 26(4), 757–781. doi:10.1080/08838158209364046
Lennon, M. M., & Girard, T. (2012, January 20). iPadagogy: Enhancing business
education with the iPad. Retrieved from http://www3.MarkPhD.com
Lessiter, J., Freeman, J., Keogh, E., & Davidoff, J. (2001). A cross-media presence
questionnaire: The ITC-Sense of Presence Inventory. Presence: Teleoperators
and Virtual Environments, 10(3), 282–297.
Li, H., Daugherty, T., & Biocca, F. (2002). Impact of 3-D advertising on product
knowledge, brand attitude, and purchase intention: The mediating role of
presence. Journal of Advertising, 43–57.
PRESENCE AND ENGAGEMENT 183
Liang, J.-K., Liu, T.-C., Wang, H.-Y., Chang, B., Deng, Y.-C., Yang, J.-C., et al. (2005).
A few design perspectives on one-on-one digital classroom environment. Journal
of Computer Assisted Learning, 21(3), 181–189.
Liao, L. F. (2006). A flow theory perspective on learner motivation and behavior in
distance education. Distance Education, 27(1), 45–62.
Lin, K.-M. (2007). The Lacanian spectator: Lacanian psychoanalysis and the cinema.
(Doctoral dissertation). Retrieved from ProQuest. (UMI No. 3275744)
Linek, S. B., Marte, B., & Albert, D. (2011). Background music in educational games:
Motivational appeal and cognitive impact. International Journal of Game-Based
Learning (IJGBL), 1(3), 53–64.
Lombard, M. (1995). Direct responses to people on the screen: Television and personal
space. Communication Research, 22(3), 288–324.
doi:10.1177/009365095022003002
Lombard, M., & Ditton, T. (1997). At the heart of it all: The concept of presence. Journal
of Computer-Mediated Communication, 3(2).
Lombard, M., & Ditton, T. (2004). A literature-based presence measurement instrument:
The Temple Presence Inventory (TPI) (Beta). The Seventh International
Workshop on Presence (PRESENCE 2004). Philadelphia, PA: Media Interface
and Network Design (M.I.N.D.) Lab, Temple University.
Lombard, M., & Jones, M. T. (2007). Identifying the (tele)presence literature.
PsychNology Journal, 5(2), 197–206.
PRESENCE AND ENGAGEMENT 184
Lombard, M., Ditton, T. B., Crane, D., Davis, B., Gil-Egui, G., Horvath, K., & Rossman,
J. (2000). Measuring presence: A literature-based approach to the development of
a standardized paper-and-pencil instrument (pp. 1–13). Presented at the Presence
The Third International Workshop on Presence, Delft, The Netherlands.
Lurie, P. (2001). “Some trashy myth of reality's escape:” Romance, history, and film
viewing in Absalom, Absalom! American Literature, 73(3), 563–597.
doi:10.1215/00029831-73-3-563
MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and
determination of sample size for covariance structure modeling. Psychological
Methods. doi:10.1037/1082-989X.1.2.130
Mammarella, N. N., Fairfield, B. B., & Cornoldi, C. C. (2007). Does music enhance
cognitive performance in healthy older adults? The Vivaldi effect. Aging Clinical
and Experimental Research, 19(5), 394–399.
Mason, R. (2004). Spinoza and the unimportance of belief. Philosophy, 79(02).
doi:10.1017/S0031819104000269
Mayer, R. E. (1999). Multimedia Aids to Problem-Solving Transfer. International
Journal of Educational Research, 31, 611–623.
Mayer, R. E. (2005). Cognitive theory of multimedia learning. In R. E. Mayer (Ed.), The
Cambridge handbook of multimedia learning (pp. 31–48). New York, NY:
Cambridge University Press.
Mayer, R. E. (2009). Multimedia learning (2nd ed.). New York, NY: Cambridge
University Press.
Mayer, R. E. (2011). Applying the science of learning. Boston, MA: Pearson Education.
