University of Southern California Dissertations and Theses

Individual motivation loss in group settings: An exploratory study of the social -loafing phenomenon

(USC Thesis Other)

Individual motivation loss in group settings: An exploratory study of the social -loafing phenomenon

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INDIVIDUAL MOTIVATION LOSS
IN GROUP SETTINGS:
AN EXPLORATORY STUDY
OF THE
SOCIAL-LOAFING PHENOMENON
by
Jeffrey A. Miller
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Educational Psychology)
May 2001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3027754
UMI'
UMI Microform 3027754
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
The Graduate School
University Park
LOS ANGELES, CALIFORNIA 900 8 9 4 6 9 5
This dissertation, w ritten b y
Jeffrey A . Miller .........
Under th e direction o f D issertation
C om m ittee, an d approved b y a ll its m em bers,
has been p resen ted to an d accepted b y The
Graduate School, in p a rtia l fulfillm ent o f
requirem ents fo r th e degree o f
DOCTOR OF PHILOSOPHY
________________
Dean o f Graduate Studies
D ate ' May 1 1 , 2001
DISSER TA TIQ N C O M M ITTEE
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
This dissertation is dedicated, in the most loving memory, to my mother. She was my
first teacher and instilled the values to persist through even the most difficult of
circumstances. This experience tested that lesson. Mom, I think of you every day.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgements
First and foremost, I would be remiss if I didn’t begin this section with
recognition of my wife who not only pushed me to persist when I wanted to quit and
scream, but believed in me. I think that she understands the sources of self-efficacy
better than most people on the planet. Annette, I thank you every day for being such
a support and partner through the highs and lows.
This was a painful arduous task that often had me focused on negative
frustrations and just when I needed to smile or laugh or play with blocks, Flynn and
Edie showed up. You are the lights of hope and love. I can only hope that you will
someday derive joy from your children as I do from you.
Dr. Richard Clark has been more than a mentor to me over the past few years
and without him, I would not be finished with this degree. His friendship and
honesty will never be forgotten. Dr. Gretchen Guiton did far more than simply serve
on my committee. She served as a role model for how instructors should relate with
stressed out graduate students.
My family has always been an integral part of my life and I couldn’t have
earned this degree without the help of my father, Ken. Deep down, he always
believed in me and hoped I would do something on a grand scale. He always tried to
instill the importance of higher education and I am grateful for that. Also, I wanted to
acknowledge Judy Barnett. My cousin that was the impetus for my entering the
teaching profession and her husband Pat.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Iv
Ray, Bette, Aaron, Raymond, Courtney, and my “other family”-1 love you
all more than you will know and thank you for just being you. James “Mouse”
Moore I owe eternal thanks and gratitude for your words of wisdom and friendship.
Always keep your perspective. To some fellow graduate students namely Amy,
Steve, Kate, and Terry- thanks and good luck.
Thank you to Dr. Robert Pauli of Pepperdine University. Had it not been for
your Educational Psychology course I never would have even thought about
enrolling in graduate school. To Albert Bandura, Frank Pajares, Barry Zimmerman,
Ed Deci- your research and electronic advice was eye-opening.
Finally, I want to end this process by hoping one day soon all children
regardless of socioeconomic status, gender, or race will be afforded the opportunity
to pursue higher education with passion and efficacy that their accomplishments are
real and the rewards awaiting them are tangible. We all should have that right not in
words but in reality.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
Dedication.............................................. ii
Acknowledgements ......................................................... iii
List of Tables and Figures .......................... viii
Abstract.................................................. x
CHAPTER 1
Literature Review. ........................... 1
Historical Overview........................... 2
Motivational Theories .......... 4
Social impact theory............................................. .........4
Arousal reduction theory .......................................... 5
Evaluation potential .......... 8
Matching of effort ........ 8
Expectancy-value theory..................................... 10
Self-efficacy theory...................... 12
Collective-effort model. ....................................................... 17
Collective efficacy.................................................... 19
Perceived loafing................................ .........................................21
Cohesiveness and collective efficacy. ........................................ 22
Independent Variables............................ 24
Identifiability and evaluation potential.................. — 24
Coworker expectation........................................ 25
Collective efficacy..................... 27
Summary of the Literature ............. 30
Research Hypotheses.. ....................................................... 33
CHAPTER 2
Methodology..................... 39
Design of study ........................................... 39
Original design. ................................... ..........................................
Final design.......................................................................................
Subjects .......... ...40
Instrumentation ........................................... 40
Independent Variables............................ 41
Self-efficacy........................... .......................................................
Persistence ...........................................................
Identifiability... ........................................................................
Collective efficacy............................................ 42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
Dependent Variables................ 43
Effort............. .....................................................................
Method of Data Collection. .................................................. 43
Procedure...................................................... 44
Differences between previous studies and the replication... 48
Difference in the conditions................................................ 48
Differences in task presentation format............................... 49
General differences ...... 50
CHAPTER 3
Results. ......................................................... 52
Descriptive Statistics................. 52
Frequencies. ..................... 53
Correlations............... 57
Check of Identifiability Manipulation ............................ 62
Hypothesis 1............................. 63
Hypothesis 2...... 65
Hypothesis 3.............................. 66
Hypothesis 4.................................................... 67
Hypothesis 5 ................. 68
Hypothesis 6................................................................... 69
Hypothesis 7..................................... .70
Hypothesis 8........................ 71
Hypothesis 9...... 72
Other Statistical Considerations............................. 73
CHAPTER 4
Discussion ...................... 76
Difficulties With This Study ............................ 77
Problems with SEM........................ 78
Removed variables ............ 79
Other concerns. ................................................. 81
Findings of the Study............................. 82
Areas for Improvement.................................... 86
Implications and Directions for Future Research.......................................88
References. ............................................................................ 93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vii
Appendices
A- State Loafing Scale for Children. ..... 103
B- Structural Equation Models. .............. 108
Low-identifiability condition.................................................................... 109
High-Identifiability condition. ...... 110
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Tables and Figures
Table 1 —Descriptive Statistics of Variables Among Whole Subjects ...........54
Table 2—Descriptive Statistics of Variables Among
Low Identifiability Subjects. ...... 55
Table 3—Descriptive Statistics of Variables Among
High Identifiability Subjects...................................................... 56
Table 4 —Frequency of Gender in Whole Group and Both Conditions. ............ 57
Table 5—Frequency Table for Identifiability Manipulation for Both Groups 58
Table 6—Correlations Among Motivational Variables in Whole Group 59
Table 7—Correlations Among Motivational Variables in Low Identifiability....6 0
Table 8—Correlations Among Motivational Variables in High Identifiability....61
Table 9—T-Test to Verify Manipulation of Identifiability................................... 62
Table 1 0 -— Correlational Differences Between Groups Based on
Self-Efficacy-Effort......................................................................... 63
Table 11—T-Test for Group Differences Based on Identifiability. .................64
Table 12—Partial Correlation Demonstrating Relationship Between Efficacy
and Effort (Controlling for Persistence). ...........................................65
Table 13—Correlational Data Relating Self-Efficacy and Persistence.................67
Table 1 4 — Correlational Data for Collective Efficacy and
Effort- high identifiability. ...................69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 15— Correlational Data for Collective Efficacy and
Effort- low identifiability. ........ 70
Table 16— Correlational Data Relating Self-Efficacy and Persistence................71
Table 17— Correlational Data Relating Collective-Efficacy and Persistence 72
Table 18— Correlational Data Relating Self- Efficacy and Collective-Efficacy.. 73
Table 19— Correlational Data Relating Self- Efficacy and Collective-Efficacy.. 74
Figure 1 — Hypothesized model to examine social loafing effect................. 42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X
Abstract
This study examined individual motivation loss within group settings. Over 200
middle school students from an urban Los Angeles community were randomly
assigned to one of two conditions and completed a brainstorming task to assess
reduction of individual effort within group tasks. Self-efficacy, persistence, and
collective efficacy were evaluated as primary independent variables of social
loafing. Self-reported levels of invested mental effort were utilized as the
dependent variable. Subjects worked collectively to generate novel uses for a
spoon in conditions where they perceived their individual contributions to be
either (a) easily identifiable or (b) difficult to identify. Subjects worked under a
time constraint. Previous research indicated the most effective means to eliminate
the social loafing phenomenon was increased individual identifiability.
Correlation and partial correlation data analysis revealed divergent findings from
previous empirical findings. Results indicated subjects in high identifiability, high
collective efficacy conditions reduced effort compared with subjects in low
identifiability conditions. Similar results were found across conditions in that
identifiability did not eliminate the loafing effect. These divergent findings
indicate a need for continued research to further understand the primary
contributors to reduced individual effort within group settings, specifically
analyzing the emergent group-level construct of collective efficacy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
Chapter 1
Literature Review
Many of life’s challenges are met working collectively. Working within a
group is an experience an individual encounters repeatedly during a lifetime. Yet, in
spite of intense exposure to group experience, individuals have a tendency to lower
their motivational input when working in collective groups (Karau & Williams,
1993). As individuals progress through the various stages of life, group interaction
skills and productivity are reinforced. Social interaction is often a priority in
preschool curriculums, and projects dealing with group accountability begin to
surface in the elementary years and become enhanced in secondary school. These
patterns of effective group-work skills translate into important work skills as
individuals reach adulthood and engage in group-centered tasks in the work place.
Johnson and Johnson (1987) identify the critical components of groups to include
interdependence, sense of belonging, equity of influence, and shared goals. While
this provides a broad conceptualization of what a group is, the output of a group is
the measure of its effectiveness. The purpose of this dissertation is to isolate the
causes of individual motivation loss in collective settings. Specifically, this
dissertation will present an analysis intended to define the relationship between
identifiability, self-efficacy, collective efficacy, persistence, and effort within the
social-loafing paradigm.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
This exploratory study seeks to extend previous research findings from the
laboratory to the junior high school classroom. In addition, the study will examine
the effects of collective efficacy on social loafing and mental effort.
The literature review will first provide a historical overview of social loafing,
then present relevant theoretical explanations of the social-loafing phenomenon. In
addition, a discussion of independent variables and exploratory hypotheses are
offered to examine social loafing.
Many work-related tasks rely on groups for their successful completion.
Musicians in symphonies rely on the performance of other group members to
accomplish their collective goals. Business teams bring multiple perspectives and
experiences to help achieve corporate goals.
Historical Overview
Social loafing research began in 1913 as a result of a rope-pulling test by
Ringlemann (cited in Kravitz & Martin, 1986). Male volunteers were recruited to
pull on a rope, tug-of-war fashion, in groups of varying sizes. Effort was measured
by a dynamometer that assessed total group effort. As group size increased, the
overall performance of the group was increasingly lower than the sum of the
individual performances. Ringlemann explained this decrease of group performance
as a function of reduced coordination among members. Steiner (1972) brought
attention to the hypothesis that in addition to poor coordination, reduced motivation
also may account for lowered group effort. Beyond poor coordination accounting for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
decreased performance, lack of synchronicity may account for poor performance, as
some group members may pause while others are pulling. Ringlemann requested
maximum effort be exerted for a 4- to 5-second period. However, the primary
variable was momentary maximum force, and this measure, rather than an averaged
effort throughout the time period, may invalidate Steiner.
In an attempt to distinguish further coordination loss and effort reduction,
Ingham, Levinger, Graves, and Peckham (1974) conducted a rope-pulling task in
actual and pseudo-groups, in which members were blindfolded and perceived they
were participating in a group activity when they actually were pulling alone. While
working in the pseudo-group conditions, performance still decreased as group size
increased. This finding strengthens the argument that reduced performance is caused
by effort reduction rather than solely inefficient coordination.
Latane, Williams, and Harkins (1979) replaced the rope-pulling task with
clapping and shouting. Volunteers were asked either to clap or shout loudly in a
variety of conditions, including alone and in a group of six. Repeated trials were
measured with a sound meter. In an effort to increase participation, subjects were
informed that researchers were interested in “how much noise people make in social
settings” (p. 824). As expected, as group size increased, noise increased. However,
the level of noise did not grow in proportion to the number of people. This also
provided support that a primary cause of performance reduction was effort loss
rather than coordination as originally presumed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
Since the 1970s, nearly 100 studies have focused on the social-loafing
phenomenon. There have been a variety of attempts to clarify the causal agents to the
phenomenon as well as theoretical frameworks to support the social-loafing
phenomenon. In a meta-analysis of 166 studies (Karau & Williams, 1993),
individuals, across tasks, reduced their effort working collectively compared to
working coactively or alone. Social loafing has been defined as a reduction of
individual effort when working collectively compared with when working
individually or coactively (Latane et al., 1979). Karau and Williams (1993) found the
social-loafing effect to be moderate in magnitude and generalizable across tasks.
Motivational Theories
Social impact theory. Social-impact theory (Latane, 1981) viewed individuals
as either sources or targets of social impact. For example, an experimenter or other
authority figure could constitute a source, and group or team members served as
targets. Social impact was determined by examining the strength, immediacy, and
number of sources and targets who were present when the impact occurred. When
Latane et al. (1979) developed the social impact theory, they reasoned that as the
target-to-sources ratio increases, individual identifiability decreases, and the
opportunity to reduce individual effort increases. Therefore, social loafing occurs. As
the number of targets increase, level of responsibility remains constant, and
additional targets absorb the cumulative responsibility. Loafing and group size are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
thought to be positively correlated. The main contribution of social-impact theory is
the ability to predict group size influence on individual effort.
Arousal reduction theory. Jackson and Williams (1985) speculated that a
drive-based component to social-impact theory might be a more appropriate
explanation for loafing behavior. They argued that a similarity exists between social-
loafing research and social-facilitation research. The similarity occurs because both
areas of research are concerned with performance effects of working with others.
However, social-facilitation literature predicts that the presence of others enhances
performance on easy tasks (Zajonc, 1965). Social loafing, on the other hand, predicts
performance decrements caused by the presence of others on simple tasks. Because
of these differences, the two areas may disagree in the categorization of task
difficulty. Social facilitation researchers argue that all tasks used to date in loafing
research have been difficult. If this is true, it can be argued that when individuals are
faced with difficult tasks, arousal increases and performance suffers. This would
indicate previous findings in social-loafing research were incorrectly analyzed.
However, this assumption may be premature as tasks such as shouting, clapping, and
rope pulling cannot logically be described as difficult.
Jackson and Williams (1985) speculated that the social-loafing findings were
best explained by arousal reduction theory. By citing social-loafing research as well
as the work of Harkins and Petty (1982), it was clearly demonstrated that loafing
might be controlled but not eliminated as suggested in the literature. Previous loafing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
research found that working collectively negatively influenced achievement on
simple tasks. However, increasing task difficulty also reduced loafing by increasing
the challenge of the task (Harkins & Petty, 1982). Increasing task difficulty or
providing alternative tasks thereby reducing the perception of redundant
contributions reduced loafing. They concluded that perception of their unique
contribution accounted for participants either increasing or reducing effort.
Social-facilitation research indicates that the presence of others increases
arousal and motivates people to perform. Jackson and Williams (1985) cited the
work of Schachter (1959) to demonstrate this conclusion may not always be true.
People preferred to be in the company of others when faced with a fearful situation.
Although Schachter does not deal specifically with group motivation or achievement,
he does provide insight into the perspective of groups facing challenge or adversity.
If working on a task can be considered fearful, or slightly intimidating, these findings
may be accurate because of heightened arousal. In addition, these findings led this
study to predict that working in a group should actually be drive reducing, not drive
inducing, as previous facilitation research indicated. As the number of targets
increase, the level of individual identifiability decreases. It has been well-
documented that when working in the presence of others, people invest more mental
effort on simple tasks and disengage on complex tasks (Hunt & Hillery, 1973;
Jackson & Williams, 1985; Markus, 1978; Sanders & Baron, 1975). The prevailing
argument of the time was that working in a group was drive inducing, and this
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
heightened drive would increase the likelihood of a dominant, or most probable,
response as an outcome. Performance could be improved on difficult tasks while
working collectively by increasing drive. The dominant, or most probable, response
to a simple task is an investment of effort that leads to success. Therefore, on an easy
task for which drive is accelerated by working in front of others, the most probable
response is increased effort and decreased loafing. The dominant response to a
complex, or novel, task is a behavior that leads to poor performance, or reduced
effort and increased loafing behavior.
By subjecting participants to a computer maze manipulation, Jackson and
Williams (1985) found support for previous findings that collective grouping tends to
decrease identifiability compared with subjects working alone. However, they also
found that accompanying this reduced identifiability construct was a feeling of
reduced pressure. These drive reducing effects enabled participants working in
collective groups to perform better on the difficult maze than the collective group
working on the easy maze. They concluded that working collectively on difficult
tasks improves performance by reducing stress on individuals. They further claim
that there may be situations in which group production may benefit during difficult
tasks when working collectively, as this is drive reducing.
The concept of self-referent processes (e.g. self-efficacy, self-esteem, self-
concept) potentially contributing to increased effort during group tasks was an
emergent component in reducing social-loafing behavior. Beyond the drive-based
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
approach to defining social loafing, the influence of self-efficacy and other self
referent processes were considered as potential explanations for effort loss.
However, research (Kerr & Bruun, 1983; Williams, Harkins, & Latane, 1981)
quickly found that identifiability alone was not a significant predictor of the
elimination of loafing behavior.
Evaluation potential. One of the critical moderators of social loafing is the
identifiability perception that evaluation will follow group activity. Several
researchers (Harkins, 1987; Harkins & Jackson, 1985; Harkins & Szymanski, 1987,
1988,1989; Kerr & Bruun, 1983; Szymanski & Harkins, 1987; Williams, Harkins, &
Latane, 1981) have defined social loafing as an individual perception of
identifiability or evaluation. Karau and Williams (1993) caution against defining
social loafing in terms of its causes as that limits the opportunity to uncover other
causal agents. Szymanski and Harkins (1987) argue that the critical understanding of
the social-loafing effect is the combined effect of individual identifiability and
anticipated evaluation. This reduced performance resulting from an expectation has
been demonstrated to be effective whether the evaluator was an experimenter (Geen,
1979), or the self (Szymanski & Harkins, 1987). Historically in social-loafing
research, motivational loss has been measured with an inclusion of an identifiability
standard.
Matching of effort. Matching of effort suggests that group members seek to
achieve a motivational equity for the group. Jackson and Harkins (1985) proposed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
that individuals match the level of effort of other coworkers when working
collectively. The loafing effect occurs when individuals expect others in the group to
reduce effort and, as a result of this expectation, reduce their own effort. The
problem with Jackson and Harkins is that coworker expectation and overall
evaluation of the experiment’s worth were confounded. In the high-effort condition,
the confederate claimed the expenment was high interest and the intention to invest
effort was prominent. However, in the low-effort condition, the confederate not only
claimed she intended to invest minimal effort, but she indicated the experiment was
boring. After Williams and Karau (1991) re-examined the influence of coworker
expectation by administering a task with high meaningfulness across conditions, the
opposite effect emerged. When teamed with an individual expected to perform
poorly, invested effort increased.
In addition to conceptualizing coworker effort as a strategically planned
action or a non-conscious response, further inquiry into the measurement of effort
may clarify the intentionality of effort investment. Williams, Karau, and Bourgeois
(1993) claim that most studies use self-report instruments to measure levels of effort.
It is generally assumed that loafing is a non-conscious, unintentional process. Since
measures are typically self-reported Likert scales, accuracy of participant response
may have a negative influence on examining relationships among variables
accurately. No studies have incorporated physiological measures to assess individual
levels of invested effort. Although this level of assessment is beyond the scope of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
this study, it is possible that further exploration into the relationship between
physiological and self-reported measures may reveal interesting data to address the
question of whether loafing is strategic and influenced by co workers, or non-
conscious and unintentional.
Each of the previous theories suggest specific, individualized prescriptions to
eliminate the loafing effect. On one hand, if identifiability is increased, loafing will
be reduced, assuming the individual infers some form of evaluation will follow.
However, perceptions regarding the redundancy of individual input significantly
reduce the effort of individuals working with group tasks. What appears to be
limiting the understanding of the loafing effect is an inclusive multivariable
discussion. Several theories have presented compelling arguments to reduce the loss
of motivation, yet remain too narrow to eliminate or understand the effect.
Two specific theories explore the process of motivation appropriate to this
study. Expectancy-value theory explores the individual assessment and decision
making strategies of individuals. Self-efficacy theory, a tenet of social-cognitive and
social-learning theory, examines aspects of expectancy-based motivational concepts.
The two theories are related in that both address the students’ perception of
competence, an integral component to this study.
Expectancy-value theory. Atkinson’s (1957) original work ultimately evolved
into the Collective Effort Model (Karau & Williams, 1993). Although the work of
Atkinson draws heavily from the work of Lewin, the primary contribution of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1
Atkinson is the expectancy construct (cited in Pintrich & Schunk, 1996). Atkinson
proposed behavior to be a multiplicative function of three components: motives,
probability for success, and incentive value. Reduction, or loss of effort, occurred if
any of the three components had a value of 0. The motive component represented
learned but stable dispositional attributes with two distinct motives: the desire to
seek success or the need to avoid failure. Within each motive, are an expectation, and
perhaps an orientation, toward emotional anticipation. Covington and others devised
a quadripolar model to categorize the four types of students based on the two
motivational orientations, success orientation and failure avoidance (Covington,
1992; Covington & Omelich, 1991; Covington & Roberts, 1994). While the
recognition of these types of students is perhaps tangentially related to this study in
the construction of classroom climate and the teacher’s ability to identify individual
student orientations, it is beyond the scope of this paper to elaborate on them.
An individual’s estimated probability for success functions as an efficacy and
outcome expectancy. This can be demonstrated using a ring-toss game and asking
the subjects to estimate their probability for success. While motives may be a more
internalized expectancy measure, the probability for success may reflect more
externally based judgements. By incorporating environmental cues into methods
such as the ring-toss, an individual maintains an overall expectation by combining
motives and expectancy for success with an affective measurement, incentives. The
third component of Atkinson’s (1957) model is the incentive value of success. An
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
inverse relationship is presumed to exist between the incentive value of the task and
estimated probability for success. It was assumed that easier tasks had a higher
probability of success but a lower incentive for succeeding. Pintrich and Schunk
(1996) illustrate that this inverse relationship predicts highest level of effort on tasks
with intermediate levels of difficulty. Taken together, the perception of expectancy
and task difficulty create a construct of expectancy directly related to self-efficacy
theory.
Self-efficacy theory. Bandura (1997) emphasized the importance of self
referent thought in the overall dynamic of human capability as this method of self-
evaluation often leads to behavioral changes. One aspect of self-referent thought is
self-efficacy. Bandura (1997) defines self-efficacy as the “beliefs in one’s
capabilities to organize and execute the courses of action required to produce given
attainments” (p. 3). Perceived self-efficacy influences the courses of action
individuals choose to pursue (Betz & Hackett, 1981,1983; Hackett & Betz, 1981;
Zimmerman, 1985), the amount of invested effort toward the pursuit of this goal
(Salomon, 1984), and the level of persistence in the face of distractions (Lent,
Brown, & Larkin, 1986; Zimmerman & Ringle, 1981) in order to achieve specific
levels of attainment. While a variety of influences can affect individual self-efficacy,
it has been determined that self-efficacy has been correlated positively with a
number of achievement outcomes (Schunk, 1983,1984, 1991); including goal
setting (Button, Mathieu, & Aiken, 1996; Locke & Latham, 1990) and self
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
regulation (Bandura, 1991; Malpass, 1994; Pajares, 1996; Zimmerman & Martinez-
Pons, 1990).
Wigfield and Eccles (1992) concluded that self-efficacy is similar to other
expectancy beliefs as both examine perceived capability and anticipated success.
They differ in that self-efficacy beliefs are domain and task specific, defined as an
individual appraisal of anticipated success on a specific task. Other expectancy value
theories have questioned the boundary of certain domains in construct development.
In contrast, self-efficacy estimates have been assessed by exposing individuals to
specific types of mathematics problems and ascertaining their estimate of success on
a similar problem (Schunk, 1991). Pajares (1996) argued against creating a general
self-efficacy measure, as it would change a judgement of capability into a general
personality trait that would have poor predictive power.
As the domain generality of assessment illustrates differences between the
self-efficacy and expectancy value theories, the primary and perhaps most critical
similarity relevant to this study is the expectancy-based construct common to both.
The role self-efficacy beliefs have in establishing effort and persistence has been
categorized as an interaction between efficacy expectancies and outcome
expectancies. Efficacy expectancies concentrate on the belief that the individual is
capable of performing the required task. Outcome expectancy is the belief that a
specific set of behaviors will lead to a desired or valued outcome. This closely
resembles the expectancy-value literature in that individuals make two primary
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
analyses before consciously choosing to invest mental effort to achieve a goal. First
is whether they can achieve the goal and second is why achieving the goal is
important to them.
The importance of these two theoretical constructs became more apparent in
a study by Sanna (1992). Efficacy expectancy was manipulated by administering
false performance feedback to subjects on a specific task. Researchers demonstrated
outcome expectancy by assigning individuals to various grouping conditions (alone,
co-action, or collective) in which group assignment predicted the expected level of
identifiability and evaluation. Two experiments were conducted to reaffirm previous
findings that feedback impacted self-efficacy and in turn affected performance.
Earlier, Sanna and Shotland (1990) manipulated quality of feedback on a rote
memorization task and found that after success, subjects held higher expectations for
success than individuals receiving negative feedback. Further, they found a positive
correlation between expectancy and performance. These findings were supported by
other research that found performing simple tasks developed expectations of success
where subjects performing difficult tasks held lower expectations (Carver & Scheier,
1981). Sanna (1992) not only found support for previous findings that perception of
identifiability/evaluation affected behavior, but also indicated that efficacy
expectancy and outcome expectancy j ointly affected performance on the vigilance
task. Feedback did induce either high or low efficacy expectancies as expected.
Efficacy expectancy was significantly related to better performance. The second
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
experiment, manipulating versions of the Remote Associates Test (RAT), attempted
to create a deeper understanding of the formation of efficacy beliefs based on task
difficulty. The RAT is an assessment in which the subject is provided with three
stimulus words and asked to generate the appropriate word that completes the set.
Easier tasks produced significantly stronger efficacy expectancies than difficult
tasks. In addition to these findings, specific attention was given to the interaction
between efficacy expectancy and outcome expectancy. Subjects in the easy-list,
evaluation condition answered more items than either the alone or coactive
condition, supporting social facilitation research. However, subjects in the difficult-
list, evaluation condition answered fewer items correctly, supporting previous social-
loafing research. This study was significant in its ability to conceptualize social
loafing and social facilitation in terms of self-efficacy theory.
Several researchers focused on individuals’ behavior and motivational
orientation in the process of goal attainment (Sanna, 1992; Schunk, 1983,1984,
1991; Wigfield & Eccles, 1992). There have been several attempts to integrate
groups theoretically as collective entities to explain motivation. One such theory
synthesized motivational models and incorporated them into the established
framework of expectancy value models to create a coherent, yet incomplete, account
of collective group motivation.
Locke, Motowidlo, & Bobko (1986) presented a compelling argument for
using self-efficacy to reconcile divergent predictions regarding the relationship of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
expectation and performance in expectancy-value and goal-setting literature. Goal-
setting literature predicts a linear relationship between goal difficulty and
performance. This assumes that low outcome expectancy breeds high performance.
On the other hand, expectancy-value theories argue performance is an outcome of
three variables: expectancy, instrumentality, and valence. As self-efficacy is an
expectancy construct, it tends to affect effort positively at two different stages: goal
choice and performance. Therefore, self-efficacy, along with valence, would affect
goal choice and goal commitment (Locke & Shaw, 1984). Performance would be
affected by self-efficacy, goals, and feedback. While the Locke et al. (1986) model
intentionally omitted variables that would affect performance (such as ability), there
is support for it in the literature (Taylor, Locke, Lee, & Gist, 1984).
While self-efficacy theory has been demonstrated to provide a framework to
better understand the social loafing phenomenon, what emerges from the theory is
the individual’s role within the group. As social loafing is a group-level construct,
and evaluation of individual effort reduction is critical to understand the entire
phenomenon, an important gap remains when considering the group level efficacy
construct that must be analyzed to clearly understand the entire process of reduced
motivation. Further, from simply using individual efficacy scores to conceptualize
the loafing phenomenon, a glaring omission of how the group expectancy construct
influences the final investment of effort emerges. Sanna (1992) clearly demonstrate
the need for identifying individual efficacy expectancy in conceptualizing social
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
loafing, however, the critical omission of the group level construct highlights the
need for the research presented in this study.
Collective-effort model. Social-loafing studies completed after 1993 (Karau
& Hart, in press; Karau & Williams, 1997) predominantly utilize this collective
model of motivation that stems from earlier work on expectancy-value models
(Vroom, 1964) of work motivation. By combining expectancy-value models and
self-evaluation processes, Karau and Williams (1993) developed the Collective-
Effort Model (CEM). The model suggests that individuals will exert effort on a task
only to the degree they expect their efforts to be instrumental in obtaining a group
outcome they personally value. Immediately, a potential confound emerges within
using the model in experimental studies. Primarily, this demonstrates a need for an
intrinsically motivating task that will heighten the value participants assign to the
task. The model predicts that if the task is low in meaningfulness or interest,
individuals are not motivated to invest effort. The group product is reliant on the
individual outcome orientations of each group member. Karau and Williams
integrated the predominant factors affecting the magnitude of the social-loafing
effect into their model in a meta-analysis. Therefore, they predict social loafing to be
reduced on the grounds of previously created theories. For example, the CEM
predicts loafing to be reduced based on: (a) the perception that collective
performance will be evaluated by others (Harkins, 1987; Harkins & Jackson, 1985;
Harkins & Szymanski, 1987), (b) reduction of group size (Latane, 1981), (c)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
redundancy of contributions (Kerr, 1983), (d) establishing of a standard (Harkins &
Szymanski, 1988), (e) work on tasks that are high in meaningfulness or valence
(Wigfield & Eccles, 1992), (f) heightened group identity (Abrams & Hogg, 1990),
(g) control of coworker expectation (Jackson & Harkins, 1985), and (h) positive
outcome expectations (Bandura, 1986).
While this model tends to predict collective outcomes based on accepted
research in the tradition of expectancy-value models, there are certain critical aspects
of the group dynamic that are inadvertently omitted. While Sanna (1992) introduced
the potential of self-efficacy as a framework to understanding social loafing, it is the
more recent work of collective efficacy (Bandura, 1997) that merits special attention
as it attempts to uncover the underlying structure of group motivation.
Karau and Williams (1997) provide a clear, well-synthesized model that
integrate the crux of social-loafing findings over the previous 30 years. Primarily,
this model identifies that valence is the critical variable in eliciting individual
motivation within group tasks. While this is a critical distinction to provide, it fails to
address the unfortunate reality that many tasks are not intrinsically interesting to
students. It is for this reason that the same researchers that provide the Collective
Effort Model also suggest continually that individuals may prefer tasks that they can
loaf. This contradiction again highlights they need to explore a critical group-level
variable that may account for the loafing phenomenon. Further, the paucity of
research in the social-loafing field highlights the urgency to evaluate whether there is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
a variable that considers group level expectancy and self-efficacy. Again, the need
for the research in this study is identified as filling a glaring gap in the social loafing
research.
Collective efficacy. Collective efficacy is defined as “a group’s shared belief
in its conjoint capabilities to organize and execute courses of action required to
produce given levels of attainments” (Bandura, 1997, p. 477). While there have been
other attempts to define this variable in relation to organizational climates (Lindsley,
Brass, & Thomas, 1995; Mulvey & Klein, 1998), these definitions lack the
individual assessment of ability as well as the belief in the groups potential to
succeed at a specific task. This construct has not been sufficiently researched in
relation to social loafing to draw specific conclusions regarding the magnitude or
ability to generalize the effect. However, enough has been written and researched to
warrant an inquiry into its role in controlling, or at least, explaining the loafing
effect.
Collective efficacy is a function of two components, the first composed of
aggregating each member’s individual assessment of their personal capabilities
(Bandura, 1997). Bandura is quick to point out that beliefs of personal efficacy are
not detached from the larger group entity. It would be foolhardy to assume an
individual in a collective environment would determine their level of personal
efficacy without regard to the abilities or potential contributions of the other
members. A conductor must consider the influence other members of the symphony
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
have in helping him succeed. Not only does each member need to maintain a strong
sense of individual efficacy, but there must be a collective expectation that the task
can be completed successfully. Considering other external variables relevant to the
task becomes an important part of developing personal efficacy on a collective task.
The other component of the measure of collective efficacy is the coordinative
beliefs of the aggregated members regarding the group as a whole (Bandura, 1997).
This illustrates a fundamental difference between individual-level efficacy and
collective efficacy. The referent of the efficacy perceptions in collective efficacy is
the group (analysis) compared (to) the individual unit of analysis in traditional self-
efficacy theory (Mulvey & Klein, 1998). Bandura (1997) also argues that the
primary difference between personal and collective efficacy is the unit of agency.
Lindsley et al. (1995) defines collective efficacy as “the group’s collective belief that
it can successfully perform a specific task” (p. 648). The similarity that is identifiable
within each definition of collective efficacy is the perception of the group as a
decision-making entity that focuses on a task-specific belief system. This task-
specific construct differs from other similar collective research on collective esteem
(Crocker & Luthanen, 1990) and potency (Guzzo, Yost, Campbell, & Shea, 1993)
that are less task-specific and more global in their definition. However, Gist (1987)
argued for the importance of individual efficacy. This supports by including an
individual self-efficacy construct as part of the larger, more inclusive view of
collective efficacy. Task specificity is a critical component in understanding how
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
personal and group efficacies are interrelated, and this distinguishes it from other
group-based motivational theories discussed earlier. As with individual efficacy,
collective efficacy has similar sources of influence, determines level of invested
effort, and directs staying power or persistence based on the specific task.
The emergence of collective efficacy as a potential explanation of social
loafing brings social loafing research back to the findings of Ringlemann: that
loafing was strictly a function of coordination loss.
Judgments of personal efficacy are not omitted from this gauging of
collective efficacy. In fact, Bandura (1997) argues that the evaluation of group
processes enhances or hinders individual investment of effort. Central to the
appraisal of the coordinative capabilities of the group are efficacy expectations each
member holds for the anothers. It has yet to be determined what role, if any, status
has in creating expectations for other group members. However, collective efficacy
centers on the appraisal of others’ personal efficacies as a predictor of performance,
and status may need further consideration in empirically based studies.
Perceived loafing. Mulvey and Klein (1998) point to the descriptive study of
Comer (1995) as an illustration of perceived loafing. This construct seeks to
understand member perceptions of individual loafing rather than the effects the
loafer has on the group. In some instances, loafing may simply go unnoticed and,
therefore, group loafing may not have existed. Traditionally, in loafing research,
emphasis is placed on the individual who engages in loafing behavior. Social loafing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
by definition is an individual motivation loss in a collective setting. This definition
does not take into consideration the effects on other group members aside from the
overall collective effort and performance. It is argued that recognized perceived
loafing, as compared to unrealized loafing, has the effect of lowering expectations
for group success and hence, reduces performance. They argue that for actual loafing
to occur, individual assessment of every group member would be required. Further, it
is argued that if the loafing goes unnoticed, the individual reduction of effort would
not have a negative influence on members’ effort or performance. While this is a
valid point, it seems to misrepresent the loafing literature and is not accounted for in
the present study. Historically, loafing examines the individual loafing in the
collective task, and, as a result, group productivity is decreased. The argument
forwarded by Mulvey and Klein tends to overlook productivity, the direct outcome
of social loafing. This tenet of collective efficacy connects previous research of
cohesiveness with the more task-specific theory of efficacy.
Cohesiveness and collective efficacy. The comparison between the two
constructs of cohesiveness and collective efficacy is novel as both concepts are
relatively new variables to group dynamics. Previous research on group cohesion
identified cohesion as an effective construct (Johnson & Johnson, 1987; Karau &
Williams, 1997). Johnson and Johnson view cohesion as a trait although evidence
was not given whether this was task specific, as collective efficacy stipulates. The
level of cohesion fluctuates based on such factors as: level of trust, individuality, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
“the amount of fun members have” (p. 409). Karau and Williams define
cohesiveness as the degree to which group membership was valued. Cohesiveness
has been identified in groups of strangers or friends or coworkers (Karau & Hart, in
press; Karau & Williams, 1997).
Taken together, the previous theories attempting to clarify the causes of
social loafing fall significantly short of providing a clear picture that integrates all
applicable variables. Sanna (1992) provides a thorough analysis of self-efficacy as an