PRESENCE AND ENGAGEMENT 185
Mayer, R. E., & Moreno, R. (1998). A cognitive theory of multimedia learning:
Implications for design principles. Presented at the Annual meeting of the ACM
SIGCHI Conference on Human Factors in Computing Systems.
Mayer, R. E., & Moreno, R. (2003). Nine ways to reduce cognitive load in multimedia
learning. Educational Psychologist, 38(1), 43–52.
doi:10.1207/S15326985EP3801_6
Mayer, R. E., Heiser, J., & Lonn, S. (2001). Cognitive constraints on multimedia
learning: When presenting more material results in less understanding. Journal of
Educational Psychology, 93(1), 187–198. doi:10.1037//0022-0663.93.1.187
Mayher, J. S., & Brause, R. S. (1986). Learning through teaching: Is your classroom like
your grandmother’s? Language Arts, 63(6), 617–620.
Melhuish, K., & Falloon, G. (2010). Looking to the future: M-learning with the iPad.
Computers in New Zealand Schools Learning, Leading, Technology, 22(3), 1–16.
Mennenga, H. A., & Hendrickx, L. (2008). Faculty concerns about requiring laptops in
the classroom. Nurse Educator, 33(4), 151–154.
doi:10.1097/01.NNE.0000312197.74191.8d
Merrill, M. D. (2002). First principles of instruction. Educational Technology Research
and Development, 50(3), 43–59.
Meurant, R. C. (2010). The iPad and EFL digital literacy. In Signal Processing and
Multimedia (Vol. 123, pp. 224–234). Springer Berlin Heidelberg.
Meyer, D. K., & Turner, J. C. (2002). Discovering emotion in classroom motivation
research. Educational Psychologist, 37(2), 107–114.
Minsky, M. (1980, June). Telepresence. Omni, 2(9), 44–53.
PRESENCE AND ENGAGEMENT 186
Moreno, R., & Mayer, R. E. (2000). A coherence effect in multimedia learning: The case
for minimizing irrelevant sounds in the design of multimedia instructional
messages. Journal of Educational Psychology, 92(1), 117–125.
doi:10.1037//0022-0663.92.1.117
Moreno, R., & Mayer, R. E. (2002). Learning science in virtual reality multimedia
environments: Role of methods and media. Journal of Educational Psychology,
94(3), 598–610. doi:10.1037//0022-0663.94.3.598
Moreno, R., & Mayer, R. E. (2004). Personalized messages that promote science learning
in virtual environments. Journal of Educational Psychology, 96(1), 165–173.
doi:10.1037/0022-0663.96.1.165
Moreno, R., & Mayer, R. E. (2005). Role of guidance, reflection, and interactivity in an
agent-based multimedia game. Journal of Educational Psychology, 97(1), 117–
128. doi:10.1037/0022-0663.97.1.117
Mousavi, S. Y., Low, R., & Sweller, J. (1995). Reducing cognitive load by mixing
auditory and visual presentation modes. Journal of Educational Psychology,
87(2), 319–334.
Myers, K. K., & Sadaghiani, K. (2010). Millennials in the workplace: A communication
perspective on millennials’ organizational relationships and performance. Journal
of Business and Psychology, 25(2), 225–238. doi:10.1007/s10869-010-9172-7
Nakamura, J., & Csikszentmihalyi, M. (2002). The concept of flow. In C. R. Snyder & S.
J. Lopez (Eds.), Handbook of positive psychology (pp. 89–105). New York, NY:
Oxford University Press.
PRESENCE AND ENGAGEMENT 187
Nakatsu, R., Rauterberg, M., & Vorderer, P. (2005). A new framework for entertainment
computing: From passive to active experience. In F. Kishino, Y. Kitamura, H.
Kato, N. Nagata (Eds.), Entertainment Computing-ICEC 2005 Fourth
International Conference on Entertainment Computing. Lecture Notes in
Computer Science, 3711, 1–12.