appropriate method to frame the loafing literature prior to the emergence of the
collective efficacy variable. In fact, it was after the publication of Sanna (1992) that
Bandura integrated the construct of collective efficacy into social cognitive theory.
Therefore, the inclusion of collective efficacy is a natural extension of the previous
research. As demonstrated above, collective efficacy, if measured appropriately,
provides a method to analyze individual efficacy, group level expectation of effort
and performance. This natural extension further highlights the need and relevance of
including collective efficacy as a potential defining variable in social loafing
research.
Karau and Williams (1997) did not generate novel variables to clarify social
loafing. Rather, they provide an integrated framework of previous research in an
attempt to converge prior research in one model. While this is an important
organizational strategy, little new information is generated to understand the
underlying causes of social loafing. Further, the collective effort model clearly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
identifies the need for the research in this study. By collapsing previous research into
one common model, it becomes evident that group-level expectancy is not included.
Therefore, including collective efficacy is a relevant and appropriate extension of
previous literature and stands to make a significant contribution toward
understanding the underlying factors in social loafing.
Independent Variables
Identifiability and evaluation potential. Assessment of social loafing is
contingent on the perceptions of group members. Perceptions focus on whether
individual contributions to the collective effort are identifiable and able to be
evaluated. The source of evaluation can be an experimenter, a standard, or the self.
The constructs of identifiability and evaluation were originally separate, but later
research combined the two variables into one simple construct (Harkins & Jackson,
1985). After manipulating identifiability and comparability (as a measure of
evaluation), identifiability alone was the significant predictor of the loafing effect.
Individuals needed to feel their contributions were going to be evaluated. Results
indicated that individuals in the high-identifiability condition performed significantly
better than those subjects who felt their contributions were not identifiable. Further,
participants who felt their contributions were compared with others performed better
than those whose performance was not compared to others. Harkins and Jackson
stated, “this potential for evaluation.. .could motivate performance,” (p. 462) rather
than simply prevent or delay its decay.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
Szymanski and Harkins (1987) conducted a partial replication of a
brainstorming task (Harkins & Jackson, 1985) but manipulated the evaluation
variable so subjects believed evaluation occurred on two levels: self-evaluation and
experimenter evaluation. Self-evaluation had significantly eliminated the loafing
effect. Further, no significant differences between experimenter and self-evaluative,; ; I
conditions emerged.
Not only do individuals reduce effort when identifiability is decreased, but
the opportunity for evaluation must exist. Individuals faced with a task that will be
evaluated, regardless of the source, invest more mental effort toward the completion
of the task than those receiving no evaluation.
Coworker expectation. Much has been said to this point regarding the
expectations an individual holds for themselves and others in a collective group.
However, relatively little has focused on the impact expectations of others has on the
productivity of the group. Olson (1965) studied free riding as an opportunity one has
when one is not required to work to complete a project successfully. Johnson and
Johnson (1987) address the potential to ride free in their characterization of a group
as needing to be interdependent. As demonstrated earlier, a disagreement existed
among the few published studies on coworker expectation. Kerr (1983) found
loafing would occur with the expectation of high coworker involvement. This
supports the free-riding concept. However, in a study that confounded coworker
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
expectation with task/experiment worth, Jackson and Harkins (1985) found that
individuals tended to match the efforts of their coworkers.
In a study seeking to create order to these contradictory findings, Williams
and Karau (1991) found support for the social-compensation hypothesis and for the
social-loafing hypothesis. In this study, participants were tested on interpersonal
trust, and these levels created coworker expectation conditions. They found a main
effect on trust was medium trusters were less productive than either high or low.
However, since they did not assess the reasons for low-trust compensation, a second
experiment was designed to manipulate coworker expectation. Again, they found
significant support for the theory that subjects engaged in loafing behavior in the
presence of a coworker who was expected to invest high levels of effort.
Although the results are significant, studies that conclude this variable is a
sound theoretical explanation of the social-loafing effect are lacking. There is
substantial need to assess the potency of coworker expectation in a field-based study
to generate real-world data. When individuals are expected to perform poorly, other
group members achieve group balance by compensating for the low expectancy.
However, when individuals expected their coworker to perform well (high trust),
they reduced mental effort. While coworker expectation is not directly evaluated as
an independent variable in this study, it does serve as an adequate and appropriate
introduction to collective efficacy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
Collective efficacy. Karau and Williams (1997) broadly defined cohesiveness
as a measure of acquaintanceship. In fact, the researchers combine the definition of
cohesion and cohesiveness into one variable they label cohesiveness. Cohesiveness
was identified by having students select group membership. By advising subjects to
enroll as experimental subjects with friends, high cohesiveness was attained. Groups
composed of individuals that valued the opportunity to work together were classified
as highly cohesive. Giving subjects the freedom to establish working groups was
assumed to include having a pre-specified set of values based on friendship. Previous
research on cohesiveness examined cheerleaders working with their own squad or
cheering with a different squad (Hardy & Latane, 1988). Further, Shirikashi (1985)
had members of a Japanese sport club clap and shout with other subjects that either
were members of the club or strangers. Although the hypothesis that highly cohesive
groups would perform better received only modest support, a second experiment
sought to examine specifically the effects of coworker expectation. Subjects
participated in an idea-generation task, similar to Harkins and Jackson (1985).
Within the high-ability condition, a significant interaction between grouping
condition and cohesiveness emerged. These studies and Karau and Hart (in press) are
the only studies that directly manipulate group cohesiveness.
Karau and Hart (in press) identified cohesiveness by compiling subject
responses to a social-issues questionnaire. Groups were then assigned to a grouping
condition similar to Karau and Williams (1997). They found evidence of a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
significant social-loafing effect based on cohesiveness. It is important to note the
potential shortcomings of the definition of cohesiveness to which the social-loafing
literature prescribes. In both studies specifically addressing the relationship between
cohesiveness and social loafing, the measure of group cohesiveness was established
by inquiring into the level of likeability each set of partners shared for one another,
the likelihood they would collaborate in the future, and perceptions regarding
coworker similarity.
There are several critical problems with the cohesiveness construct that limit
a more complete understanding of the underlying factors contribution to social
loafing (Karau & Williams, 1997). The cohesiveness construct omitted group
expectancy. The main component of the CEM is the relationship between outcome
valence and performance related to individuals and the group. Self-efficacy has
proven to be an antecedent to many critical educational outcomes. In addition, it has
been used as a framework to understand the expectancy component of social loafing
better (Sanna, 1992). One of the critical components to working with others is the
affective domain, and the previous studies of cohesiveness address the affective
domain sufficiently. However, as the crux of this argument rests in the social
cognitive theory, group level analysis of efficacy and environmental influences
becomes a necessity. Working on a cooperative task has an element of capability
expectation that is not addressed sufficiently.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
With this in mind, one of the goals of this study is to offer a new definition of
group cohesiveness in relation to social loafing. As discussed earlier, collective
efficacy maintains a two-dimensional component to group motivation (Bandura,
1997). Within these two dimensions, aggregated members’ perceptions of capability
and belief in the cooperative potential of the group are assessed. Not only do these
create a more meaningful understanding of group expectancy, but they offer the
opportunity to test in multiple domains on specific tasks. By utilizing this perspective
of group cohesiveness, the researcher moves away from a simple affective
description of a group and into a cognitive assessment of judging perceived efficacy.
Bandura posits that this level of group analysis moves group assessment from a
detached “group mind” (p. 478) into a system in which judgments, tasks, and
perspectives are interdependent.
Further, Bandura (in press) suggests group attainments are dependent on
interactive, coordinative, and synergistic dynamics, which support the definition of
group (Johnson & Johnson, 1987). Bandura also states that collective efficacy is not
just a summation of individual efficacy beliefs but an emergent group-level property.
For the purpose of this study, cohesiveness will be defined relative to the
components of collective efficacy. Cohesiveness will be analyzed as the result of
gaining an aggregated appraisal of individual member’s personal efficacy specific to
the task being assessed, as well as the aggregation of the group members’ appraisals
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
regarding the capability of the group to accomplish and coordinate on specific tasks
successfully.
One concern with the inclusion of this new construct is an empirically
reliable measure of collective efficacy. While beliefs of collective efficacy have been
shown to predict performance (Hodges & Carron, 1992; Spink, 1990), Bandura
(1997) argues for the strong positive correlation between collective efficacy and
effort, regardless of whether the efficacy expectancy was developed naturally or
experimentally.
Summary of the Literature
Recent research has identified an updated collectivist view of the expectancy-
value model in the form of the Collective Effort Model. In addition, Bandura (1997)
proposed the concept of collective efficacy as a way to gain a deeper understanding
of group processes. Throughout this discussion, several similarities have emerged.
Historically, the combined influence of identifiability and evaluation potential has
had a dramatic impact on the reduction of loafing behavior. Efficacy expectancies
influence an individual’s desire to choose to engage in an activity, invest effort, and
persist in the face of distractions in order to ensure successful completion of the task.
The expectations that individuals hold for themselves extend to group-level
constructs, such as collective efficacy, as individuals inherently consider their role in
the group process. A quarterback may be highly efficacious but may hold
dramatically lower group-level beliefs in other individuals or the group as a whole.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
However, individual efficacy influencing collective efficacy in social loafing is a
rational extension of the self-efficacy literature. Taken together, self-efficacy and
collective efficacy extend social-loafing research by examining the relationship
between group-level efficacy expectancies and invested mental effort within a group
task. Self-efficacy, identifiability, and basic levels of group-based outcome
expectation (coworker expectation) have at various points in social-loafing research,
surfaced as critical components to explain individual motivation loss. Introducing
collective efficacy as a predictive variable extends the literature by incorporating
task-specific group-level efficacy expectations, in addition to individual efficacy
expectations, and evaluating its influence as an emergent group-level property
(Sanna, 1992).
After an examination of the literature, it appears that out of approximately
170 studies, less than 20% have been conducted outside the laboratory. An even
smaller percentage focused on populations other than college students (Karau &
Williams, 1993). There has been minimal effort to move group-motivation-loss
literature to the classrooms or to examine motivational loss with younger, more
diverse populations. Cross-cultural studies examined social-loafing differences
between collectivist and individualistic societies (Gabrenya, Wang, & Latane, 1985).
Other studies involving children have engaged children in shouting and clapping
tasks rather than classroom tasks for which students may be held accountable
(Gabrenya, Latane, & Wang, 1983). Assessing the loafing effect in a younger
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
population and different setting extends the literature beyond the laboratory into the
real world. Further, using a highly sensitive population offers a benchmark for future
comparisons.
This study analyzes basic assumptions regarding individual effort within
group situations. By examining the relationships between self-efficacy, collective
efficacy, persistence, and effort with younger populations, practical suggestions to
improve cooperative and collective learning environments emerge. Further, clarity
regarding intra-group perceptions can facilitate teacher preparation and instructional
delivery.
The concept that loafing occurs is not in question. This study contributes
significantly to the understanding of social loafing on three levels. First, research
findings move away from the laboratory and into the real world. This study subjects
critical variables, demonstrated to control the loafing effect, to practical learning
situations with random groups of young students. This study examines a highly
sensitive population of sixth-grade students in urban Los Angeles. Second, this
dissertation examines the emergent dynamics of collective efficacy within groups to
assess the influence on effort reduction. This moves the theoretical framework of
Sanna (1992) to collective-efficacy expectations rather than solely individual-
efficacy and outcome expectancies. Finally, this study identifies whether empirical
findings on social loafing extend to young children. If social loafing is an emergent
phenomenon in young children, loafing as a socialized process can be examined. In
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
addition, findings will increase the effectiveness of such teaching strategies as
cooperative learning, peer tutoring, and jigsaw learning. The knowledge of social
loafing will be significantly enhanced.
Research Hypotheses
As this study proposes to confirm prior experimental findings as well as
suggest novel extensions with regard to collective efficacy, the presentation of the
hypotheses are separated based on participant perception of identifiability,
historically influential in loafing research. What follows is a brief explanation of
each hypothesis included in the present study. With four primary measures- self-
efficacy, effort, persistence, and collective efficacy- a 4 x 4 x 2 correlational matrix
is presented where there are hypothesized to be 12 hypotheses based on the
identifiability construct. However, the present study collects data both before and
after participating in the task. Therefore, certain distinctions between constructs are
not salient when accounting for identifiability and are discussed as having influence
across conditions.
Bandura (1982,1986,1997) discusses the critical role of self-efficacy in
understanding motivation from a social cognitive perspective. Within the discussion,
he clarifies the direct, immediate and profound influence self-efficacy has on not
only choosing to engage in a task but on the level of invested mental effort and
persistence toward attainment of a task specific goal. Sanna (1992) used the
expectancy aspect of efficacy theory to frame social loafing research.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
Harkins (1987) and Harkins and Jackson (1985) clarify the influence of
perceived identifiability in eliminating or, at least, reducing the loafing phenomenon.
By manipulating participants perceived level of identification by the observer, and
further by having them perceive their contributions would be evaluated (Szymanski
& Harkins, 1987), loafing diminishes. The exact method of creating two distinct
conditions of high and low identifiability is explained later. Taken together, it can be
hypothesized that the relationship between self-efficacy and effort will remain
positive. However, it is expected that under conditions of low identifiability, the
relationship will not be as strong as under conditions of high identifiability. The
strength of the identifiability construct is anticipated to emerge when the opportunity
to loaf increases. Therefore, under conditions of high identifiability, there will be a
significant and positive correlation between self-efficacy and effort. However, under
conditions of low identifiability, there will be a positive but not significant
correlation between self-efficacy and effort.
Persistence is generally understood as invested mental effort over time. With
this understanding, it would be expected that a strong correlation between effort and
persistence in this study emerges. Historically, identifiability has had a significant
influence in social loafing and that influence is expected in this study. Under
conditions of high identifiability, there will be a significant and positive relationship
between effort and persistence. Under conditions of low identifiability, ther will be a
positive but not significant relationship between effort and persistence.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
The emergence of collective efficacy in this study as a potential factor in
understanding the social loafing phenomenon from a group level makes its inclusion
exploratory. Therefore, the results are merely directional and potentially lend to a
greater clarity in understanding the influence collective efficacy has on a variety of
variables. Bandura (1997) argues for a strong positive correlation between collective
efficacy and effort. He further suggests this relationship should surface regardless
whether the efficacy component was developed experimentally or develops
naturally. Therefore, a natural extension in the present study is to hypothesize the
significant and highly positive correlation between collective efficacy and effort.
However, the influence of identifiability is hypothesized to influence this
relationship. The collective efficacy construct in this study requested middle school
students to assess not only how they expected to perform (self-efficacy) but also how
they expected their group as a whole to perform. While this study did not find it
appropriate to have each group member evaluate every member in the group,
primarily because of age, data will be collected on group-level expectation.
Identifiability is hypothesized to exert a significant influence on this relationship
based on the findings of Sanna (1992). As Bandura (1997) has suggested that self-
efficacy and collective efficacy are related, similar predictions can be made
regarding the influence of identifiability on the correlational relationship between
collective efficacy and effort. Therefore, under conditions of high identifiability,
there will be a significant and positive relationship between effort and collective
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
efficacy. However, under conditions of low identifiability, there will be a positive
but not significant relationship between effort and collective efficacy.
The remaining hypotheses examine relationships between pre-task variables
in this study. Therefore, the influence of the identifiability construct is not relevant
and distinguishing them on that basis would be inappropriate.
As stated earlier and mentioned above, self-efficacy has a profound influence
on effort and persistence. Therefore, similar predictions should be expected between
self-efficacy and effort and self-efficacy and persistence. Self-efficacy remains
constant in that it is a task-specific construct. However, effort in this study is
measured as a state variable at the conclusion of the task while persistence is
measured as a trait variable. Therefore, this particular hypothesis engages two trait
variables measured prior to the specific task and should not be subject to influence
from the identifiability construct. Bandura (1986,1991,1997) has suggested this is
one of the primary expectations held within self-efficacy theory. Therefore, nothing
unique should be discovered within the data of this study and it is hypothesized that a
significant and highly positive correlation between self-efficacy and persistence
should emerge.
Similarly, individuals that hold a positive view of the ability of their group to
cooperate and work together to achieve a similar goal are anticipated to persist. As
the task in this study is not detailed enough to measure task specific persistence, and
a measure of persistence (Lufi & Cohen, 1987) was adopted, similar results to the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
hypothesized relationship between self-efficacy and persistence are expected here.
Therefore, it is hypothesized that a significant and positive correlation between
collective efficacy and persistence should emerge.
Bandura (1997, 1999) has intimated that beliefs of personal efficacy are not
detached from the larger group entity. In fact, one of the components of measuring
collective efficacy is evaluating the coordinative capacity of the group, of which on
is a member and must include themselves in the rating. Further, Femandez-
Ballesteros, Diez-Nicolas, Caprara & Barbaranelli & Bandura (in press) and Bandura
(in press) suggest that although collective efficacy is multifaceted, this component is
in accord with the evidence pertaining to the structure of individual efficacy beliefs.
This suggests that although the constructs are empirically unique, a strong
correlational relationship between self-efficacy and collective efficacy is expected in
this study. While they are unique constructs and are expected to have distinguishable
influence on group performance and individual effort, a strong positive relationship
should emerge. Therefore, it is hypothesized that there will be a significant and
highly positive relationship between self-efficacy and collective efficacy.
In sum, the following hypotheses will be examined in the present study:
Hypothesis One - Under conditions of high identifiability, there will be a significant
and positive correlation between self-efficacy and effort.
Hypothesis Two - Under conditions of low identifiability, there will be a positive but
not significant correlation between self-efficacy and effort.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Hypothesis Three - Under conditions of high identifiability, there will be a
significant and positive relationship between effort and persistence.
Hypothesis Four - Under conditions of low identifiability, there will be a positive but
not significant relationship between effort and persistence.
Hypothesis Five - Under conditions of high identifiability, there will be a significant
and positive relationship between effort and collective efficacy.
Hypothesis Six - Under conditions of low identifiability, there will be a positive but
not significant relationship between effort and collective efficacy.
Hypothesis Seven- Across conditions, there will be a significant and positive
correlation between self-efficacy and persistence.
Hypothesis Eight- Across conditions, there will be a significant and positive
correlation between collective efficacy and persistence.
Hypothesis Nine- Across conditions, there will be a significant and highly positive
relationship between self-efficacy and collective efficacy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
Chapter 2
Methodology
Design of the study
Original design. The original design of this study was a nonexperimental,
multigroup structural equation model (see figure 1). The proposed design measured
the following independent variables: self-efficacy, group cooperation, and group
ability. These constructs explored causality and prediction between the latent
variables—individual effort and group effort—using a structural equation model
while incorporating persistence as a potential mediating variable. Finally, the
construct identifiability will be examined as a possible moderating variable that can
separate the subjects into two unique and distinct groups. Confirmatory factor
analysis will analyze the measurement model to test the hypothesis that relationships
exist between variables. This technique will determine the adequacy of the
hypothesized factor loadings and construct intercorrelations. Structural-equation
modeling will assess the causal relationships in the proposed model. Structural-
equation modeling was completed and can be seen in appendix C. Specific
information regarding the omission of structural-equation modeling as the primary
method of data analysis will be discussed in the subsequent chapters.
Final design. The final design incorporated a comparison of correlational
data, including partial correlations and simultaneous regression, to distinguish the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
effects and examine relationships among the variables. T-tests were incorporated to
analyze group differences.
Subjects
Eight sixth-grade classes were invited to participate in this study. Out of a
possible 250 subjects, more than 80% (n = 214) completed the study. The sample
consisted of (at least) 200 predominantly African-American and Hispanic, low-
socioeconomic-status students in grade 6 from the Foshay Learning Center, a
multicultural urban school serving students from kindergarten through grade 12. As
distinguishing effort loss between ethnic groups was beyond the scope of this study,
those data were not analyzed, but demographic data was collected. Foshay is located
within the Manual Arts Cluster of the Los Angeles Unified School District. More
than 90% of students at Foshay is eligible for the federal ffee-meal program and
qualify as low socioeconomic status. Subjects were assigned randomly to one of two
conditions (low and high identifiability) by taking sequential groups of four students
alphabetically from the roster sheets that identify individual classroom composition.
Instrumentation
The State Social Loafing Scale for Children (SSLSC) was created to assess
each of the variables in the study and adapted from various scales previously
included in related studies (see appendix A). This measure sought to define five
primary variables as follows: self-efficacy, collective efficacy, identifiability, effort,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
and persistence. This model will be analyzed using correlational data with groups
representing grade 6.
Independent Variables
Self-efficacy. Self-efficacy was assessed using a four-item confidence
measure generated by Esposito (1998). This instrument was developed in the spirit of
Bandura (1997). Self-efficacy received a reliability rating of .818 (Esposito, 1998).
The self-efficacy scale is found in Appendix A.
Persistence. Persistence items were incorporated from Lufi and Cohen (1987)
to identify trait-level individual persistence on the pretask measure. A 40-item
true/false scale reported internal reliability of .66 and .77 in a test-retest after 6
months. To maintain the consistency of the Likert format used in this study, the
items will be adapted to fit the Likert scale. In addition, utilizing 40 items would be
too labor intensive on a young population so an 11-item scale measured
persistence(see Appendix A).
Identifiability. Identifiability was assessed by including three items designed
to define the level to which individuals felt their contributions were: (a) able to be
identified by others and (b) whether their contribution was subject to evaluation by
others (see Appendix A). Prior research defined others as evaluators (Harkins &
Jackson, 1985) or the self (Szymanski & Harkins, 1987).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
Self-efficacy
Individual
effort
G roup
cooperation
persistence
— ► y G roup effort
G roup ability
Figure 1 : Hypothesized model to examine social-loafing effect. This is a multi-group
sample in which the construct identifiability will serve as the potential moderating
variable and distinguish groups based as belonging to either high or low levels of
perceived identifiability.
Collective efficacy. Collective efficacy was defined by including two
subscales: (a) belief in other members’ abilities, and (b) general belief in the group’s
ability to work cooperatively. This scale was adapted from Esposito (1998). The
scale has received preliminary reliability indications of .901 as a general scale, and
the subscales support Bandura’s (1997) definition of collective efficacy. Collective
efficacy and self-efficacy items can be found in Appendix A).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
Dependent Variables
Effort. Effort-based items were adapted from the Amount of Invested Mental
Effort scale (Salomon, 1983). This scale measures posttask invested mental effort.
Items were included to generate an individual’s perception of both individual and
group effort. Two items measured individual effort. One item was used on the
SSLSC to measure perceived group effort directly, while an average of all members’
perceptions of group effort was created to generate a second variable used to
measure group effort.
Method of Data Collection
Students completed the State Social Loafing Scale for Children (SSLSC)
model during one instructional period (see appendix A). Since the Foshay Learning
Center utilizes block scheduling of classes, each instructional period is
approximately 2 hours. Therefore, data collection and completion of the SSLSC can
be completed in less than one instructional period. The SSLSC consists of the four
scales mentioned above. Students indicated their responses on a 5-point Likert scale,
ranging from never, once in a while, sometimes, very often, and always. All items
were read aloud to both groups, and a standard script was used to enhance reliability
among each class. Students were assured of their anonymity, and participation was
strictly voluntary. Students who did not feel comfortable had every opportunity to
end their involvement with the project. No students chose to stop participating.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
After participation in the performance task, students were asked to complete
the posttask portion of the State Social-Loafing model. This was completed during
the same visit. Supervising classroom teachers moderated all classrooms, and no
classes required extra time. There were no students who required specific learning
environment arrangements, such as required in special-education law.
Procedure
All experimental sessions were conducted at the school setting by the
researcher. This study required one classroom visit that took less than 2 hours. The
first half of the visit had participants complete the first portion of the State Social-
Loafing Model designed to measure collective efficacy, self-efficacy, and
persistence. Subjects completed two ability measures (see appendix B). The second
half of the visit had subjects complete group tasks and collect data regarding
achievement. This final part of the visit had students complete the State Social-
Loafing Model, measuring reported levels of effort invested on the task and
perceived levels of identifiability. This served as an opportunity to debrief
participants. No students were suspicious or had questions regarding any part of the
study.
Setting. The experimental setting consisted of 10 medium tables, each with 4
chairs, located throughout the room. The experimenter quietly monitored the
progress of groups but did not sit down or answer questions once the activity began.
Classes participated collectively, with all groups sitting in the same room. This was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
an efficient way to compile data on an entire class without spending unnecessary
effort on organizational issues or taking students away from instructional time and
interfering with the instructional program.
The experimenter greeted subjects as they entered the task-performance
room. Each student signed the roster of participants, moved to the assigned task
table, and sat at the chair in front of a packet of work with their name written on an
index card on the top. They were instructed to either keep the card or dispose of it at
the end of the activity. At the table were four stacks of index cards, sharpened
pencils, and one box with a slit on the top large enough to fit a folded index card.
Index cards in the highly identifiable condition were color coded to determine
individual contributions. Each student in a group received a stack of either pink,
yellow, blue, or purple index cards. Index cards in the low identifiability condition
were all white. However, slight discriminating dots were placed on them to ensure
identifiability to the researcher but not the participant. All students had 30 index
cards to avoid having to borrow from others and, thus, interfere with the accuracy of
data collection. No student used all index cards, and no student verbalized an
acknowledgement regarding the coding of the index card. This manipulation,
coupled with the success of the identifiability manipulation mentioned earlier,
indicated that the task was completed without complication.
The experimenter introduced the brainstorming task. Participants were
handed a direction sheet and instructed to read along. Subjects were asked to raise
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
their hands if they had any questions. At that point, the experimenter read the
following instructions to every group:
“Today you will take part in what is called a brainstorming task. You will be
given the name of an object, as well as provided with a sample, and your task will be
to come up with as many uses for this object as you can. Don’t be concerned about
the quality of your answers as the responses can be ordinary or very unusual.
However, responses will be evaluated so there is no repetition of a verb. For
example, a spoon can be used to eat but will not be counted as a unique idea if you
include; eat ice cream, eat cake, eat cereal, etc.... It is important for you to know that
there is no grade on this task. Do not be concerned over writing incomplete sentences
or spelling every word correctly. Simply list as many uses on the provided paper as
possible. Please write one use per index card.”
Subjects in the high identifiable condition were also read the following:
“I am going to ask you to write only one use on each index card, fold the
card, and then put that slip of paper into the box in the middle of the table. You will
notice each of you have different colored cards. Please do not mix up the index
cards. This is our way of making sure everyone feels their input is important and that
each of you bears the full responsibility for generating as many uses as you can. In
addition, next to your identification number on the top of the Social-Loafing scale, I
would like you to write the color of your index card. You will have 9 minutes to list
as many uses as possible. However, responses will be evaluated so there is no
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
repetition of verb. For example, a spoon can be used to eat but will not be counted as
a unique idea if you include; eat ice cream, eat cake, eat cereal, etc.... In other
words, if you repeat the action this object will be used for but use different things the
object can be used with, only one credit will be given.”
Subjects in the low identifiability condition were read the following:
“I am going to ask you to write only one use on each index card, fold the
card, and then put that slip of paper into the box in the middle of the table. It is not
important if you come up with the same uses as your partners. I want to see how
many uses your group can come up with. Each of you shares the responsibility with
your partners for coming up with as many uses as possible. You will have 9 minutes
to list as many uses as possible. However, responses will be evaluated so there is no
repetition of verb. For example, a spoon can be used to eat but will not be counted as
a unique idea if you include; eat ice cream, eat cake, eat cereal, etc.... In other
words, if you repeat the action this object will be used for but use different things the
object can be used with, only one credit will be given.”
Both groups were given the following motivator: “It has been found that
effort is directly linked to intelligence. Research seems to indicate that the harder
you work, the smarter you are. I understand you are working with a time limit, so do
not worry if you misspell a word.”
At that point, the experimenter called for questions, reminded the students
they had 9 minutes, and told the groups to start working. The experimenter did not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
communicate with the groups until their 9 minutes elapsed. Answers were not read
for quality. Any answer that was placed in the box was acceptable according to the
guidelines stated. There were no instances of repeated verbs.
At the conclusion of the performance task, participants were thanked and
dismissed back to class.
Differences Between Previous Studies and This Replication
There are some important differences in outcomes between Harkins’ and
Petty’s (1982) experiment 1 and this replication. In addition, this experiment was
significantly different from other conceptual replications. The intention of this
experiment was to combine the positive attributes of multiple studies and create one
study that would examine previously discovered moderating variables and subject
them to testing outside the laboratory. The major differences were found between the
conditions, task presentation, as well as several general differences.
Differences in the conditions. In Harkins and Petty (1982), data was collected
with either one subject or groups of five. Loafing research is primarily compiled by
examining behavioral outcomes of collective and coactive conditions. Social
facilitation research, on the other hand, compares alone and collective conditions.
While this poses an interesting combination of the fields later explored by Harkins
(1987), it does not qualify as a true loafing paradigm. In addition, social-impact
theory predicts different effects based on increasing group size. By controlling for
this effect, the experiment eliminates the influence of group size and easier
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
comparison between groups is facilitated. A true loafing paradigm has been utilized
by various researchers to compare coactive and collective conditions (Sanna, 1992;
Williams & Karau, 1991), and this study seeks to replicate these studies partially.
Recently, loafing studies have been run in groups of 2 participants. This use of dyads
is an interesting manipulation o f the definition of group and has been used frequently
(Karau & Hart, in press; Karau & Williams, 1997). Karau argued the use of dyads
was created for convenience because of the increasing difficulty with scheduling
larger groups (personal communication, April 22,1998). The findings may be
suspect as the dyad pushes the concept of group to the limit regardless of definition.
It is assumed groups larger than 2 are preferable. The dyad poses obvious conceptual
problems. The distinction between coactive and collective activities may have a
tendency to become confounded. While preferred group size relevant to social-
loafing research may be helpful, the preferred method for assessing the loafing
phenomenon is achieved by comparing coactive and collective groups larger than 2
participants, and this dissertation employs that method of comparison.
Differences in task presentation format. In Harkins’ and Petty’s (1982)
experiment 1, subjects in the group condition were isolated from one another during
the task. Regardless of condition, belief in the ability of the group to cooperate to
accomplish a task successfully is a necessary condition. By separating participants,
the experiment no longer remains a group task and becomes a summative measure of
individual contribution without regard for cooperation. It is plausible that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
membership in a group with minimal interaction makes it difficult to maintain
expectancies of others. This is not to say there are no situations in which groups
work apart from each other. Many groups, such as assembly line workers, rely on the
performance of others, thereby relying on trust. That is why the Rotter (1971)
Interpersonal Trust scale was utilized. As this study examines individual effort
reduction within group settings, subjects work within cooperative settings, regardless
of condition. No participants were separated in this experiment.
Szymanski’s and Harkins’s (1987) experiment 1 presented the task format in
a manner similar to this study. Subjects sat around a table for a certain amount of
time and composed uses for a knife. The difference between the two experiments
was that identifiability was determined by providing subjects with 70 slips of paper
and counting the remaining slips to determine individual input. Subtracting the
remaining slips from 70 determined individual achievement, and this methodology is
problematic. The assumption of this study is that any paper folded and put in the box
must constitute a use. The major difference between Szymanski and Harkins and this
study is that all subjects were run individually. This method was inappropriate for
this study as the independent variable was self-evaluation. This study is not
concerned with the origination of evaluation. It is assumed the experimenter provides
the potential for evaluation.
General differences. There are several distinctions that distinguish this study
from previous studies on the same phenomenon. It may be construed that this study
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
is a partial replication, as it does not reproduce exactly any one experiment, yet
borrows facets of many. The primary distinction between this dissertation and
previous research is the setting. According to Karau and Williams (1993), 166
studies have been conducted on social loafing. Of these studies, 24 have occurred
away from the laboratory. One claim repeatedly offered by social-loafing researchers
emphasizes the need for real world studies. By providing this data, comparative
analyses can be generated to determine whether the moderating variables tested here
apply outside the laboratory.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
Chapter 3
Results
Descriptive Statistics
Table 1 shows that the measures for the various motivational variables have
low to good reliabilities (from a low of .44 to a high of .78). The low reliabilities
were possibly a result of including too few items to measure the variable sufficiently;
therefore, the low reliabilities were not surprising. For example, individual effort was
originally measured with only two items. In addition, group effort was measured
with one item, and low correlations resulted in the variable being consolidated to
create one general-effort construct. This remains in line with research as the unit of
analysis in this study is the individual, and all measures of both group and individual
effort are taken from the perspective of the individual. In addition, reliability of the
persistence variable in this study (.44) was dramatically lower than the original
instrument (.66) by Lufi and Cohen (1987). This difference in reliability can be
explained possibly by the age difference between subjects. Although the items were
modified so they would be age appropriate, there is a probability these changes may
have influenced the reliability of the variable. In addition, descriptive data is
provided for the items that measure each observed variable. Tables 2 and 3 provide
descriptive data for both of the conditions; table 2 describes subjects in the low-
identifiability condition while table 3 describes those in the high-identifiability
condition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
Within collective efficacy, descriptive statistics are provided specifically for
group appraisals of cooperation and to keep the spirit of Bandura’s (1997) research.
Specific to reports of group-level effort, each group member’s appraisals of the
collective effort was aggregated to create an item used in conjunction with item 27
on the State Social Loafing Scale for Children (see appendix A).
Measured variables are skewed and have kurtosis indices that indicate large
deviations from normality. It may be concluded that the overwhelming tendency
toward negatively skewed distributions found in Table 1 was a direct result of the
age of the participants, the familiarity and desire to please the experimenter, or lack
of involvement with studies of this kind. Subjects were all between the ages of 10
and 13 years of age. While Marsh (1989) reported that young subjects are able to
self-report feelings of self-worth accurately, the plausibility of this conclusion is
questionable in the current sample.
For a summary of the descriptive data relative to subjects in the low-
identifiability condition, refer to table 2. For a summary of the descriptive data
relative to subjects in the high-identifiability condition, refer to table 3.
Frequencies
Data was not collected on ethnic differences in the present study as it is
beyond the scope of the study. However, it is relevant to note that as this study
intended to focus on high-risk urban students who are representative of other urban
schools, a report issued by the Los Angeles Unified School District (CBEDS, 2000)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
indicated Foshay Learning Center was composed of 70% Hispanic and 29% African-
American students. This suggests the sample is representative of urban schools
primarily in Los Angeles.
Table 1
Descriptive Statistics of Variables Among Whole Subjects (« = 214)
Variable Mean sd Skew Kurtosis Alpha
SE 16.92 2.68 -1.00 .70 .65
Item 1 4.46 .82 -1.6 2.00
Item 5 4.27 .91 -1.2 .86
Item 9 4.07 .96 -1.0 .77
Item 13 4.13 1.13 -1.2 .70
CE 28.86 4.7 -.803 -.003 .78
Item 2 4.12 .92 -1.0 .75
Item 4 4.0 .99 .79 -.03
Item 6 4.4 .77 -1.5 3.0
Item 8 4.0 1.15 -1.0 .15
Item 10 4.5 .79 -1.5 -1.8
Item 12 3.8 1.16 -.79 -.261
Item 14 4.0 1.21 -1.13 .296
Cooperation 13.03 1.86 -1.2 1.4 .62
Ability 1.6. 3.37 -.668 .422 .74
Persistence 38.87 5.5 -.335 -.007 .44
Effort 16.75 2.31 -1.183 1.837 .60
Note. SE = self-efficacy; CE= collective efficacy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
Table 2
Descriptive Statistics of Variables Among Low Identifiability Subjects (n — 108)
Variable Mean sd Skew Kurtosis Alpha
SE .724
Item 1 4.57 .76 -2.2 5.4
Item 5 4.29 .87 -1.04 .25
Item 9 4.18 .99 -1.2 .88
Item 13 4.08 1.22 -1.28 .622
CE .809
Item 2 4.37 .84 -1.274 1.43
Item 4 4.07 .96 -.916 .513
Item 6 4.42 .77 -1.5 3.0
Item 8 4.00 1.14 -.98 .087
Item 10 4.51 .77 -1.68 2.50
Item 12 3.84 1.15 -.826 -.097
Item 14 4.04 1.19 -1.19 .575
Cooperation 11.9 2.5 -.596 -.456 .645
Ability 13.2 1.8 -1.36 1.953 .742
Persistence 46.08 5.0 -.352 .217 .311
Effort 16.75 2.44 -1.275 2.121 .60
Note. SE = self-efficacy; CE= collective efficacy.
This study did collect data on gender, but Table 4 identifies that gender was
evenly distributed in the whole sample as well as within each manipulated condition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
Table 3
Descriptive Statistics of Variables Among High Identifiability Subjects (n = 106)
Variable Mean sd Skew Kurtosis Alpha
SE .556
Item 1 4.35 .87 -1.12 .24
Item 5 4.25 .95 -1.25 1.252
Item 9 3.95 .92 -.954 .996