Nass, C., & Moon, Y. (2000). Machines and mindlessness: Social responses to
computers. Journal of Social Issues, 56(1), 81–103. doi:10.1111/0022-
4537.00153
Niemeyer, D. C. (2003). Hard facts on smart classroom design. Lanham, MD: Rowman
& Littlefield.
Norman, D. A. (1990). The design of everyday things (Reprint). New York, NY:
Doubleday.
Norman, D. A., & Bobrow, D. G. (1975). On data-limited and resource-limited processes.
Cognitive Psychology, 7(1), 44–64.
Novak, T. P., Hoffman, D. L., & Yung, Y.-F. (1998). Modeling the structure of the flow
experience among web users. Presented at the INFORMS Marketing Science and
the Internet Mini-Conference MIT.
Novak, T. P., Hoffman, D. L., & Yung, Y.-F. (2000). Measuring the customer experience
in online environments: A structural modeling approach. Marketing Science,
19(1), 22–42.
Oliveira, I., Tinoca, L., & Pereira, A. (2011). Online group work patterns: How to
promote a successful collaboration. Computers & Education, 57(1), 1348–1357.
doi:10.1016/j.compedu.2011.01.017
PRESENCE AND ENGAGEMENT 188
Oudart, J.-P. (1978). Cinema and suture. Screen, 18(4), 35–47.
Oyserman, D. D., Terry, K. K., & Bybee, D. D. (2002). A possible selves intervention to
enhance school involvement. Journal of Adolescence, 25(3), 313–326.
doi:10.1006/jado.2002.0474
Paas, F., Gog, T., & Sweller, J. (2010). Cognitive load theory: New conceptualizations,
specifications, and integrated research perspectives. Educational Psychology
Review, 22(2), 115–121. doi:10.1007/s10648-010-9133-8
Palmer, M. T. (1995). Interpersonal communication and virtual reality: Mediating
interpersonal relationships. In F. Biocca, M. R. Levy (Eds.), Communication in
the Age of Virtual Reality (pp. 277–302). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Park, B., Moreno, R., Seufert, T., & Brünken, R. (2011). Does cognitive load moderate
the seductive details effect? A multimedia study. Computers in Human Behavior,
27(1), 5–10. doi:10.1016/j.chb.2010.05.006
Piaget, J. (1971). Science of education and the psychology of the child. (D. Coltman,
Trans.). New York, NY: Prentice Hall.
Picciano, A. G. (2002). Beyond student perceptions: Issues of interaction, presence, and
performance in an online course. Journal of Asynchronous Learning Networks,
6(1), 21–40.
Pintrich, P. R. (2003). A motivational science perspective on the role of student
motivation in learning and teaching contexts. Journal of Educational Psychology,
95(4), 667–686. doi:10.1037/0022-0663.95.4.667
PRESENCE AND ENGAGEMENT 189
Pliner, S. M., & Johnson, J. R. (2010). Historical, theoretical, and foundational principles
of universal instructional design in higher education. In M. Adams, W. J.
Blumenfeld, C. R. Castañeda, H. W. Hackman, M. L. Peters, & X. Zúñiga (Eds.),
Readings for diversity and social justice (pp. 474–476). New York, NY:
Routledge.
Pogue, D. (2013). Out with the real. Scientific American, 308(2), 29.
doi:10.1038/scientificamerican0213-29
Prober, C. G., & Khan, S. (2013). Medical education reimagined: A call to action.
Academic Medicine, 88(10), 1407–1410. doi:10.1097/ACM.0b013e3182a368bd
Radford, A. W. (2011). Learning at a distance: Undergraduate enrollment in distance
education courses and degree programs (Statistics in Brief: NCES 2012-154).
Washington, DC: United States Department of Education.
Reeves, B., & Nass, C. (1996). The media equation: How people treat computers,
television, and new media like real people and places. New York, NY:
Cambridge University Press.
Reinwein, J. J. (2012). Does the modality effect exist? and if so, which modality effect?
Journal of Psycholinguistic Research, 41(1), 1–32. doi:10.1007/s10936-011-
9180-4
Rice, R. E. (1992). Task analyzability, use of new media, and effectiveness: A multi-site
exploration of media richness. Organization Science, 3(4), 475–500.