Item 13 4.17 1.05 -1.16 .540
CE .771
Item 2 4.0 .95 -.821 .478
Item 4 3.97 1.02 -.672 -.432
Item 6 4.36 .77 -1.49 3.26
Item 8 4.04 1.17 -1.96 .276
Item 10 4.42's .79 -1.36 1.445
■ Item 12 3.82 1.19 -.763 -.370
Item 14 4.05 1.29 -1.094 .107
Cooperation 15.85 3.42 -.719 -.389 .728
Ability 12.77 1.88 -1.097 1.226 .585
Persistence 37.69 5.70 -.222 -.153 .523
Effort 16.75 2.19 -1.06 1.44 .542
Note. SE = self-efficacy; CE= collective efficacy.
Table 5 represents the pattern of answers in response to self-reported levels
of identifiability based on condition. Initial examination of the data indicates that a
significant percentage of subjects were in the appropriate group, with those subjects
in the low-identifiability group indicating it would be difficult to detect individual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
contribution and those in the high-identifiability group indicating their contributions
would not be difficult to detect.
Table 4
Frequencies of Gender in Whole Group and Both Conditions
____________ Gender____________________ n______________ %
Whole Group Male 110 51.4
Female 104 48.6
Low Identifiability Male 56 51.9
Female 52 48.1
High Identifiability Male 54 50.9
Female 52 49.1
Correlations
Table 6 shows the bivariate zero-order correlations for all the motivational
and achievement variables, with associated p-values indicating two-tailed
significance for the whole sample. It can be noted that while the majority of
correlations are significant, there is one general area where significance is lacking.
Gender tends to have no distinct relationship with any pretask trait measures of self-
efficacy, anticipated group efficacy, or persistence. Therefore, it may be suggested
that gender is not as important in evaluating motivation or group membership as
other variables such as self- and collective efficacy.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
Table 5
Frequency Table for Identifiability Manipulation for Both Groups
Low Identifiability—item 1
F % Cum%
Very hard 70 64.8 64.8
Sort of hard 14 13 77.8
Not very hard 12 11.1 88.9
Sort of easy 7 6.5 95.4
Very easy 5 4.6 100
Low Identifiability—item 2
F % Cum %
Very hard 50 46.3 46.3
Sort of hard 20 18.5 64.8
Not very hard 12 11.1 75.9
Sort of easy 16 14.8 90.7
Very easy 10 9.3 100
Low Identifiability—item 3
F % Cum %
Very hard 70 64.8 64.8
Sort of hard 13 12 76.9
Not very hard 9 8.3 85.2
Sort of easy 5 4.6 89.8
Very easy 11 10.2 100
High Identifiability—item 1
F % Cum %
Very hard 6 5.7 5.7
Sort of hard 6 5.7 11.3
Not very hard 10 9.4 20.8
Sort of easy 23 21.7 42.5
Very easy 61 57.5 100
High Identifiability—item 2
F % Cum %
Very hard 9 8.5 8.5
Sort of hard 8 7.5 16
Not very hard 21 19.8 35.8
Sort of easy 15 14.2 50
Veiy easy 53 50 100
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
Table 5 (cont.)
High Identifiability—item 3
F % Cum %
Very hard 7 6.6 6.6
Sort of hard 3 2.8 9.4
Not very hard 12 11.3 20.8
Sort of easy 33 31.1 51.9
Very easy 21 48.1 100
Table 6
Correlations Among Motivational Variables in Whole Group (n = 214)
Gender SE CE P Effort
Gender 1.00
SE .123 1.00
CE .067 .465** 1.00
P .047 .452** .409** 1.00
Effort .132 .214** .326** .240** 1.00
Note. SE = self-efficacy; CE = collective efficacy; P = persistence.
*2 < .05, ** g < .01.
Table 7 provides the correlational matrix for subjects in the low-
identifiability condition. Minor differences emerge between the entire sample and the
low-identifiability condition. This may indicate that when given the opportunity to
reduce effort, students with higher efficacy expectations reduce their effort and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
become the social loafer in the group situation. Further, students with higher
individual efficacy rate the group’s ability to cooperate and complete the task as
sufficient. While this is not proof in and of itself, the possibility that students with
high efficacy reduce their effort when they feel the group is capable of completing
the task itself is noteworthy.
Table 7
Correlations Among Motivational Variables in Low Identifiability (n = 108)
Gender
Gender
1.00
SE CE P Effort
SE .159 1.00
CE -.073 .437** 1.00
P .001 .460** .373** 1.00
Effort .155 .217* .541** .264** 1.00
Note. SE = self-efficacy; CE = collective efficacy; P = persistence.
*p< .05, **p< .01.
Table 8 presents the correlational matrix for subjects in the high-
identifiability condition. The relationship between tables 7 and 8 examines collective
efficacy and effort and is noteworthy. Subjects in the high-identifiability condition
appear to exert little if any effort in the task. From an initial evaluation, this
contradicts what would be expected. High levels of collective efficacy indicate a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
belief that the group not only possess the prerequisite skills to accomplish the task,
but also indicates the group members believe they can work together to cooperate.
Further, subjects believe that each member’s contribution will be easily identifiable.
In the high-identifiability condition, in which subjects claimed they felt their
contributions were identifiable, subjects with high self-efficacy invested a significant
amount of self-reported effort. This, along with table 7, may indicate that self-
efficacy and persistence are the primary predictors of invested mental effort.
Table 8
Correlations Among Motivational Variables in High Identifiability (n = 106)
Gender SE CE P Effort
Gender 1.00
SE .082 1.00
CE -.092 .369** 1.00
P .095 ' .439** .277** 1.00
Effort .100 .211* .078 .224* 1.00
Note. SE = self-efficacy; CE = collective efficacy; P = persistence; IE = Individual
Effort; GE = group effort.
* p < .05, ** p < .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
Check of Identifiability Manipulation
The sample was manipulated to differentiate two distinct groups based on
identifiability (low identifiability and high identifiability). Half of the participants
were given similar colored index cards were intended to perceive their contributions
to the group product as difficult to identify. The remaining half of the participants
received a unique set of colored index cards so that they would perceive their
contributions to the group product would be easily identified.
Table 9 reports the t-test to verify whether the identifiability manipulation
was successful. Group 2 (M = 4.07) indicated that they felt it would difficult to
detect individual contributions in the group task, and this was significantly different
than group 1 (M = 1.93), t (212) = -16.98, p < .001.
Table 9
T-test to Verify the Manipulation of Identifiability ___________ __________
Identifiability N Mean SD t df sig(2-tailed)
Low Identifiability 108 1.93 .94 -16.98 212 .001
Hi Identifiability 106 4.07 .90
This t-test indicates that a separation of the subjects into two distinct groups
was successful. Subjects, who participated in the study, using similar index cards to
prompt difficulty in individual identification, reported overwhelmingly that it was
difficult for the researcher to identify their contributions. Along with the t-test, an
examination of group means further demonstrates significant differences, as the self-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
report instrument utilized a 5-point Likert scale and there is more than a 2-point
difference between groups.
Hypothesis 1
Under the high-identifiability condition, the correlation between self-efficacy
and effort will be significant.
Table 10 indicates no significant differences between both low- and high-
identifiability conditions. However, there is a slight statistical difference when
compared to the whole group sample. This appears to be an insignificant finding
because of the need to explore the differences that the separate conditions create.
Table 10
Correlational Differences Between Groups Based on Self-Efficacy Effort
Self-Efficacy
Effort (WG) Effort (LI) Effort (HI)
Pearson .214** .217* .211*
Significance .002 .024 .031
N 212 107 105
Note. WG = Whole Group; LI = Low Identifiability; HI = High Identifiability.
* j) < .05, ** g < .01
While no significant difference between conditions based on level of
identifiability emerge, the hypothesis appears to receive support in that the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
relationship between the variable is significant. Although this difference is not
statistically significant, as all conditions were significantly related to self-efficacy,
the influence of identifiability needs further consideration. Further, social loafing
research suggests identifiability mediates relationships between variables such as
self-efficacy and effort. Future research needs to clarify this finding.
A t-test determined whether any significant group differences emerged when
controlling for identifiability. Table 11 shows the results of the t-test and
demonstrates the only significant variable is persistence, which is also the only trait
variable in this study.
Table 11
T-Test for Group Differences Based on Identifiability
F Sig. T df sig (2-tailed)
SE 1.607 .206 1.133 210 .259
PER 2.134 .146** 3.820 205 .000
CE .007 .931 .919 206 .359
Effort .155 .694 .016 212 .987
Note. **g < .01.
By controlling for persistence and utilizing a partial correlation, the
relationship between self-efficacy and effort can be evaluated more thoroughly.
Table 12 demonstrates that when controlling for persistence, no relationship exists
between self-efficacy and effort. In fact, a stronger relationship emerges within the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
low-identifiability condition. This indicates that when individuals have the
opportunity to reduce their efforts, highly efficacious students emerge as the loafers,
reducing effort. This contradicts previous findings and is analyzed with caution
based on nonsignificant data presented in table 11.
Table 12
Partial Correlation Demonstrating Relationship Between Efficacy and Effort
(Controlling for Persistence)______________ ■ _______________________ _______
Effort (low ident.) Effort (high ident.)
SE .1153 (p - .258) .0810 (p = .426)
Note. SE represents self efficacy.
It appears this hypothesis did not receive support but requires further
examination of the influence of self-efficacy and effort within the two conditions.
That individuals with higher efficacy appraisals may be the group members that
disengage is an interesting finding and worthy of more stringent examination.
Hypothesis 2
Under conditions of low identifiability, there will be a positive but not
significant correlation between self-efficacy and effort.
Table 10 provides data relevant to this hypothesized relationship that under
conditions of low identifiability, students with lower efficacy rating would reduce
effort. Hypothesis 2 is not supported as a significant relationship emerged across all
conditions, specifically within low identifiability. Effort reduction within this
hypothesis primarily tests the influence of the loafing phenomenon specific to self
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 6
reported levels of efficacy and invested mental effort. Table 10 illustrates a positive
and significant correlation that partially supports the hypothesis. However, the
loafing effect did not emerge when examining the data through Pearson correlation.
This suggests the relationship between self-efficacy and effort maintains a strength
regardless of whether or not individuals have an opportunity to hide. When
controlling for persistence, the hypothesis receives limited support. Table 11
suggests that group differences did emerge and the partial correlation presented in
Table 12 suggests that persistence may, in fact, mediate this relationship and further
statistical evaluation is deemed appropriate.
An underlying concern within the primary data analysis focuses on the
comparative strength of the relationship based on identifiability. Social loafing
theory anticipates effort reduction within the low identifiability condition. However,
both Tables 10 and 12 suggest that participants in the low identifiability condition
invested more effort than those in the high identifiability condition. This comparison
further suggests a need more a more rigorous evaluation of the relationship between
self-efficacy and effort with younger populations.
Hypothesis 3
Under conditions of high identifiability, there will be a significant and
positive relationship between effort and persistence.
Data provided in table 13 indicate support for this hypothesis. The
relationship between state level effort and trait level persistence is significant and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
positive. As discussed earlier, a significant relationship between effort and
persistence was anticipated. In situations where individuals perceived their
opportunity to hide as limited, invested mental effort should logically increase.
Comparing individuals in the high identifiability condition with the whole group
sample, it appears that the behavior of the participants was slightly counterintuitive.
This suggests a need to examine the relationship between effort and persistence
under conditions of low identifiability.
Table 13
Correlational Data Relating to Effort and Persistence_____________________
Pers
(whole group)
Effort .240**
Sig .000
N 207
* g < .05,"** p< 01
Hypothesis 4
Under conditions of low identifiability, there will be a positive but not
significant relationship between effort and persistence.
Table 13 suggests that there is virtually no effect to differentiate the
relationship between effort and persistence when comparing individuals in the low
identifiability condition with the whole group. Further, compared to individuals
Pers Pers
(low ident.) (high ident.)
.264** .224*
.007 .022
105 102
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
within the high identifiability condition, the difference is negligible. All conditions
suggest a significant and positive relationship between the variables that initially
does not offer support for the hypothesis. While the data supports the historical
relationship between these variables, little insight is generated when evaluating its
effect on the loafing phenomenon. Data suggests the level of identifiability has no
differentiating effect on the relationship.
Hypothesis 5
Under conditions of high identifiability, there will be a significant and
positive relationship between effort and collective efficacy.
Table 14 shows both Pearson and partial correlations (controlling for
persistence) for collective efficacy and effort. This data suggests a profound rejection
of the hypothesis. This data soundly rejects the hypothesis that when individuals
believe their contributions to the group productivity will be readily identifiable,
higher ratings of collective efficacy result in increased mental effort. This finding is
further supported when controlling for persistence. Taken together, this suggests that
when individuals with high expectations for the group believe their contributions to
be easily identifiable and evaluable, reduce their effort. These findings suggest a
dramatic divergence from historical expectation as well as the study at hand. While
data analysis remains correlational and does not allow for discussion of causality,
findings are important and novel. The exploratory findings suggested in Table 14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
soundly reject this hypothesis and suggest further examination into the relationship
between these variables relevant and appropriate.
Table 14
Correlational Data for Collective Efficacy and Effort- high identifiability
Effort
WG HI HIPC
Collective Efficacy
Pearson .326** .078 -.069
Sig (2-tailed) .000 .438 .492
N 208 102 102
Note. WG= whole group; HI = high identifiability; HIPC = high identifiability with
partial correlation.
* g < .05, ** j) < .01.
Hypothesis 6
Under conditions of low identifiability, there will be a positive but not
significant relationship between e ffo rt and collective efficacy.
Table 15 shows the relationship between collective efficacy and effort under
conditions of low identifiability. The data supports prior hypotheses regarding the
relationship between the two variables. Results suggest decreased effort occurs when
it is readily apparent who is contributing to the group product. Further, this condition
creates an opportunity for participants to hide and therefore reduce invested mental
effort. Table 15 suggests that individuals not only failed to reduce effort, but did so
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
in a significant, positive manner. This positive effect was not surprising as group
work is thought to be motivating. However, compared to Table 14, this data suggests
the exact opposite of the hypothesized relationships occurred. There is a lack of
research examining the relationship between collective efficacy and effort, which
makes this data exploratory. Taken together, the data within this study and the
paucity of research in this field suggests that further statistical examination may
enlighten these unexpected findings.
Table 15
Correlational Data for Collective Efficacy and Effort- low identifiability______
Effort
WG LI LIPC
Collective Efficacy
Pearson .326** .541** .513**
Sig (2-tailed) .000 .000 .000
N 208 106 106
Note. WG= whole group; LI = low identifiability; LIPC = low identifiability with
partial correlation.
* £ < .05, ** £ < .01.
Hypothesis 7
Across conditions, there will be a significant and positive correlation between
self-efficacy and persistence.
Data supports the hypothesis as the relationship between self-efficacy and
persistence is significant and highly positive. Both variables were trait level, pre-task
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
variables in this study and supported previous research in the field. Compared to
Table 10, the data suggests that traditionally anticipated relationships identified
under social cognitive theory gained support. Individuals that reported high levels of
efficacy claim to persist on academic, social, and personal tasks. As both variables
are pre-task variables, similarity is found across conditions as identifiability has no
influence.
Table 16
Correlational Data Relating Self-Efficacy and Persistence___________________
Pers
(whole group)
SE .479**
sig .000
n 214
* g < .05, ** g < .01.
Hypothesis 8
Across conditions, there will be a significant and positive correlation between
collective efficacy and persistence.
Table 17 presents data for collective efficacy and persistence. Support was
found for this hypothesis as suggested in social cognitive theory and logical
connections between hypotheses 8 and 9 are suggested. Previous research suggests
that self and collective efficacy are similar in sources of influence and variables they
influence, specifically effort and persistence. No surprising findings surfaced within
Pers Pers
(low ident.) (high ident.)
.501** .460**
.000 .000
108 106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
this hypothesis as individuals that believe their group has the ability to complete the
task and cooperate persist on academic, social and personal tasks. Table 17 suggests
no novel information and confirms previous research findings.
Table 17
Correlational Data Relating Collective-Efficacy and Persistence
Pers Pers Pers
(whole group) (low ident.) (high ident.)
CE .409** .373** .277**
sig .000 .000 .000
n 214 108 106
* p < .05, ** p < .01. “
Hypothesis 9
Across conditions, there will be a significant and positive relationship
between self-efficacy and collective efficacy.
Table 18 suggests data providing support for the hypothesis under question.
As suggested in earlier tables, theoretical expectations for the influence of self-
efficacy and collective efficacy received mixed support. However, the same
theoretical basis suggests that similar influences emerge from both individual and
collective efficacy measurements. Table 18 suggests the exploration of this
relationship is both significant and highly positive.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
Other Statistical Considerations
In line with the exploratory nature of this study, it was necessary to consider
whether the variables collectively made a stronger contribution on effort beyond the
correlational data. A simultaneous regression was conducted to answer this question,
Table 18
Correlational Data Relating Self- Efficacy and Collective-Efficacy___________
CE CE CE
(whole group) (low ident.) (high ident.)
SE .465** .437** .369**
Sig .000 .000 .000
n 214 108 106
* g < .05, ** g < .01. “ — —
and the results are presented in table 19. As the table demonstrates, the only variable
that emerges as significant in the regression is collective efficacy. In
addition,regression analysis accounted for only 11% of the total variance.
Unfortunately, this does not provide additional insight beyond that found in the
correlational analysis. Further, the analysis provides no additional insight regarding
the relationship between self-efficacy and effort or persistence and effort. However,
the regression analysis demonstrates a critical need to further examine this
relationship in “real world” settings.
In addition to the correlational data and the regression data, a structural
equation model was created to investigate causal relationships among the variables.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
The final model can be found in appendix C. While it is not relevant to discuss that
the findings ofthe structural equation model were not significant, it is worth
mentioning. The structural model did not work because of several critical reasons
Table 19
Simultaneous Regression With Effort as Dependent Variable
V a r ia b le b m i s ig !
Self-efficacy 1.987E-03 .070 .002** .977
Persistence 4.078E-02 .028 .117 .147
Collective efficacy .247 .080 .264 .002
Dependent variable: effort
* g < .05, ** p < .01.
including, but not limited to, small sample size, non-normal data, and poor
measurement of variables and subsequent factors (see table 1). Collectively, these
contribute to the power of the data and subsequent analysis. However, it is worth
noting that future social-loafing research may benefit from in-depth examination of
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
75
the phenomenon through the stringent requirements of structural-equation modeling
to determine causality among variables.
Results seem to indicate that when groups have the opportunity to hide, they
invest mental effort and reduce effort when they perceive their contributions to the
group product can be identified. Divergent findings indicate a need for further
analysis to clarify the relationships. Results indicate an emergent area of study that
will be discussed subsequently.
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
76
Chapter 4
Discussion
The primary purpose of this study was to subject previous findings in the
social-loafing realm to real-world experimentation. By subjecting previous findings
to analysis outside the laboratory and applying the research to school age children,
this research became exploratory. There is a noticeable paucity of research in the
social-loafing area that focuses on children, and this study begins to fill that void and
illustrates direction for the future. In addition, the introduction of collective efficacy
was introduced to clarify further the role of the individual within group-level
motivational situations. This has direct relevance to cooperative learning and other
collaboratively based activities in classroom situations.
It was anticipated that students who had the opportunity to hide when their
efforts go unnoticed would take the opportunity to do so. This was the primary
contribution of prior studies on social loafing. In this study, students did not reduce
effort when expected and, in fact, exhibited the opposite behavior. Students who
reported they believed their group had the prerequisite skills and ability to cooperate
reduced their effort to the group task when their contributions were identifiable.
Virtually every previous study in the social-loafing literature contends that making
individual contributions identifiable eliminates the loafing effect. Further, there was
no significant relationship between self-efficacy and effort. This is surprising as self-
efficacy historically has been a primary source of increased effort.
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
77
The data set obtained in this study had high levels of non-normal data in
addition to a sample size that was smaller than desired. This all contributed to an
inability to rely sufficiently on the analysis afforded by a structural-equation model
and called for analysis using more simple statistical analysis. Although this will be