Riva, G., & Waterworth, J. A. (2003). Presence and the self: A cognitive neuroscience
approach. Presence Connect, 3(3).
PRESENCE AND ENGAGEMENT 190
Rose, D. H., & Meyer, A. (2002). Teaching every student in the digital age: Universal
design for learning. Alexandria, VA: Association for Supervision and Curriculum
Development.
Rush, K. L. (2008). Connecting practice to evidence using laptop computers in the
classroom. Computers, Informatics, Nursing, 26(4), 190–198.
doi:10.1097/01.NCN.0000304803.38746.ad
Russell, M., & Higgins, D. B. J. (2004). Laptop learning: A comparison of teaching and
learning in upper elementary classrooms equipped with shared carts of laptops
and permanent 1:1 laptops. Journal of Educational Computing Research, 30(4),
313–330.
Salkind, N. J. (2012). Tests and measurement for people who (think they) hate tests and
measurement (2nd Ed.). Los Angeles, CA: SAGE Publications.
Sanchez, C. A., & Wiley, J. (2006). An examination of the seductive details effect in
terms of working memory capacity. Memory & Cognition, 34(2), 344–355.
Schneider, W., & Shiffrin, R. M. (1977). Controlled and automatic human information
processing: I. Detection, search, and attention. Psychological Review, 84(1), 1–66.
doi:10.1037/0033-295X.84.1.1
Schunk, D. H. (2008). Metacognition, self-regulation, and self-regulated learning:
Research recommendations. Educational Psychology Review, 20(4), 463–467.
doi:10.1007/s10648-008-9086-3
Schunk, D. H., Pintrich, P. R., & Meece, J. L. (2008). Motivation in education: Theory,
research, and applications (Third edition.). Upper Saddle River, NJ:
Pearson/Merrill Prentice Hall.
PRESENCE AND ENGAGEMENT 191
Schwalbe, C. B. (2009). Leveraging the digital media habits of the millennials: Strategies
for teaching journalism courses. Southwestern Mass Communication Journal,
25(1), 53–68.
Seymour, H., Ashton, J., & Edwards, P. (1986). Health museums or theme parks: A new
approach to intersectoral collaboration. Health Promotion (Oxford, England),
1(3), 311–317.
Sheridan, T. B. (1992). Musings on telepresence and virtual presence. Presence:
Teleoperators and Virtual Environments, 1(1), 120–126.
Sheridan, T. B. (1994). Further musings on the psychophysics of presence. IEEE
International Conference on Systems, Man and Cybernetics (Vol. 2, pp. 1073–
1077). IEEE. doi:10.1109/ICSMC.1994.399986
Shernoff, D. J., Csikszentmihalyi, M., Shneider, B., & Shernoff, E. S. (2003). Student
engagement in high school classrooms from the perspective of flow theory.
School Psychology Quarterly, 18(2), 158.
Shield, B., Greenland, E., & Dockrell, J. (2010). Noise in open plan classrooms in
primary schools: a review. Noise & Health, 12(49), 225–234. doi:10.4103/1463-
1741.70501
Shiffrin, R. M., & Schneider, W. (1977). Controlled and automatic human information
processing: II. Perceptual learning, automatic attending, and a general theory.
Psychological Review, 84(2), 127–190. doi:10.1037/0033-295X.84.1.1
Short, J., Williams, E., & Christie, B. (1976). The social psychology of
telecommunications. New York, NY: John Wiley & Sons.
PRESENCE AND ENGAGEMENT 192
Sinclair, M. F., Christenson, S. L., Lehr, C. A., & Anderson, A. R. (2003). Facilitating
student engagement: Lessons learned from Check & Connect longitudinal studies.
The California School Psychologist, 8, 29–42.
Skinner, E. A., & Belmont, M. J. (1993). Motivation in the classroom: Reciprocal effects
of teacher behavior and student engagement across the school year. Journal of
Educational Psychology, 85(4), 571.