explained later, the analysis of this data did generate interesting findings that point to
a significant need to continue investigating individual effort within group projects,
especially in social-loafing research. Considering the population in this study, the
importance of continuing research in the area of young children and effort reduction
within group tasks emerged. With the introduction of collective efficacy, this study is
an exploratory analysis of social loafing in children. The novelties of real-world data,
coupled with the findings, are worthy of discussion.
Discussion of this study will occur in four sections: difficulties with the
study, findings, areas for improvement, and, finally, future implications and
directions for research will conclude the paper.
Difficulties With This Study
Immediate analysis of data revealed multiple levels of non-normal data.
Further examination indicated the majority of self-reported items were negatively
skewed. This study did not collect qualitative data, but historical evaluation of
success in urban schools indicates a poor rate of success when evaluating success as
a measure of academic performance. The inflated scores in this study indicate a
desire to please the researcher. The high level of familiarity between subjects and the
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
78
researcher may have resulted in variables that were not truly representative of the
sample. This concern may have had a profound influence on the subsequent analysis
using multivariate statistics. Further, while this study primarily incorporated less /
sophisticated statistical analysis, the correlational data presented herein does
sufficiently address the primary objectives of the study; namely to subject social
loafing research to real-world younger populations.
The primary difficulty with this study was problematic nature of using
structural-equation modeling as the primary method of analysis. What follows is a
brief explanation regarding the specific concerns with the collection of data.
Problems with SEM. All results within this study should be viewed with
caution based on the high levels of non-normal data. This does not dismiss this
study’s foundational conclusions but results in difficulty analyzing data with
advanced statistical programs. Findings reported in the previous chapter were
subjected to extensive statistical analysis. As demonstrated in appendix C, analysis
by a structural-equation model is deemed unreliable. Primarily, difficulties emerged
from several design flaws. First, structural-equation models are sensitive to sample
size and require sample sizes of at least 200 per group to be evaluated accurately. As
this study was conducted at the end of a school year in a highly sensitive area, access
to subjects was limited, and this jeopardized the final analysis using structural-
equation modeling. Participation was limited because of the proximity of the state-
mandated achievement examinations (Stanford Nine). Small sample size was the
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
79
primary obstacle impeding the reliability of the structural-equation model as the
primary method of analysis. In addition, goodness-of-fit measures were extremely
low and prevented an accurate assessment of the influence of specific variables.
Major modifications to the model would have sacrificed the accuracy of the data
analysis and were rejected.
However, using alternative measures of fit, a model was generated (see
appendix C). While the model appears to work, the analysis indicates that no
significant findings were available that clarified the relationships under investigation.
Further, there was no additional insight regarding the relationships among major
variables that would have aided the understanding of social loafing in a causal
manner. High levels of error made evaluation with structural-equation modeling
difficult at the best (see appendix C), and small sample size increases the possibility
of error. Interpreting data in that manner is unrealistic. It was deemed appropriate to
abandon this method and proceed with a more simple, and accurate, method of
analysis.
Removed variables. The initial study included ability measures that can be
found in appendix B. These measures were age appropriate and relatively straight
forward. Unfortunately, students misunderstood the expectations of the task and
guessed when it was not appropriate. In these measures, students received points for
correct answers but had a fraction of a point deducted for incorrect answers.
Subsequently, a majority of students received negative values for scores, and the
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
80
scores were deemed unreliable and removed. This directly contradicts advice
traditionally given by teachers during the Stanford Nine Achievement Examinations,
which are being given within 1 month of this study. Using grade point averages as
indicators of ability was problematic as access was limited. In addition, as there is no
set rubric for evaluating student work and distributing grades, reliance on this .
measure as an accurate measure of ability is not acceptable. As the inclusion of an
ability measure was primarily introduced to support the achievement (number of
ideas generated) variable, its removal was not detrimental to the study considering
the achievement variable had been removed.
Removing the ability measure became an opportunity to clarify the loafing
phenomenon further. Historically, social loafing has been evaluated in terms of
achievement. Replicating previous social-loafing studies encouraged the use of total
ideas generated as the primary dependent variable. After ability-measurement
problems emerged, it became apparent that using an outcome measure, such as the
number of uses generated or using ability measures, led to an inappropriate analysis
to study the loafing phenomenon accurately as described in previous studies.
Examining the definition of motivation as forwarded by Pintrich and Schunck
(1996), the most appropriate way to measure motivation loss is to assess effort
reduction. Traditionally, this was done in the early loafing literature where subjects
were involved in a tug-of-war exercise. Later studies on social loafing did include a
self-report measure of loafing but did not use the variable as a dependent variable as
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
8 1
occurs in this study. All of this was taken into consideration when the decision was
made to remove not only the achievement and ability measures but the structural-
equation model. This led to the adoption of effort as the primary dependent variable.
This variable becomes a new concept in measuring the loafing phenomenon. It
directly differs from the traditionally held method of examining product-based
outcomes as opposed to process-based outcomes. In fact, the earliest studies of social
loafing primarily assessed effort but did so by examining a proportional reduction of
effort compared to total number of individuals. This was tested on a physical
exertion task and has little applicability to the academic domain.
Other concerns. While the primary concerns were mentioned above, other
minor concerns that surfaced during this study are worth mention as they serve as
potential explanations for data analysis concerns. This study was conducted at the
Foshay Learning Center. This school also was the researcher’s work place, and there
was a high level of familiarity between the students and the researcher. This
potentially allows for students, especially young students like those in sixth grade, to
answer items in a manner that may be inconsistent with their beliefs. Although their
anonymity was guaranteed, there is a distinct possibility that students answered items
in a way they believed the researcher may have wanted them. This was especially
observable in the self-efficacy and collective efficacy measures. Both measures were
negatively skewed and indicated students had a high degree of confidence in both
themselves and their group. This is particularly noteworthy when considering no
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
82
relationship was found between self-efficacy and effort. The absence of significance
may indicate that there was a poor measurement of variables.
This study encountered multiple statistical problems that make the analysis of
the data inteqpretable only with caution. Regression analyses beyond the analysis
contained within this study indicated that there were no significant relationships. All
ofthe above statements do not discount the importance of this study. Statistically, the
study was flawed and will be discussed next. This exploratory study uncovered
findings that demonstrate the need for more in-depih research in this area. This
discovery makes this study a unique contribution to the field of educational
psychology.
Findings of The Study
As mentioned, there was no conclusive data beyond a comparison of
correlation and partial correlation matrices. Regression analyses did not provide
further insight into how the variables interacted. However, several surprising
findings emerged as a result of this study. These findings indicate a significant gap in
the social-loafing literature.
Students in this study rated the ability of the group to succeed and cooperate
as high. In addition, the manipulation of the identifiability variable was successful.
Together, it becomes possible to assess the loafing effect. While collective efficacy
continues to gain popularity in the empirical literature, it is generally accepted that
belief in a group of which one is a member is conducive to high levels of effort. It
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
83
would be logical to infer that members understood that their contributions were
distinguishable from the group project, increased individual effort would follow.
However, this study demonstrated that groups, such as that mentioned above,
show no significant relationship between collective efficacy and effort. In other
words, their expectation for success was not equally met with effort. Partial
correlations suggest there is an inverse relationship when contributions are
identifiable. This has significant relevance to student-centered instruction,
specifically learning methods such as cooperative learning.
These results suggest that students, in groups in which they feel competent,
collectively reduce effort because they assume someone else will accomplish the
task. Considering the age of the students, this suggestion is entirely possible.
Students may feel their contributions would be redundant or simply that someone
else will come up with the idea. Both notions are detrimental in the learning process.
In addition, these findings facilitate the discussion regarding the benefits or
shortcomings of ability grouping in the classrooms. For those teachers who subscribe
to the belief that, in classroom situations, students should work with other students of
similar ability level, the findings in this study reject that belief conclusively. While
this is not a study on the effects of cooperative learning or, specifically, creating
effective learning environments, the findings have direct implications that will be
discussed later.
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
84
As has been discussed throughout this paper, social loafing has been
evaluated historically by examining achievement variables rather than effort. It is
with this thought that concern arises that the cart has been placed before the horse in
this study. In this study, the measurement of specific variables, or the poor
measurement of certain variables, contributed to the overall poor fit of the structural-
equation model and also may have contributed to the lack of statistical
sophistication. However, as the hindsight bias is a strong force in dissertations, and
research in general, one major discovery, which could have practical implications in
teaching, is the need for a statistically evaluated social-loafing scale. This study
contributes several important components to the field, such as the need to include
collective efficacy. However, the scale used and found in appendix A, is a
combination of several scales used to assess constructs, such as effort, persistence,
self-efficacy, and collective efficacy. Collective efficacy emerged as a critical
variable in defining social loafing.
It seems logical to assume that the relationship between self-efficacy and
effort has an important contribution to this finding. Individuals with higher ratings of
efficacy may feel their effort would be so profound and unique that it would create a
free-rider effect and, therefore, would disengage from the group task.
With that finding in mind, it may be logical to infer that when contributions
are easy to identify, the groups with the high ratings of efficacy feel other members
of the group will complete the task and disengage. In other words, a situation may
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
85
occur that becomes similar to a group-level free-rider effect. Conversely, group
members who rate themselves as having high levels of collective efficacy, but have
the opportunity to hide, invest significantly more effort than subjects who report they
do not feel as though they can disengage, from the group task and hide. This may
indicate a lack of trust or simply may be indicative of subj ects answering the
collective efficacy measure in an attempt to please the researcher. Table 1 tends to
refute the possibility that subjects answered in an attempt to please the researcher as
the data is well within the range of normality. However, an item-by-item analysis
(see table 1) may indicate an overwhelming tendency for students to rate collective
efficacy higher than actually occurs on similar tasks.
Further, it can be speculated that difficulty arose evaluating effort on two
levels, group and individual effort. This study viewed effort as an individual’s
interpretation of their own contribution as well as that of the group. This was done
intentionally in order to maintain consistency with the collective efficacy construct.
This may have contributed to the lack of significance between self-efficacy and
effort. Further, neither the regression analysis nor the structural-equation model
demonstrated any relationship between self-efficacy and effort and even showed that
it was one of the weaker relationships when introducing other variables
simultaneously.
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
86
Areas for Improvement
The end of any product is logically the beginning of the subsequent one.
Within that line of logic falls specific issues that, upon further investigation of the
social-loafing phenomenon in school-age children, are appropriate. With respect to
this particular study, the major contributions to the field may not necessarily be in
the statistical findings of this study, but rather in the unique identification of future
areas of improving the way social loafing is defined. This study uncovered one
critical area that needs improvement and may be the most significant contribution of
this research, task selection.
At the outset of this paper, a definition for group was offered that included
interdependence, sense of belonging, equity of influence, and shared goals (Johnson
& Johnson, 1987). Further Bandura (in press) suggests that a critical aspect of
collective efficacy is interdependence of human functioning. When analyzing the
brainstorming task used in this study, it becomes clear that the task used to assess
effort loss within group tasks was not a group-level task at all! In fact, future
research would benefit from examining the different influences that emerge from
simply working in a shared space but not on a cooperative task. While this suggests
an inclusion of social comparative literature, the relevance is worth noting. More
critically, as this study was a replication of prior studies on group-level tasks, it is
suggested that perhaps many of the foundational concepts used to analyze social
loafing were conducted in situations that were not group tasks representative of
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
87
collaborative situations seen in either the classroom or the work place. This poses
obvious concerns when reviewing the research.
Does this mean that all the previous findings can be considered obsolete?
Probably not, but it does allow for the consideration that there is more occurring
within this field than previously thought. While it is understood that group dynamics
is a broad field, and this dissertation is far too narrow in scope to deal with all the
possible solutions, it is important to engage in a discussion about how much research
does transfer to the real-world environments, as that is thought to be the primary
purpose of empirical research. One must consider whether subjects at the age of
those in this study define group within the narrow constraints offered above or if
they simply view shared space as group. Specifically, are the findings within this
study explainable simply under the mere presence hypothesis based on the
assumption this exercise was removed from group work as defined above and
restricted to a discussion of shared workspaces.
Regardless, this discovery is one that not only has empirical relevance but
practical importance as well. As mentioned, this study encountered several statistical
obstacles but still affords a significant amount of insight into the social-loafing
literature. Failing to use a group task to measure group-level attributes creates a
fundamental error in understanding the relevance of social loafing. It has significant
implications with the interpretation of prior research and may indicate that findings
earlier in this field may not be as definitive as the literature suggests. Since the
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
88
majority of social-loafing research has used a brainstorming task (for a detailed list,
refer to Karau & Williams, 1993), this study extends existing knowledge by
demonstrating the need for group tasks that are both interdependent and equitable.
Implications and Directions for Future Research
If this dissertation has shown nothing else, it has contributed to the field by
highlighting critical gaps in the existing research. The findings in this paper have
significant and profound implications for how instruction not only is prepared but
how it is presented.
First, educators would greatly benefit from a user-friendly instrument that
would allow greater assessment of group interaction as well as a method for
monitoring long-term student improvement. By creating a social-loafing scale that is
focused primarily on children, several critical areas would be addressed. Primarily, it
would give teachers a common starting place to begin a dialogue on the critical
components of group work. From data collected in this study, it can be inferred that,
at this point in time, the primary variables involved in assessing individual
participation in group tasks include: self-efficacy, collective efficacy, effort, and
possibly persistence as a trait-level construct. It is important to consider trait-level
constructs in addition to state constructs in order to monitor change. As persistence
was the only variable to be influenced by the identifiability manipulation, it seems
like a logical starting point to begin pedagogical dialogue. Future research would
benefit from creating a scale that would allow teachers to measure individual
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
89
participation within group projects. Specifically, this would allow teachers a
common language to begin a dialogue about pedagogy relative to group tasks. In
addition, a reliable measure of collaborative activity would enable teachers to predict
and address specific needs for students. Taken together, the creation of a reliable
instrument would enable instructors to create productive environments for
collaborative learning as well as prescriptive advice.
Second, as mentioned earlier, a critical area in educational psychology, as it
applies to instruction, is goal orientation of the student. It is anticipated that
examining how groups differ, based on goal orientation, also would shed light on the
influence of social loafing. Social loafing has been clearly defined by Sanna (1992,)
in terms of self-efficacy theory, in that efficacy expectations are an important
variable in understanding effort reduction. As many school tasks, and some may
argue school in general, are performance-goal oriented with few intrinsically
motivating elements, additional insight may be gained by understanding social
loafing from a student’s perspective on goal theory. This study was not designed to
be competitive, and no prizes were offered to increase effort beyond participation in
task. Subjecting younger populations to this research, but including comparative data
regarding mastery and performance goal orientation, will extend the understanding
of social loafing and enhancing classroom environments.
Cooperative learning has been a key activity and buzzword in pedagogical
discourse in classrooms throughout the last 20 years. Group organization based on
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
ability level and gender, to name a few, have created heated debates in educational
settings and have provided models of instruction that include, but are not limited to,
multi-age classrooms, mainstreaming, and basic individual philosophies regarding
cooperative group composition. This study indicates no gender differences exist
within this population. While this may be a little extreme to consider as a definitive
conclusion, it opens the door to begin a discussion about group-level attributes.
Further, grouping based on ability is out of fashion. The findings herein may present
evidence against programs such as gifted and talented education and argue for
educational concepts such as mainstreaming, but the effects may not be so severe.
Practically speaking, this research indicates it is important for the instructor to
understand the dynamics of the classroom with regard to the variables presented in
this study. Although the concept of group familiarity, or working with one’s friends,
was not addressed in this study, future research may benefit from examining the
influence of group familiarity on social loafing. While Karau and Williams (1997)
evaluated this construct previously, a study comparing the construct’s cohesiveness,
trust, and collective efficacy would clarify the social-loafing phenomenon further,
specifically with children. While this tends to lean toward the psychological domain
of in-group and out-group membership, looking simply at how friends work together
may provide some practical information to teachers. A flaw in the design of this
study was the intentional omission of the cohesiveness construct (Karau & Williams,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
1997). Future research would greatly benefit from including this construct, as well as
collective efficacy, and comparing the influence on invested mental effort.
Finally, this study provides a starting point to discuss whether activities and
the educational process contribute to socialized loafing. In other words, are students,
during the course of their elementary school career, being inherently taught to
disengage from tasks? Previous research mentioned in this paper hinted at the need
to explore cohesiveness as a means of increasing productivity in the classroom. In
addition, task value and individual status within groups also must be considered. As
this study was not developmental in nature, it appears that middle school students are
easily distracted from the task at hand for reasons discussed earlier. Replicating this
study with the suggested modifications may add to our knowledge further about how
patterns of behavior develop in school-age children. Finally, it has been suggested in
the social-loafing literature that individuals may prefer tasks in which they can
reduce effort. All of the factors above need to be examined to provide empirical and
practical evidence regarding group dynamics in the classroom.
In conclusion, this study extends existing social-loafing research by
uncovering several interesting points that need to be addressed in future research.
Students in urban middle schools have a tendency to reduce their effort when they
feel secure in the group’s ability, and relax this effort even when they recognize their
individual work will be easily recognized. Future research would benefit from
addressing social loafing from a collective perspective. In addition, it is important to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
conduct comparative analyses across socioeconomic groups to clarify further the
differences in motivational composition in young populations. It is critical to gain a
full understanding of the area under investigation. By increasing our understanding
and comprehending the processes, our knowledge regarding the theory and practice
will be greatly enhanced.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
References
Abrams, D., & Hogg, M. A. (1990), Social Identity Theory: Constructive and
Critical Advances. New York: Springer-Verlag.
Anderson, J. R. (1983). The Architecture of Cognition. Mahwali, New Jersey:
Lawrence Erlbaum Associates.
Atkinson, J. W. (1957). Motivational determinants of risk-taking behavior.
Psychological Review. 64 .359-372.
Bandura, A. (1982). Self-Efficacy Mechanism in Human Agency. American
Psychologist. 3 7 .122-147. '
Bandura, A. (1986). Social Foundations of Thought and Action: A Social
Cognitive Theory. Englewood Cliffs, N.J.: Prentice Hall.
Bandura, A. (1991). Self-regulation of motivation through anticipatory and
self-regulatory mechanisms. In R. A. Dienstbier (Ed.) Perspectives on Motivation:
Nebraska Symposium on Motivation (Vol. 38, pp. 69-164).
Bandura, A. (1997). Self-Efficacy: The Exercise of Control. New York: W.
H. Freeman and Company.
Bandura. A. (in press). Exercise of human agency through collective efficacy.
Current Directions in Psychological Science.
Betz, N. E. & Hackett, G. (1981). The relationship of career-related self-
efficacy expectations to perceived career options in college women and men. Journal
of Counseling Psychology. 28, 399-410.
Betz, N. E. & Hackett, G. (1983). The relationship of mathematics self-
efficacy expectations to the selection of science-based college majors. Journal of
Vocational Behavior. 23, 329-345.
Button, S. B., Mathieu, J. E., & Aikin, K. J. (1996). An examination of the
relative impact of assigned goals and self-efficacy on personal goals and
performance over time. Journal of Applied Social Psychology. 26 (12). 1084- 1103.
Byrne, B. (1996). Measuring Self-Concept Across the Life Span: Issues and
Instrumentation. Washington D.C. American Psychological Association.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
Carver, C. S. & Scheier, M. F. (1981). The self-attention induced feedback
loop and social facilitation. Journal of Experimental Social Psychology. 17. 545-568.
Comer, D. R. (1995). A Model of Social Loafing in Real Work Groups.
Human Relations. 48 (6). 647-665.
Covington, M. V. (1992). Making the Grade: A Self-Worth Perspective on
Motivation and School Reform. Cambridge: Cambridge University Press.
Covington, M. V. & Omelich, C. L. (1991). Need achievement revisited:
Verification of Atkinson’s original 2 x 2 model. In C. D. Spielberger, I. G. Sarason,
Z. Kulcsar, & G. L. Van Heck (Eds.), Stress and Emotion: Anxiety. Anger, and
Curiosity (Vol. 14, pp. 85-105). Washington D.C.: Hemisphere.
Covington, M. V. & Roberts, B. (1994). Self-worth and college achievement:
motivational and personality correlates. In P. R. Pintrich, D. R. Brown, & C. E.
Weinstein (Eds.), Student Motivation, Cognition, and Learning: Essays in Honor of
Wilbert JMcKeachie (pp. 157-187). Hillsdale, NJ: Erlbaum.
Dweck, C. S., & Elliot, E. S. (1983). Achievement Motivation. In P. H.
Mussen (Ser. Ed.) & E. M. Heatherington (Vol. Ed.), Handbook of Child
Psychology: Vol 4. Socialization, Personality, and Social Development (4t h ed., pp.
643-691). New York: Wiley.
Eccles, J. S., & Midgely, C. (1989). State-environment fit: Developmentally
appropriate classrooms for young adolescents. In R. Ames & C. Ames (Eds.),
Research on motivation in education. Vol. 3, Goals and Cognitions (pp. 139-186).
New York: Academic.
Eccles, J. S., Midgley, C., & Adler, T. (1984). Grade related changes in the
school environment: Effects on achievement motivation. In J. Nichools (Ed.),
Advances in Motivation and Achievement: The Development of Achievement
Motivation (Vol. 3, pp. 283-331). Greenwich, CT: JAI PresSi
Eccles, J. S. & Wigfield, A. (1995). In the mind of the actor: The structure of
adolescents’ achievement task values and expectancy-related beliefs. Personality and
Social Psychology Bulletin, 2 1 ,215-225.
Esposito, M.K. (1998). Team efficacy beliefs: Their application and
measurement. Unpublished doctoral dissertation. University of Southern California.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
Femandez-Ballesleros, R., Diez-Nicolas, J., Caprara, G. V., & Barbaranelli,
C. & Bandura, A. Determinants and Structural relation of personal efficacy to
collective efficacy, (in press).
Gabrenya, W. K., Jr., Latane, B., & Wang, Y. E. (1983). Social loafing in
cross-cultural perspective: Chinese in Taiwan. Journal of Cross-Cultural Psychology.
14,368-384.
Gabreyna, W. K., Jr., Latane, B., & Wang, Y. E. (1985). Social loafing in
cross cultural perspective: Chinese in Taiwan. Journal of Cross-Cultural
Psychology, 14, 368-384.
Geen, R. G. (1979). Effects of being observed on learning following success
and failure experiences. Motivation and Emotion, 3 (4), 355-371.
Geen, R. G. (1979). Social Motivation. Annual Review of Psychology, 42,
377-399.
Gist, M. E. (1987). Self-efficacy: Implications for organizational behavior
and human resource management. Academy of Management Review. 12,472-485.
Guzzo, R. A., Yost, P. R., Campbell, R. J., & Shea, G. P. (1993). Potency in
groups: Articulating a construct. British Journal of Social Psychology, 32, 87-106.
Hackett, G. & Betz, N. E. (1981). A self-efficacy approach to the career
development of women. Journal of Vocational Behavior. 18. 326-339.
Hackman, J. R. & Morris, C. G. (1975). Group tasks, group interaction,
process, and group performance effectiveness: A review and proposed integration. In
L. Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol 8, pp. 45-99).
San Diego, CA: Academic Press.
Hardy, C. J. & Latane, B. (1988). Social loafing in cheerleaders: Effects of
team membership and competition. Journal of Sport and Exercise Psychology, 10,
109-114.
Harkins, S. (1987). Social loafing and social facilitation. Journal of
Experimental Social Psychology. 2 3 ,1-18.
Harkins, S. G. & Jackson, J. M. (1985). The role of evaluation in eliminating
social loafing. Personality and Social Psychology Bulletin, 11 (4), 457-465.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
Harkins, S. G. & Petty, R. E. (1982). Effects on task difficulty and task
uniqueness on social loafing. Journal of Personality and Social Psychology. 43 (6),
1214-1229.
Harkins, S. G. & Szymanski, K. (1987). Social loafing and social facilitation:
New wine in old bottles. In C. Hendrick (Ed.), Review of Personality and Social
Psychology (Vol. 9, pp. 167-188). Newbury Park, CA: Sage.
Harkins, S. G. & Szymanski, K. (1988). Social loafing and self-evaluation
with an objective standard. Journal of Experimental Social Psychology, 24, 354-365.
Harkins, S. G. & Szymanski, K. (1989). Social loafing and group evaluation.
Journal of Personality and Social Psychology. 56. 934-941.
Harter, S. (1990). Causes, correlates, and the functional role of self-worth: A
life-span perspective. In R. J. Sternberg & J. Kolligan (Eds.), Competence
Considered (pp. 67-97), New Haven, CT: Yale University Press.
Harter, S. (1981). A new self-report scale of intrinsic versus extrinsic
orientation in the classroom: Motivational and informational components.
Developmental Psychology, 17,300-312.
Hodges, L., & Carron, A. V. (1992). Collective efficacy and group
performance. International Journal of Sport Psychology, 2 3 .48-59.
Hu, L., & Bentler, P. (1998). Fit indices in covariance structure modeling:
Sensitivity to underparameterized model misspecification. Psychological Methods, 3
(4), 424-453.
Hu, L., & Bentler, P. (1999). Cutoff criteria for fit indices in covariance
structure analysis: Conventional criteria versus new alternatives. Structural Equation
Modeling, 6 (1), 1-55.
Hunt, P. J., & Hillery, J. M. (1973). Social facilitation in a coactive setting.
Journal of Experimental Social Psychology, 9, 563-571.
Ingham, A. G., Levinger, G., Graves, J, & Peckham, V. (1974). The
Ringelmann e ffe c t: Studies of group size and group performance. Journal of
Experimental Social Psychology, 10. 371-384.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
Jackson, J. M., & Harkins, S. G. (1985). Equity in effort: An explanation of
the social loafing effect. Journal of Personality and Social Psychology, 4 9 .1199-
1206.
Jackson, J. M., & Williams, K. D. (1985). Social loafing on difficult tasks:
Working collectively can improve performance. Journal of Personality and Social
Psychology, 49 (4), 937-942.
Johnson, D. W., & Johnson, F. P. (1987). Joining Together: Group Theory
and Group Skills. Englewood Cliffs, NJ: Prentice-Hall.
Karau, S. J., & Hart, J. W. (in press). Group cohesiveness and social loafing.
Karau, S. J., & Williams, K. D. (1993). Social Loafing: A meta-analytic
review and theoretical integration. Journal of Personality and Social Psychology. 65
(4), 681-706.
Karau, S. J., & Williams, K. D. (1995). Social loafing: Research findings,
implications, and future directions. Current Directions in Psychological Science, 4
(5), 134-139.
Karau, S. J., & Williams, K. D. (1997). The effects of group cohesiveness on
social loafing and social compensation. Group Dynamics, 1 (2), 156-168.
Kerr, N.L. (1983). Motivation losses in small groups: A social dilemma
analysis. Journal of Personality and Social Psychology, 45, 819-828.
Kerr, N. L., & Braun, S. E. (1981). Ringelmann revisited: Alternative
explanations for the social loafing effect. Personality and Social Psychology
Bulletin, 7 (2). 224-231.
Kerr, N. L., & Braun, S. E. (1983). The dispensibility of member effort and
group motivation losses: Free rider effects. Journal of Personality and Social
Psychology, 44, 78-94.
Kerr, N. L., & Stanfel, J. A. (1993). Role schemata and member motivation
in task groups. Personality and Social Psychology Bulletin. 19 (4), 432-442.
Kravitz, D. A., & Martin, B. (1986). Ringelmann rediscovered: The original
article. Journal of Personality and Social Psychology, 50, 936-941.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
Latane, B. (1981). The psychology of social impact. American Psychologist.
36,343-356.
Latane, B., Williams, K., & Harkins, S. (1979). Many hands make light the
work: The causes and consequences of social loafing. Journal of Personality and
Social Psychology, 37 (6). 822-832.
Lent, R. W., Brown, S. D., & Larkin, K. C. (1986). Self-efficacy in the
prediction of academic performance and perceived career options. Journal of
Counseling Psychology, 33 (3). 265-269.
Lindsley, D. H., Brass, D. J., & Thomas, J. (1995). Efficacy-performance
spirals: A multilevel perspective. Academy of Management Review. 20. 645-678.
Locke, E. A., & Latham, G. P. (1990). A Theory of Goal Setting and Task
Performance. Englewood Cliffs, NJ: Prentice Hall.
Locke, E. A., Motowidlo, S. J., & Bobko, P. (1986). Using self-efficacy
theory to resolve the conflict between goal setting theory and expectancy theory in
organizational behavior and industrial/organizational psychology. Journal of Social
and Clinical Psychology, 4 (3), 328-338.
Locke, E. A., & Shaw, K. N. (1984). Atkinson’s inverse-U curve and the
missing cognitive variables. Psychological Reports. 5 5 .403-412.
Los AngelesUnified School District. (2000). CBEDS Report. [On-Line].
Available: http://search.lausd.kl2.ca.us/cgi-
in/fccgi.exe?w3exec=cbeds3&which=8132.
Malpass, J. R. (1994). A Structural Model of Self-Regulation, Goal
Orientation, Self-Efficacy, Worry, and Mathematics Achievement. Unpublished
doctoral dissertation, Los Angeles: University of Southern California.
Markus, H. (1978). The effect of mere presence on social facilitation: An
unobtrusive test. Journal of Experimental Social Psychology, 14, 389-397^
Marsh, H. W. (1989). Age and sex effects in multiple dimensions of self-
concept: Preadolescence to early adulthood. Journal of Educational Psychology, 81
(31417-430.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
Mulvey, P. W., & Klein, H. J. (1998). The impact of perceived loafing and
collective efficacy on group goal processes and group performance. Organizational
Behavior and Human Decision Processes. 74 (1). 62-87.
Olson, M. (1965). The Logic of Collective Action: Public Goods and the
Theory of Groups. Cambridge, MA: Harvard University Press.
Pajares, F. (1996). Self-efficacy beliefs in academic settings. Review of
Educational research, 66 (4), 543-578.
Piers, E. V., & Harris, D. B. (1964). Age and other correlates of self-concept
in children. Journal of Educational Psychology, 55 (2). 91-95.
Pintrich, P. R. (1989). The dynamic interplay of student motivation and
cognition in the college classroom. In C. Ames & M. L. Maehr (Eds.), Advances in
Motivation and Achievement: Motivation Enhancing Environments (Vol. 6, pp. 117-
160). Greenwich, CT: JAI Press.
Pintrich, P. R., & DeGroot, E. (1990). Motivational and self-regulated
learning components of classroom academic performance. Journal of Educational
Psychology, 82, 33-40.
Pintrich, P. R., & Garcia, T. (1991). Student goal orientation and self
regulation in the college classroom. In M. L. Maehr & P. R. Pintrich (Eds.)
Advances in Motivation and Achievement: Goals and Self-Regulatory Processes
(Vol. 7, pp. 371-402). Greenwich, CT: JAI Press.
Pintrich, P. R., & Schunk, D. H. (1996). Motivation in Education: Theory,
Research, and Applications. Englewood Cliffs, NJ: Prentice Hall.
Rotter, J. (1971). Self-disclosure and interpersonal trust: An exploratory
study. Psychological Reports, 28,235-240
Ruble, D. N., & Frey, K. (1991). Changing patterns of comparative behavior
as skills are acquired: A functional model of self-evaluation. In J. Suls & T. Wills
(Eds.), Social Comparison: Contemporary Theory and Research (pp. 79-113).
Hilsdale, NJ: Erlbaum.
Salomon, G. (1984). Television is “easy” and print is “tough”: The
differential investment of mental effort in learning as a function of perceptions and
attributions. Journal of Educational Psychology, 76 (4), 647-658.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 0
Sanders, G. S., & Baron, R. S. (1975). The motivating effects of distraction
of task performance. Journal of Personality and Social Psychology, 3 2 ,956-963.
Sanna, L. J. (1992). Self-efficacy theory: Implications for social facilitation
and social loafing. Journal of Personality and Social Psychology, 62 (51, 774-786.
Sanna, L. J., & Shotland, R. L. (1990). Valence of anticipated evaluation and
social facilitation. Journal of Experimental Social Psychology, 26, 82-92.
Schachter, S. (1959). The Psychology of Affiliation: Experimental Studies of
the Sources of Gregariousness. Stanford. CA: Stanford University Press.
Schunk, D. H. (1983). Developing children’s self-efficacy and skills: The
roles of social comparative information and goal setting. Contemporary Educational
Psychology, 8, 76-86.
Schunk, D. H. (1984). Self-efficacy perspective on achievement behavior.
Educational Psychologist. 19.48-58.
Schunk, D. H. (1991). Self-efficacy and academic motivation. Educational
Psychologist, 2 6 ,207-231.
Singley, M. K., & Anderson, J. R. (1989). The Transfer of Cognitive Skill.
Cambridge: Harvard University Press.
Shirikashi, S. (1985). Social loafing of Japanese students. Hiroshima Forum
for Psychology, 10, 35-40.
Sorrentino, R. M., & Higgins, E. T. (1996). Handbook of Motivation and
Cognition (Volume 3). New York: The Guilford Press.
Spink, K. S. (1990). Collective efficacy in the sport setting. International
Journal of Sport Psychology, 2 1 ,380-395.
Steiner, I. (1972). Group Processes and Productivity. San Diego, CA:
Academic Press.
Swain, A. (1996). Social Loafing and Identifiability: The Mediating Role of
Achievement Goal Orientations. Research Quarterly for Exercise and Sport, 67 (31,
337-344.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
Szymanski, K., & Harkins, S. G. (1987). Social loafing and self-evaluation
with a social standard. Journal of Personality and Social Psychology. 53 (5). 891-
897.
Taylor, M. S., Locke, E. A., Lee, C., & Gist, M. E. (1984). Type A behavior
and faculty research productivity: What are the mechanisms? Organizaitonal
Behavior and Human Performance, 3 4 ,402-418.
Taylor, S. E., Wayment, H. A., & Carrillo, M. (1996). Social comparison,
self-regulation, and motivation. In R. M. Sorrentino & E. Tory Higgins (Eds.),
Handbook of Motivation and Cogniton: Vol. 3. The Interpersonal Context. New
York: The Guilford Press.
Vroom, V. H. (1964). Work and Motivation. New York: Wiley.
Waller, J. E. (1996). Social loafing and the group-evaluation effect: Effect of
dissimilarity in a social comparison standard. Psychological Reports, 78,177-178.
Wigfield, A., & Eccles, J. (1992). The development of achievement task
values: A theoretical analysis. Developmental Review, 12,265-310.
Williams, K. D., Harkins, S., & Latane, B. (1981). Identifiability as a
deterrent to social loafing: Two cheering experiments. Journal of Personality and
Social Psychology, 40,303-311.
Williams, K. D., Jackson, J. M., & Karau, S. J. (1995). Collective hedonism:
A social loafing analysis of social dilemmas. In D. A. Schroeder (Ed.), Social
Dilemmas: Perspectives on Individuals and Groups. Westport: Praeger.
Williams, K. D., & Karau, S. J. (1991). Social loafing and social
compensation: The effects of expectations of co-worker performance. Journal of
Personality and Social Psychology, 61 (4), 570-581.
Williams, K. D., Karau, S. J., & Bourgeois, M. (1993). Working on collective
tasks: Social loafing and social compensation. In M. A. Hogg and D. Abrams (Eds.),
Group Motivation: Social Psychological Perspectives (pp.130-148). New York:
Harvester Wheatsheaf.
Zajonc, R. (1965). Social facilitation. Science, 149,269-274.
Zimmerman, B. J. (1985). The development of “intrinsic” motivation: A
social learning analysis. Annals of Child Development. 2 ,117-160.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
Zimmerman, B. J., & Martinez-Pons, M. (1990). Student differences in self
regulated learning: Relating grade, sex, and giftedness to self-efficacy and strategy
use. Journal of Educational Psychology. 82, 51-59.
Zimmerman, B. J., & Ringle, J. (1981). Effects of model persistence and
statements of confidence on children’s efficacy and problem solving. Journal of
Educational Psychology, 73 ,485-493. ■
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix A
State Social Loafing Scale for Children
ID#
Gender M F
Using the 5-point scale below, indicate how you feel about the following items.
1 2 3 4 5
no confidence very little confidence some confidence much confidence very confident
Sample: I am confident that I can give directions to my house.
1.............2.............3..............4..............5
Part I
1. I am confident that I can follow the directions of this task.
1..............2..............3............. 4.............5
2. I am confident that my group will successfully follow the directions of the task.
1............ 2 ............ 3..............4........ 5
3. I am confident that my group members will work hard on this task.
1..............2..............3............. 4.............5
4. I am confident that my group members will all contribute equally to this task.
1............. 2........ ....3..............4............. 5
5. I am confident that I will think of new uses for this object.
1.............. 2..............3............. 4.............5
6. I am confident that my group members have the ability to complete this task.
1..............2..............3............. 4........ ....5
7. I am confident that my group members will try to complete the task.
1..............2..............3............. 4.............5
8. I am confident that my group members will all cooperate with each other.
1............. 2.............3............4............. 5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
9. I am confident that I will complete this task in the time given.
1.............2...........3.............4..............5
10.1 am confident that my group can think of uses for an object.
1.............2 ............ 3.............4..............5
11.1 am confident that my group will keep working even if they are distracted.
1.............2.............. 3............ 4............. 5
12.1 am confident that my group will be able to cooperate without arguing to finish
the task.
1............. 2.............. 3............ 4............. 5
13.1 am confident that I can work for the entire time, even if I am distracted.
1.............2.............. 3............ 4............. 5
14.1 am confident that my group will get along.
1............. 2.............. 3.............4............. 5
Partll
Using the 5-point scale below, indicate how you feel about the following items.
1 2 3 4 5
never once in a while sometimes very often
always
Sample: I like to eat hot dogs for breakfast.
1............. 2............ .3..............4..............5
15. I complete many of the activities I begin.
1.............2...... 3..............4 ..............5
16.1 work as hard as I can no matter what I am doing.
1.............2............. 3..............4..............5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
17. While I am doing my homework, I like to take breaks.
1............ 2..............3.......... 4...........5
18.1 work as hard as possible, even if I am being bothered.
1........... 2..............3 ........... 4..............5
19. When I watch television, I like to see the programs from beginning to end.
1............ 2............. 3..............4..............5
20.1 put a lot of effort in most of the things I do.
1............ 2 ............ 3.............4.............. 5
21.1 give up easily when I do not succeed.
1............ 2............. 3..............4..............5
22. If I make a spelling error at the end of a paper, I will rewrite the entire paper so it
is perfect.
1..............2....... 3............. 4 ............. 5
23.1 work hard at everything I do, even if it is really hard.
1.............2............3............. 4........ 5
24.1 try several solutions to math problems before I ask for help from the teacher.
1...... 2..............3............. 4 ..............5
25. When I play video games, I finish the game even if I am doing poorly.
1............. 2............. 3..............4..............5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
STOP. Do not continue working. Close this document and wait for directions.
Using the 5-point scale below, indicate how you feel about the following items.
1 2 3 4 5
not at all not very hard quite a bit rather hard
extremely hard
Sample: How hard do you try to get a perfect score on a test?
1.............2...........3.............4..............5
1. How hard did you try to understand the brainstorming task?
1.............2............. 3 ............4..............5
2. How hard did the other members of the group try to understand the brainstorming
task?
1. ............2 ............3............. 4 ...............5
3. How much did you concentrate while participating in the brainstorming task?
1 .. .......... 2 ............3 .............4............. 5
Using the 5-point scale below, indicate how you feel about the following items.
1 2 3 4 5
very hard sort of hard not very hard sort of
easy very easy
Sample: Athletes work_______________ to perform well during competitions.
1.............2............. 3..............4..............5
very hard sort of hard not very hard sort of easy very easy
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
1. It will be__________________for Mr. Miller to identify which ideas in the box
are mine.
1.............2............. 3..............4 ..............5
very hard sort of hard not very hard sort of easy very easy
2. It will be__________________ for Mr. Miller to identify who did the most work
in the group.
1.... 2.............3..............4.............5
very hard sort of hard not very hard sort of easy very easy
3. After putting the paper in the box, it will b e_________________________to
tell who worked hard and who did not.
1.............2..............3............. 4 ............. 5
very hard sort of hard not very hard sort of easy very easy
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix B
Structural Equation Models
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
o
j r
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
licai 14
110
\
£
o «
o
m
«r
I
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Factors that influence students' investment of mental effort in academic tasks: A validation and exploratory study