Skinner, E., Furrer, C., Marchand, G., & Kindermann, T. (2008). Engagement and
disaffection in the classroom: Part of a larger motivational dynamic? Journal of
Educational Psychology, 100(4), 765–781. doi:10.1037/a0012840
Smith, D. (2002). The theory heard ’round the world. Monitor on Psychology, 33(9), 30–
33.
Smith, K. A., Sheppard, S. D., Johnson, D. W., & Johnson, R. T. (2005). Pedagogies of
engagement: Classroom-based practices. Journal of Engineering Education,
94(1), 87–101.
Smith, P. J., & Smith, S. N. (1999). Differences between Chinese and Australian
students: some implications for distance educators. Distance Education, 20(1),
64–80.
Spinoza, B. (1677/1992). The ethics; Treatise on the emendation of the intellect; selected
letters. (2nd ed.). (S. Shirley, Trans.) Indianapolis, IN: Hackett Publishing.
Spring, J. (2008). Globalization of education: An introduction. New York, NY:
Routledge.
PRESENCE AND ENGAGEMENT 193
Sproull, N. (1973). Visual attention, modeling behaviors, and other verbal and nonverbal
meta-communication of prekindergarten children viewing Sesame Street.
American Educational Research Journal, 10(2), pp. 101–114.
Stacey, E. (2007). Collaborative learning in an online environment. The Journal of
Distance Education/Revue de l'Éducation à Distance, 14(2), 14–33.
Stanton, D., Bayon, V., Neale, H., Ghali, A., Benford, S., Cobb, S., … Pridmore, T.
(2001). Classroom collaboration in the design of tangible interfaces for
storytelling (pp. 482–489). Presented at the Proceedings of the SIGCHI
conference on Human factors in computing systems.
Steiger, J. H., & Lind, J. M. (1980). Statistically based tests for the number of common
factors. Presented at the Annual meeting of the Psychometric Society, Iowa City,
IA.
Sulbaran, T., & Baker, N. (2001). A study of engineering students interaction with
distributed virtual reality in the classroom. Presented at the 2000 ASEE Southeast
Section Conference.
Swan, K. (2001). Virtual interaction: Design factors affecting student satisfaction and
perceived learning in asynchronous online courses. Distance Education, 22(2),
306–331.
Sweller, J. (1994). Cognitive load theory, learning difficulty, and instructional design.
Learning and Instruction, 4(4), 295–312. doi:10.1016/0959-4752(94)90003-5
Sweller, J., & Chandler, P. (1994). Why some material is difficult to learn. Cognition and
Instruction, 12(3), 185–233. doi:10.2307/3233760
PRESENCE AND ENGAGEMENT 194
Sweller, J., Van Merrienboer, J. J., & Paas, F. G. (1998). Cognitive architecture and
instructional design. Educational Psychology Review, 10(3), 251–296.
Tao, L., & Zhang, M. (2013). Understanding an online classroom system: Design and
implementation based on a model blending pedagogy and HCI. IEEE
Transactions on Human-Machine Systems, 43(5), 465–478.
doi:10.1109/THMS.2013.2281436
Tellegen, A., & Atkinson, G. (1974). Openness to absorbing and self-altering experiences
(“absorption”), a trait related to hypnotic susceptibility. Journal of Abnormal
Psychology, 83(3), 268.
Thompson, W. F., Schellenberg, E. G., & Letnic, A. K. (2012). Fast and loud background
music disrupts reading comprehension. Psychology of Music, 40(6), 700–708.
doi:10.1177/0305735611400173
Tudge, J. R., & Winterhoff, P. A. (2010). Vygotsky, Piaget, and Bandura: Perspectives
on the relations between the social world and cognitive development. Human
Development, 36(2), 61–81.
Uline, C., & Tschannen-Moran, M. (2008). The walls speak: the interplay of quality
facilities, school climate, and student achievement. Journal of Educational
Administration, 46(1), 55–73. doi:10.1108/09578230810849817
van Merriënboer, J. J. G., Kirschner, P. A., & Kester, L. (2003). Taking the load off a
learner's mind: Instructional design for complex learning. Educational
Psychologist, 38(1), 5–13. doi:10.1207/S15326985EP3801_2
Vorderer, P., Klimmt, C., & Ritterfeld, U. (2004). Enjoyment: At the heart of media
entertainment. Communication Theory, 14(4), 388–408.
PRESENCE AND ENGAGEMENT 195
Vygotsky, L. S. (1926/1997). Educational Psychology. (R. Silverman, Trans.) Boca
Raton, FL: CRC Press.
Warburton, S. (2009). Second Life in higher education: Assessing the potential for and
the barriers to deploying virtual worlds in learning and teaching. British Journal
of Educational Technology, 40(3), 414–426. doi:10.1111/j.1467-
8535.2009.00952.x
Webster, J., Trevino, L. K., & Ryan, L. (1994). The dimensionality and correlates of flow
in human-computer interactions. Computers in Human Behavior, 9(4), 411–426.
Weinstein, C. S. (1979). The physical environment of the school: A review of the
research. Review of Educational Research, 49(4), 577–610.
doi:10.3102/00346543049004577
Wood, E., & Wolf, B. (2008). Between the lines of engagement in museums: Indiana
University and the Children’s Museum of Indianapolis. The Journal of Museum
Education, 33(2), 121–130.
Wylie, R. C. (1974). The Self-concept: A review of methodological considerations and
measuring instruments. Lincoln, NE: University of Nebraska Press.
Yang, J. C., & Lin, Y. L. (2010). Development and evaluation of an interactive mobile
learning environment with shared display groupware. Educational Technology &
Society, 13(1), 195–207.
Yee, N., Bailenson, J. N., Urbanek, M., Chang, F., & Merget, D. (2007). The unbearable
likeness of being digital: The persistence of nonverbal social norms in online
virtual environments. CyberPsychology & Behavior, 10(1), 115–121.
doi:10.1089/cpb.2006.9984
PRESENCE AND ENGAGEMENT 196
Appendix A:
Original Early Sense of Presence Inventory
Media Questionnaire
Thank you very much for agreeing to complete this questionnaire.
The questions on these pages ask about the media experience you just had. This may
have been watching television, watching an IMAX or Omniverse film, or using a virtual
reality (VR) system.
There are no right or wrong answers; please simply give your first impressions and
answer all of the questions as accurately as possible, even questions that may seem
unusual or to not apply to the particular media experience you just had. For example, in
answering a question about how much it felt like you were "inside the environment you
saw/heard," base your answer on your feeling rather than your knowledge that you were
not actually inside that environment.
Throughout the questions, the phrases "the environment you saw/heard" and "objects,
events, or people you saw/heard" refer to the things or people that were presented in the
media experience, not your immediate physical surroundings (i.e., the actual room you
were in during the media experience).
Please circle the responses that best represent your answers. All of your responses will
be kept strictly confidential.
__________________________________
How much did it seem as if the objects and people you saw/heard had come to the place you
were?
Not at all 1 2 3 4 5 6 7 Very much
How much did it seem as if you could reach out and touch the objects or people you saw/heard?
Not at all 1 2 3 4 5 6 7 Very much
How often when an object seemed to be headed toward you did you want to move to get out of
its way?
Never 1 2 3 4 5 6 7 Always
To what extent did you experience a sense of 'being there' inside the environment you
saw/heard?
Not at all 1 2 3 4 5 6 7 Very much
To what extent did it seem that sounds came from specific, different locations?
Not at all 1 2 3 4 5 6 7 Very much
How often did you want to or try to touch something you saw/heard?
Never 1 2 3 4 5 6 7 Always
Did the experience seem more like looking at the events/people on a movie screen or more like
looking at the events/people through a window?
Like a movie screen 1 2 3 4 5 6 7 Like a window
PRESENCE AND ENGAGEMENT 197
PRESENCE AND ENGAGEMENT 198
PRESENCE AND ENGAGEMENT 199
PRESENCE AND ENGAGEMENT 200
PRESENCE AND ENGAGEMENT 201
Appendix B:
Modified Early Sense of Presence Inventory (as distributed)
PRESENCE AND ENGAGEMENT 202
PRESENCE AND ENGAGEMENT 203
PRESENCE AND ENGAGEMENT 204
PRESENCE AND ENGAGEMENT 205
Appendix C:
Early Sense of Presence Inventory Subscales
PRESENCE AND ENGAGEMENT 206
PRESENCE AND ENGAGEMENT 207
Abstract (if available)
Abstract
The sense of presence—a feeling of engagement with a mediated or artificial experience that so closely mirrors the experience of its natural, unmediated analogue that the perceived difference becomes blurred or lost completely (Lee, 2004)—has been extensively studied in fields such as computer science and telecommunications, but rarely in an education context. The present study sought to examine the influence of presence on learning engagement, using mixed methods to assess both constructs among master’s degree students in both online and on‐campus class settings. A self‐report instrument that combined an established learning engagement measure with an established presence measure was distributed to participants (n = 148
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Examining the use of online storytelling as a motivation for young learners to practice narrative skills
PDF
An examination of classroom social environment on motivation and engagement of college early entrant honors students
PDF
Creating effective online educational content: an evaluation study of the influences affecting course developers’ abilities to create online content with engaging learning strategies
PDF
Will school-based online faculty development be an effective tool for their professional growth?
PDF
Instructional designers in the new learning paradigm: COVID-19’s impact on roles and responsibilities of instructional designers
PDF
The sport of learning: the effect of college athletes' perception of identity on approach to learning
PDF
Middle-management's influence on employee engagement in the ambulatory practices
PDF
A residential living and learning approach to successful veteran transition: an innovation study
PDF
Examining the learning environments of urban high school educators who are culturally aware and serve a majority of students from historically marginalized populations
PDF
Online graduate-level student learning and engagement: developing critical competencies for future leadership roles: an evaluation study
PDF
The perceived barriers of nursing faculty presence in an asynchronous online learning environment: a case study
PDF
Learning 21st century skills In a multicultural setting
PDF
PowerPoint design based on cognitive load theory and cognitive theory of multimedia learning for introduction to statistics
PDF
A study of the pedagogical strategies used in support of students with learning disabilities and attitudes held by engineering faculty
PDF
Student engagement in online learning: examining undergraduate student engagement in online learning communities to improve instruction
PDF
Civic engagement of Teach For America alumni in Los Angeles
PDF
An examination of the direct/indirect measures used in the assessment practices of AACSB-accredited schools
PDF
Early identification of a learning disability and its impact on success in postsecondary education
PDF
Collaborative social networks in student affairs: an exploration of the outcomes and strategies associated with cross‐institutional collaboration
PDF
Analysis of student cyclic responses and tri-engagement levels in a global learning environment
Asset Metadata
Creator
Smith, Scott A.
(author)
Core Title
Examining presence as an influence on learning engagement
School
Rossier School of Education
Degree
Doctor of Education
Degree Program
Education
Publication Date
06/19/2014
Defense Date
05/02/2014
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
classroom design,Distance education,educational psychology,learning engagement,OAI-PMH Harvest,online classrooms,pedagogy,presence,sense of presence,Technology,text chat,virtual learning environments
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Tobey, Patricia Elaine (
committee chair
), Early, Sean (
committee member
), Yates, Kenneth A. (
committee member
)
Creator Email
magicspark@mac.com,sas@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-422183
Unique identifier
UC11286010
Identifier
etd-SmithScott-2561.pdf (filename),usctheses-c3-422183 (legacy record id)
Legacy Identifier
etd-SmithScott-2561.pdf
Dmrecord
422183
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Smith, Scott A.
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
classroom design
educational psychology
learning engagement
online classrooms
pedagogy
sense of presence
text chat
virtual learning environments