PDF

Change beliefs and academic outcomes: A construct validation

PDF

An adolescent -parent conflict resolution skills training program for ethnically diverse families: A program evaluation study

PDF

Individual motivation loss in group settings: An exploratory study of the social-loafing phenomenon

PDF

Commitment to deliberate practice in the development of writing expertise: A survey of the motivational patterns of academic writers in Tier One research universities

PDF

Effects of teaching self -monitoring in a distance learning course

PDF

Cognitive aging: Expertise and fluid intelligence

PDF

Comparing social studies planning processes of experienced teachers and student teachers

PDF

Examining the relationship between writing self -efficacy, writing performance and general achievement for third graders

PDF

Ideological threat, social identity, and consensus estimation

PDF

Critical thinking cognitive abilities and the analytical section of the educational testing service Graduate Record Examinations(R) General Test: A conceptual study in content validity

PDF

Attachment style, interpersonal guilt, parental alcoholism, parental divorce and eating disordered symptomatology in college women

PDF

Assessment of racial identity and self -esteem in an Armenian American population

PDF

Atypicality: Benefit and bane. Provocation, trigger, typicality, and the expression of aggression and generalization

PDF

A phenomenological inquiry into the essential meanings of the most intense experience of religiosity and spirituality

PDF

Can individual creativity be enhanced by training? A meta analysis

PDF

Inaccuracies in expert self -report: Errors in the description of strategies for designing psychology experiments

PDF

A contemporary urban ethnography of Pakistani middle school students in Oslo, Norway

PDF

Alcohol treatment entry and refusal in a sample of older veterans

PDF

Capital and punishment: The privatization of prisons in America

Asset Metadata

Creator Miller, Jeffrey Allan (author)

Core Title Individual motivation loss in group settings: An exploratory study of the social -loafing phenomenon

School Graduate School

Degree Doctor of Philosophy

Degree Program Educational Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag education, educational psychology,OAI-PMH Harvest,psychology, social

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Clark, Richard (committee chair), [illegible] (committee member), Guiton, Gretchen (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-104642

Unique identifier UC11327940

Identifier 3027754.pdf (filename),usctheses-c16-104642 (legacy record id)

Legacy Identifier 3027754.pdf

Dmrecord 104642

Document Type Dissertation

Rights Miller, Jeffrey Allan

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 au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA