THE MODERATING EFFECTS OF SELF AND OTHER EFFICACY ON MOTIVATION
GAINS IN SWIMMING RELAYS
By
Kaitlynn Sedabres

A DISSERTATION
Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of
Kinesiology-Doctor of Philosophy

2013

ABSTRACT
THE MODERATING EFFECTS OF SELF AND OTHER EFFICACY ON MOTIVATION
GAINS IN SWIMMING RELAYS
By
Kaitlynn Sedabres
This dissertation investigated the moderating effects of self- and other-efficacy on the
Köhler motivation gain effect in swimming relays. The study was an extension of the motivation
gain literature in sport. The Köhler effect has been found to promote motivation gains for inferior
group members when participating with moderately more capable partners. Self-efficacy
research has also demonstrated a strong relationship with performance in sport. However,
research on efficacy beliefs about others in one’s group indicates other efficacy may be a
stronger predictor of performance compared to self-efficacy (Dunlop, Beatty, & Beauchamp,
2011). Participants were 199 swimmers at the Division II and III levels who swam the 200, 400,
or 800 yard freestyle relay at their fall invitational meets. Both relay times performed at the meet
as well as individual best times for each participant were collected. Participants also completed
questionnaires regarding their self- and other efficacy beliefs of their relay performances. Using
an HLM cross-classified model, the results indicated that the fourth ranked member performed
faster in the relay compared to their individual performance, demonstrating a motivation gain.
Further, under conditions of high self- or other efficacy, this effect was modified by gender. In
female relays, only the fourth ranked member showed a motivation gain, while in male relays
both the third and fourth ranked relay member demonstrated a motivation gain. Findings of the
study contribute to the motivation gain literature in sport and how both self- and other efficacy
can successfully change performance of weaker relay members.

iv

ACKNOWLEDGMENTS

A special thank you to my advisor and dissertation chair, Dr. Deborah Feltz for your guidance
and support throughout this program.

Thank you to my dissertation committee members, Drs. Dan Gould, John Hollenbeck, and Al
Smith for your guidance on this dissertation project.

Thank you to Emery Max and David Reyes-Gastelum whose help in this project was pivotal to
its completion.

Thank you to Alisha and Alison; you have been my rocks throughout this process. I couldn’t
have done it without you!

Lastly, a special thank you to my family, whose efficacy in my abilities has never wavered.

iii

TABLE OF CONTENTS
LIST OF TABLES ............................................................................................................ vii
LIST OF FIGURES ........................................................................................................... ix
CHAPTER 1
Introduction ............................................................................................................ 1
Purpose of the Study............................................................................................. 10
Contextual Factors ................................................................................................ 10
Hypotheses ........................................................................................................... 11
Research Questions .............................................................................................. 11
Delimitations ........................................................................................................ 12
Definitions ............................................................................................................ 12
CHAPTER 2
REVIEW OF LITERATURE
Group Motivation Research ................................................................................. 14
The Köhler Motivation Gain Effect ..................................................................... 17
Self-Efficacy and Performance............................................................................. 24
Other Efficacy ...................................................................................................... 25
Summary............................................................................................................... 32
CHAPTER 3
METHOD
PILOT STUDY
Participants ........................................................................................................... 34
Measures ............................................................................................................... 34
Demographics ................................................................................................ 34
Motivation Gain............................................................................................. 35
Procedure .............................................................................................................. 35
Standardization of Relay Times ........................................................................... 36
Results .................................................................................................................. 37
Discussion............................................................................................................. 38
DISSERTATION
Participants ........................................................................................................... 40
Measures ............................................................................................................... 40
Demographics ................................................................................................ 40
Self-Efficacy .................................................................................................. 41
Other Efficacy ............................................................................................... 41
RISE .............................................................................................................. 41
RPE ................................................................................................................ 42
Motivation Gain............................................................................................. 42
Procedure .............................................................................................................. 42
Treatment of Data ................................................................................................. 44
iv

CHAPTER 4
RESULTS
Descriptives .......................................................................................................... 48
Bivariate Correlations ........................................................................................... 53
Hypothesis Testing ............................................................................................... 53
Research Questions .............................................................................................. 74
CHAPTER 5
DISCUSSION
Performance Results ........................................................................................... 114
Moderating Effects of Efficacy Beliefs .............................................................. 116
Strengths and Limitations ................................................................................... 123
Implications ........................................................................................................ 125
Future Directions ................................................................................................ 126
Conclusion .......................................................................................................... 127
APPENDICES ................................................................................................................ 129
Appendix A Demographic Questionnaire ........................................................... 130
Appendix B Self-Efficacy Scale, Individual, Female 50 Freestyle .................... 131
Appendix C Self-Efficacy Scale, Individual, Male 50 Freestyle ........................ 132
Appendix D Self-Efficacy Scale, Individual, Female 100 Freestyle .................. 133
Appendix E Self-Efficacy Scale, Individual, Male 100 Freestyle ...................... 134
Appendix F Self-Efficacy Scale, Individual, Female 200 Freestyle. .................. 135
Appendix G Self-Efficacy Scale, Individual, Male 200 Freestyle...................... 136
Appendix H Self-Efficacy Scale, Relay, Female 50 Freestyle ........................... 137
Appendix I Self-Efficacy Scale, Relay, Male 50 Freestyle ................................ 138
Appendix J Self-Efficacy Scale, Relay, Female 100 Freestyle .......................... 139
Appendix K Self-Efficacy Scale, Relay, Male 100 Freestyle ............................. 140
Appendix L Self-Efficacy Scale, Relay, Female 200 Freestyle.......................... 141
Appendix M Self-Efficacy Scale, Relay, Male 200 Freestyle ............................ 142
Appendix N Other Efficacy Scale, Female 200 Freestyle Relay ........................ 143
Appendix O Other Efficacy Scale, Male 200 Freestyle Relay ........................... 144
Appendix P Other Efficacy Scale, Female 400 Freestyle Relay ........................ 145
Appendix Q Other Efficacy Scale, Male 400 Freestyle Relay ........................... 146
Appendix R Other Efficacy Scale, Female 800 Freestyle Relay. ....................... 147
Appendix S Other Efficacy Scale, Male 800 Freestyle Relay. ........................... 148
Appendix T Relation-Inferred Self-Efficacy Scale, Female 200 Freestyle Relay.149
Appendix U Relation-Inferred Self-Efficacy Scale, Female 400 Freestyle Relay150
Appendix V Relation-Inferred Self-Efficacy Scale, Female 800 Freestyle Relay.151
Appendix W Relation-Inferred Self-Efficacy Scale, Male 200 Freestyle Relay 152
Appendix X Relation-Inferred Self-Efficacy Scale, Male 400 Freestyle Relay . 153
Appendix Y Relation-Inferred Self-Efficacy Scale, Male 800 Freestyle Relay . 154
Appendix Z Pre Relay Questionnaire ................................................................. 155
Appendix AA Post Relay Questionnaire ............................................................ 156
Appendix AB Confidentiality Assurance. .......................................................... 157
v

REFERENCES ............................................................................................................... 158

vi

LIST OF TABLES

Table 1 Pilot Study Swimmer Demographics ................................................................... 49
Table 2 Estimated marginal performance means and confidence intervals between
ranks across conditions ..................................................................................................... 50
Table 3 Dissertation Study Swimmer Demographics ....................................................... 51
Table 4 Dissertation Study Relay Variables ..................................................................... 52
Table 5 Correlations between self-efficacy, other efficacy, RISE beliefs and individual
and relay performances for the 200 yard freestyle relay................................................... 54
Table 6 Correlations between self-efficacy, other efficacy, RISE beliefs and individual
and relay performances for the 400 yard freestyle relay. .................................................. 55
Table 7 Correlations between self-efficacy, other efficacy, RISE beliefs and individual
and relay performances for the 800 yard freestyle relay................................................... 56
Table 8 Model 1 for performance time ............................................................................. 61
Table 9 Model 1 Estimated mean times and percent changes between relay and
individual performance ..................................................................................................... 62
Table 10 Model 2 for performance time, including the moderating effects of
self-efficacy....................................................................................................................... 67
Table 11 Model 2 Estimated mean times and percent changes between relay and
individual performance ......................................................................................................68
Table 12 Model 2 for performance time, including the moderating effects of
other efficacy. ................................................................................................................... 75
Table 13 Model 3 Estimated mean times and performance changes between relay and
individual performance ..................................................................................................... 76
Table 14 Deviance comparisons between hypotheses models ......................................... 80
Table 15 ICCs of random effects pertaining to hypotheses models ................................. 80
Table 16 Model 4 for performance time ........................................................................... 83
Table 17 Model 4 Estimated mean times and percent changes between relay and
vii

individual performance ......................................................................................................84
Table 18 Model 5 for performance time, including the moderating effects of
self-efficacy and gender. ................................................................................................... 90
Table 19 Model 5 Estimated mean times and percent changes between relay and
individual performance for females ...................................................................................92
Table 20 Model 5 Estimated mean times and percent changes between relay and
individual performance for males ......................................................................................94
Table 21 Model 6 for performance time, including the moderating effects of
other efficacy and gender. ................................................................................................. 99
Table 22 Model 6 Estimated mean times and performance changes between relay and
individual performance for females ................................................................................ 101
Table 23 Model 6 Estimated mean times and performance changes between relay and
individual performance for males ................................................................................... 102
Table 24 Model 7 for performance time, including the moderating effects of
RISE beliefs. ................................................................................................................... 108
Table 25 Model 6 Estimated mean times and performance changes between relay and
individual performance ................................................................................................... 109

viii

LIST OF FIGURES

Figure 1. Individual and relay performance by rank in the 200 distance ......................... 63
Figure 2. Individual and relay performance by rank in the 400 distance ......................... 63
Figure 3. Individual and relay performance by rank in the 800 distance ......................... 64
Figure 4. Individual and relay performance by rank with low and high self-efficacy
beliefs at the 200 distance ................................................................................................. 69
Figure 5. Individual and relay performance by rank with low and high self-efficacy
beliefs at the 400 distance ................................................................................................. 70
Figure 6. Individual and relay performance by rank with low and high self-efficacy
beliefs at the 800 distance ................................................................................................. 71
Figure 7. Individual and relay performance by rank with low and high other efficacy
beliefs at the 200 distance ................................................................................................. 77
Figure 8. Individual and relay performance by rank with low and high other efficacy
beliefs at the 400 distance ................................................................................................. 78
Figure 9. Individual and relay performance by rank with low and high other efficacy
beliefs at the 800 distance ................................................................................................. 79
Figure 10. Individual and relay performance by position in the 200 distance.................. 85
Figure 11. Individual and relay performance by position in the 400 distance.................. 85
Figure 12. Individual and relay performance by position in the 800 distance.................. 86
Figure 13. Individual and relay performance by rank and gender with low and high
self-efficacy beliefs at the 200 distance ............................................................................ 94
Figure 14. Individual and relay performance by rank and gender with low and high
self-efficacy beliefs at the 400 distance ............................................................................ 95
Figure 15. Individual and relay performance by rank and gender with low and high
self-efficacy beliefs at the 800 distance ............................................................................ 96
ix

Figure 16. Individual and relay performance by rank and gender with low and high
other efficacy beliefs at the 200 distance ........................................................................ 103
Figure 17. Individual and relay performance by rank and gender with low and high
other efficacy beliefs at the 400 distance ........................................................................ 104
Figure 18. Individual and relay performance by rank and gender with low and high
other efficacy beliefs at the 800 distance ........................................................................ 105
Figure 19. Individual and relay performance by rank with low and high RISE beliefs
at the 200 distance ........................................................................................................... 110
Figure 20. Individual and relay performance by rank with low and high RISE beliefs
at the 400 distance ........................................................................................................... 111
Figure 21. Individual and relay performance by rank with low and high RISE beliefs
at the 800 distance ........................................................................................................... 112

x

CHAPTER 1
Introduction
In certain sports, competition can be categorized as both team and individually based. At
the individual level, one is focused solely on his or her performance. At the team level,
individuals have to focus on both their own performance as well as the performance of others.
Swimming is a great example of a sport in which this dual categorization can be found. The
majority of events swum are individual, however there are also team relays where the
competition is interdependent. In these situations motivation intensities may vary between
individual and group performance. Further, it is practical to assume that swimmers themselves
should hold beliefs about their own capabilities to perform, as well as the capabilities of their
teammates in a relay task. Such beliefs can also influence the performance motivation of an
athlete in sports such as swimming where there are individual and team demands.
Motivation has been a widely studied phenomenon, especially motivation in groups.
Much of the literature has focused on motivation losses, or social loafing. Social loafing is the
tendency for individuals to perform worse in a group situation compared to an individual
situation (Karau & Williams, 1993). A few reasons as to why social loafing occurs may be that
group members either feel less personally identifiable therefore giving less effort to the group
task, or they may be able to “free-ride” on others’ efforts and therefore reduce their efforts
accordingly (Baron & Kerr, 2003).
One social loafing study specifically in regards to swimming relays, has demonstrated
that these social loafing effects can be eliminated if groups are threatened with punishment if a
certain performance time is not met (Miles & Greenberg, 1993). Further in this study, follow-up
questions indicated that individuals worked harder in groups compared to working alone. They

1

also expressed more concern for letting down their team members in a group setting. The results
of this study indicate that individuals put forth more effort while in the group task and it may be
relevant to study performance gains in groups rather than performance losses.
Recent research has focused on motivation gains in which performance increases within a
group setting compared to individual performance (e.g., Weber & Hertel, 2007). Sports teams
may be a good place to study motivation gains specifically because athletes in a competitive
setting are thought to be highly motivated already as the task is important to them. Further, in
terms of practical significance, coaches are interested in understanding how to maximize
individual motivation within teams.
Two cognitive-based theories of motivation gain in groups are social compensation and
the Köhler effect. Social compensation occurs when individuals increase their performance in
order to compensate for other anticipated weaker performances of group members (Karau &
Williams, 1997). This motivation gain is usually demonstrated by the strongest member of the
group who would be able to compensate or make up for weaker group members, thus performing
better in the group setting. The Köhler effect occurs when a less capable individual performs
better in a group setting compared to performing a task individually either through social
comparison or increased feelings of indispensability (Baron & Kerr, 2003).
While both social compensation and the Köhler effect have been demonstrated in the
literature, this dissertation focused on the performance differences of the weakest member of the
group and therefore the theoretical focus was on the Köhler effect. The Köhler effect was
originally studied by Otto Köhler in the 1920s who studied rowing teams. He found that those
who worked in a dyad condition at an arm curling task persisted longer than they did when they
performed the same task individually (Kerr et al., 2007).

2

The Köhler effect has been studied in a variety of both physical and cognitive tasks
(Gockel, Kerr, Soek, & Harris, 2008; Hertel, Deter, & Konradt, 2003; Hertel, Kerr, & Messé,
2000; Hertel, Niemeyer, & Clauss, 2008; Kerr, Messé, Park, & Sambolec, 2005). The Köhler
effect has also been shown to be task dependent. According to Steiner’s (1972) task taxonomy,
conjunctive tasks, or ones where the result of the group is dependent on the performance of the
weakest group member, are seen to provide the greatest motivation gains for those weakest
group members (Weber & Hertel, 2007). Other task demands, like an additive or coactive task
demand have also shown motivation gains, although less so than conjunctive task demands.
Two different processes have been proposed for the Köhler effect: indispensability and
social comparison (Hertel et al., 2008). The first is when individuals feel indispensable to the
group and therefore perform better than their individual performance. This process should occur
especially in conjunctive task demands when the group is especially dependent on the weakest
individual (Hertel et al., 2008). Social comparison occurs when individuals compare themselves
with a superior member of the team and revise their performance goal upward to try to match the
performance of the superior member’s (Kerr et al., 2007).
These processes of indispensability and social comparison are thought to occur
simultaneously in group tasks. However, there are a few moderators of the Köhler effect on
which this dissertation focused: gender and self-efficacy. In terms of gender, women
demonstrate greater performances in group situations compared to their individual performance
under conjunctive demands. Women also display better performances in group situations under
coactive and additive demands but to a lesser degree, respectively (Kerr et al., 2007; Weber &
Hertel, 2007). Men have been shown to exhibit motivation gains when information is provided
about a superior coworker regardless of whether they are in a group task or performing

3

individually. These gender-based findings indicate that social comparison may be more salient
for men; whereas, indispensability may be more salient for women (Weber & Hertel, 2007).
One other moderator of the Köhler effect is self-efficacy. Seok (2004) found that under
conjunctive task demands, individuals performed better at a group task compared to their
individual performance. In this study, self-efficacy was manipulated through false feedback.
When participants were separated into high and low self-efficacy categories, individuals with
low self-efficacy had greater motivation gains than individuals with high self-efficacy. Seok
reasoned that the high self-efficacy individuals may have been overconfident in their abilities for
the subsequent trial. Further, these high self-efficacy participants may have felt like they did not
have to prove themselves to either the researcher or their partner, so they may have decreased
their effort.
While Seok (2004) looked directly at the relationship between the Köhler motivation gain
effect and self-efficacy, efficacy research in sport has looked directly at the relationship between
efficacy and performance. One meta-analysis has demonstrated that there is a moderate and
positive correlation between self-efficacy and performance (Moritz, Feltz, Fahrback, & Mack,
2000). This relationship is strengthened when measurements between performance and efficacy
beliefs are both concordant as well as assessed close together. Further, this relationship is seen to
be reciprocal in that self-efficacy increases performance while performance also enhances
subsequent self-efficacy (Feltz, 1982; McAuley, 1985).
One extension of self-efficacy theory was proposed by Lent and Lopez (2002) in their
tripartite model of efficacy beliefs. In this model, there are three types of efficacy beliefs: selfefficacy (e.g., an individual’s own belief in his/her abilities to swim a given distance in under a
certain time), other efficacy (e.g., an individual’s belief in his/her other relay members to swim a

4

given distance in under a certain time), and relation-inferred self-efficacy (RISE) (e.g., an
individual’s perception of other relay member’s beliefs in his/her own abilities to swim a given
distance in under a certain time). Lent and Lopez stated that self-efficacy theory ignores
interpersonal relations and argued that these relations can have an impact on self-efficacy beliefs.
This model was primarily built for groups in which members are in a close relationship, meaning
they have an impact on other group members, as well as being interdependent. Sports groups,
and especially relay teams, do have both these relational aspects, and the tripartite model of
efficacy is likely to be important in relay situations and may impact subsequent self-efficacy and
other-efficacy of relay members.
Feltz, Short & Sullivan (2008) define ability-focused other-efficacy beliefs as an
individual’s beliefs about their teammate’s abilities. However, these beliefs can also be
reciprocal in that the second teammate can also have other efficacy beliefs about the first
teammate as well as additional teammates. That is, in teams greater than two, each teammate
would have beliefs about each of his or her other teammates. An athlete’s other-efficacy beliefs
could, then, be aggregated from his/her separate beliefs of each teammate. Sources of these
beliefs can come from a variety of outside factors and can include perceptions of the teammate’s
accomplishments in past situations, experiences with the athlete, or information conveyed by
third party observers (e.g., coach’s beliefs about the teammate). In turn other-efficacy beliefs can
also influence behavior including the type and amount of effort employed in a joint task or
persistence intentions.
Similarly RISE beliefs, as stated previously, are defined as an individual’s beliefs
regarding how significant others view the individual’s efficacy at a particular task. While these
beliefs are generally a measure of individuals’ reflection on how their partners view them, they

5

may actually be very important in situations in which individuals have limited sources for selfefficacy beliefs. RISE beliefs may further help to predict other group outcomes like persistence
on a task or relationship satisfaction (Lent & Lopez, 2002).
There have been a few studies that have validated Lent and Lopez’s (2002) original
suggestions. One study by Jackson, Knapp, and Beauchamp (2008) investigated the sources of
other-efficacy beliefs. Jackson et al. found that the sources of other-efficacy included one’s
partner’s past performances individually, past performance as a dyad, comparisons with past
partners and other athletes in general, and the partner’s motivation (e.g., desire to succeed,
achieve goals, willingness to work hard, etc.). Sources for RISE included ones’ own self-efficacy
and mastery experiences, past mastery experiences as a dyad as well as dyad experience,
motivation, ones’ partner’s verbal and nonverbal behavior, and physiological factors.
Jackson, Grove, and Beauchamp (2010) studied the relational efficacy beliefs in a
coach-athlete dyad. Here they found that high other-efficacy beliefs predicted enhanced
commitment for both dyad members which supports Lent and Lopez’s (2002) original theory.
However RISE beliefs were seen to exert different effects for coaches and athletes. When
coaches believed that their athletes were confident in their coaching abilities, commitment from
both the coach and athlete increased. However when athletes thought their coach was confident
in their abilities, this led to lowered commitment from both coaches and athletes. While this
finding goes against the proposed theory, the authors speculated that athletes might become
complacent in their relationship, which would result in less commitment and social loafing.
To date, there have not been any studies that have looked at RISE beliefs in relation to
performance measures. However, there have been a handful of recent research studies
concerning performance and other efficacy. The first study conducted by Beauchamp and

6

Whinton (2005) studied self-efficacy between rider and horse in equestrian dyads. Riders
assessed their own self-efficacy as well as their other (i.e., horse’s) efficacy prior to each phase
of competition. Results indicated that other efficacy was related to performance and accounted
for 4% of the variance in performance scores beyond just self-efficacy alone. One of the
limitations of this study was the use of rider-horse dyads where only the human member has
cognitive beliefs, however the findings suggest that other-efficacy beliefs make a significant
contribution to performance and should be investigated further in other partnerships.
One such study used female high-school aged volleyball teams and studied how other
efficacy, collective efficacy, trust, and performance were related over the course of a season
(Dithurbide, 2011). In this study, there were surprisingly no significant results between other
efficacy and performance. One of the main factors as to why this result occurred may have been
due to the team nature of the other efficacy measure. While other efficacy has been demonstrated
to work well in dyad partnerships where it is easy to establish a “me versus partner” approach
where one’s partner’s contribution is different than oneself, this may not be the case in a highly
interdependent team sport like volleyball. In this type of sport, it may be difficult to separate how
an individual contributes directly to the team. In groups larger than dyads, other efficacy may be
better distinguishable if there is a separate and distinct contribution to the performance. This
way, an individual can use past performance as a basis for the other-efficacy beliefs.
One other study (Dunlop et al., 2011) also looked at the relationship between selfefficacy, other efficacy, and performance. They used dyads performing in a dance-based
videogame. Other efficacy was shown to be a significantly related to performance regardless of
self-efficacy levels. So those individuals who believed they had a high-ability partner
outperformed those who believed they had a low-ability partner. The authors postulated that

7

other-efficacy beliefs may be a stronger predictor of performance than self-efficacy beliefs in
dyad performance situations.
Dunlop et al. (2011) believed that those individuals in the high other efficacy condition
may have thought they were more capable of a good performance given their partner’s high
ability and therefore increased effort on subsequent trials. The low other efficacy individuals
may have felt like they did not have a chance due to their partner’s lack of ability and therefore
decreased effort. These findings may help to explain motivation in groups larger than dyads;
specifically groups where performance can be broken up into significant contributions, thus
making it easier for group members to identify each separate member’s contribution.
Swimming relays may be an ideal place to study motivation in group contexts and how
both self-efficacy and other efficacy modify this performance relationship. Relays are technically
an additive task according to Steiner’s (1972) taxonomy of group tasks. However, they function
as a quasi-conjunctive task due to their characteristics (Kerr et al., 2005). Relays are divisible,
meaning that each member of the relay must adequately complete his or her leg of the relay for
the team to have a successful performance. Additionally, relays are sequential, meaning that each
member of the relay must complete his or her leg of the relay before the next leg can be started.
These characteristics allow for this group to function as a quasi-conjunctive group in that any
failure to perform by any group member results in a poor performance for the relay team.
Previous research on motivation in swimming relays has demonstrated that the weakest
members of the relay perform significantly faster in the relay condition compared to the
individual condition (Osborn, Irwin, Skogsberg, & Feltz, 2012). Motivation gains were greater as
the task became more important (i.e., in a scoring relay as opposed to a preliminary race).
Another similar study looked at indispensability in swimming relays and found similar findings

8

(Huffmeier & Hertel, 2011). Those individuals in the later positions of the relay had greater
motivation gains than the individuals in the beginning positions of the relay compared to their
individual times. The authors suggested that indispensability increased as a result of the position
of the relay.
Huffmeier and Hertel’s results were supported by an additional study using archival data
by Huffmeier, Krumm, Kanthak, and Hertel (2012). Here, they used archival data to verify that
one’s serial position in a relay does increase a member’s indispensability. This increase in
indispensability was compared to other explanations of motivation gain based on the differences
between an individual start and a relay start. Swimmers were faster in the relay condition
compared to the individual condition when the swimmer’s performance was instrumental for the
team (i.e., later serial position in the relay) and when the group’s performance was important
(i.e., the relay had a higher chance of winning a medal).
While these previous studies provide support for motivation gains within relays, one
limitation is that there is no subjective information concerning performance. By using real time
relays, subjective measures like an individuals’ perceived exertion, feelings of indispensability or
whether individuals socially compare themselves to teammates can be assessed. One way to
measure this is through ratings of perceived exertion (RPE) using the Borg scale (Borg, 1970).
The Borg scale is a widely used tool to assess subjective perceptions of effort during physical
activity. The Borg scale has been validated against a variety of objective measures (SkaturdMickelson, Benson, Hannon, & Askew, 2011) and has been used in physical activities ranging
from strength training (Buckley & Borg, 2011) to cardiovascular exercise tests (Scherr,
Wolfarth, Christle, Pressler, Wagenpfeil, & Halle, 2013). RPE should correlate positively with

9

performance and this can help to validate whether individuals are accurate in their assessments of
performance.
While these studies demonstrate motivation gains in swimming relays, they were all
conducted with archival data. In order to get a better understanding of the cognitive processes
underlying motivation gains, it is necessary to use real relay teams in a competitive environment.
Purpose of the Study
The purpose of this study was to test the Köhler motivation gain effect in swimming
relays. The pilot study aimed to determine if the Köhler motivation gain effect could be
replicated in real-time relays based on findings from the analysis of archival data. The
dissertation study strived to replicate the previous Köhler effect research with real relay teams
and investigate the potential moderating effects of self- and other efficacy on performance. This
dissertation aimed to demonstrate the importance of other efficacy in groups beyond a dyadic
partnership and the importance both self- and other efficacy may have in regards to motivation.
Further this study aimed to understand the processes of the Köhler effect in groups larger than
dyads.
Contextual Factors
The context chosen for this dissertation was sport and specifically swimming relays.
Swimming relays have been used to show motivation gains in the weakest member of the group
in past research. A competitive setting was chosen for the second study as it is likely that
swimmers would already be highly motivated. Further, these relays allowed for the measurement
of performance times so that individuals could directly see the contribution that both the
individual and other team members made to the relay. Relays also comprise four members which

10

may help individuals have more accurate other-efficacy beliefs concerning their relay meets due
to knowing how exactly an individual contributes to the group’s overall performance.
Hypotheses
1) Self-efficacy and other efficacy will be negatively related to performance times.
2) Swimmers will swim faster times in the relay condition compared to their individual
times.
3) The weakest relay member will demonstrate a motivation gain in the relay compared
to their individual condition.
4) The strongest relay member will demonstrate a motivation loss in the relay compared
to their individual condition.
5) Self-efficacy will moderate the motivation gain relationship so that those with low
self-efficacy will show a motivation gain in the relay condition compared to their
individual condition. Further, high self-efficacy beliefs will result in increased
performance times from the individual to relay condition, resulting in a motivation
loss.
6) Other efficacy will moderate the motivation gain relationship, in that those with
higher other efficacy will demonstrate a motivation gain from individual to the relay
condition.
Research Questions
1) Will there be a performance-change difference from individual to relay performance
depending on the serial position of the relay?
2) Are there differences with respect to gender and the moderating effects of self- and
other efficacy on the Köhler motivation gain effect?

11

3) Will a social compensation effect emerge as a result of low other-efficacy?
4) Will there be a relationship between RISE beliefs and performance?
5) Do relay members indicate that they socially compare themselves to the others in
their team?
6) Do relay members believe their performance will make a substantial contribution to
the relay as a whole? Does this differ according to whether the assessment is before
or after the relay?
7) How much effort do relay members perceive they gave in their performance?
Delimitations
The study was delimited to collegiate level swimmers in swimming relays. The results of
the study may not be applicable to other types of groups in which the task demands are different
than a relay task. Further, it may not be applicable to other sports teams that are more
interdependent and do not have a way to measure an individual’s contribution to a group
performance directly.
Definitions
Motivation Gain: a decrease in swimming split time from the individual condition to the relay
condition.
Motivation Loss: an increase in swimming split time from the individual condition to the relay
condition.
Other Efficacy: an individual’s belief in another’s capabilities to organize and execute the
courses of action necessary to produce certain levels of attainment. In this study, other-efficacy
beliefs are an individual’s beliefs in his/her other relay members’ abilities to swim a given
distance in under a certain time.

12

Relation-Inferred Self-Efficacy (RISE): an individual’s beliefs regarding how significant others
view the individual’s efficacy at a particular task. In this study, an individual’s perceptions of
their other relay member’s beliefs in their own abilities to swim a given distance in under a
certain time.
Self-Efficacy: an individual’s beliefs in his or her capabilities to organize and execute the courses
of action necessary to produce certain levels of attainment. In this study, self-efficacy beliefs are
an individual’s beliefs in his/her abilities to swim a given distance in under a certain split time.
Social Compensation: a decrease in swimming split time from the individual condition to the
relay condition as performed by the strongest member of the group.

13

CHAPTER 2
REVIEW OF LITERATURE
The purpose of this chapter is to provide a review of literature that is relevant to the
variables and procedures in this study. The chapter begins with a review of group motivation
research and theory. Next, a review of the research and empirical literature of the Köhler
motivation gain effect is provided. This is followed by a review of two potential moderating
variables of the Köhler motivation gain effect: self-efficacy and other-efficacy.
Group Motivation Research
There are many different theories of group motivation; however in the majority of these
theories, motivation is based on a cognitive antecedent. An individual’s thoughts determine an
individual’s motivational intensity and direction (Roberts, 1992). Due to the cognitive nature of
motivation, it is difficult to measure motivation directly. However the literature commonly infers
that performance is an accurate measure of motivation. These performance differences are
inferred from effort, which is then inferred to be a result of motivation.
Motivation in groups has been widely studied in the social psychology literature for the
past 30 years. This research has primarily focused on motivation losses. Social loafing is the
tendency for individuals to perform worse in a group situation compared to an individual
situation, and thus is a motivation loss for the group (Everett, Smith, & Williams, 1992; Karau &
Williams, 1993; Williams & Karau, 1991; Williams, Nida, Baca, & Latané, 1989). Baron and
Kerr (2003) provide possible causes for this decrease in performance in a group environment.
Group members (a) may feel less personally identifiable in a group situation and therefore less
subject to evaluation; (b) may recognize that in some instances they may be able to free-ride on
other group members’ efforts; or (c) may reduce their efforts rather than contribute to what they
perceive to be more than their fair share of the collective effort.
14

Social-loafing research has been conducted specifically on ad-hoc relays in sport but has
found contradictory results (Everett et al., 1992; Williams et al., 1989). Williams et al. (1989)
found a social loafing effect in swimming when team members were unidentifiable and therefore
not held accountable for their performance. In this study, performance increased in the relay
condition compared to the individual condition when swimmers were identifiable. Everett et al.
(1992) replicated this study and also studied team cohesion. While they did not find a social
loafing effect, individuals’ performance may not have been affected by varying levels of
identifiability because team cohesion was high. Social loafing may occur when a group
comprises individuals of different abilities. More capable individuals might revise their
performance goals lower to match those of their teammates, resulting in less effort.
Another social loafing study conducted by Miles and Greenberg (1993), using swimming
relays, demonstrated that these social loafing effects can be eliminated if groups are threatened
with punishment. Swimmers were required to swim under a certain performance time that was
determined to be challenging yet achievable. Half of the participants swam as individuals and the
other half swam the task as a group. The threat of punishment was used in the event that
swimmers did not swim under the given performance time. The results demonstrated that the
threat of punishment did indeed attenuate a social loafing effect. Further, follow-up questions
indicated that individuals perceived that they worked harder in groups compared to working
individually. They also expressed more concern for letting down their team members in a group
setting. The results of this study indicate that individuals put forth more effort while in the group
task, and it may be relevant to study performance gains in groups rather than losses. However, in
terms of practical implications, threats of punishment may not a useful way to improve relay
performance on a consistent basis.

15

In terms of motivation gains in groups, two cognitive-based theories of motivation are
social compensation and the Köhler effect. Social compensation occurs when individuals
increase their performance in order to compensate for other anticipated weaker performances of
group members (Karau & Williams, 1997). The Köhler effect occurs when less capable
individuals perform better with others compared to performing the task individually (Baron &
Kerr, 2003).
Social compensation research has generally shown that this motivation gain is usually
demonstrated by the strongest members of the group who would be able to compensate or make
up for weaker group members. In one study, dyad student groups were asked to complete an idea
generating task. Those individuals who had been partnered with a low-ability partner worked
harder to come up with ideas especially when the students were working towards a collective
goal rather than working coactively (Karau & Williams, 1997).
While the social compensation effect has been shown in a variety of studies (Karau &
Williams, 1997; Todd, Seok, Kerr, & Messé, 2006; Williams & Karau, 1991), task importance
and feedback are important components that must be present in order for this effect to occur, but
individual contributions should not be identifiable (Williams & Karau, 1991). In one study by
Williams and Karau, students were asked to complete an idea-generating task. Students who
were given false feedback that the task was not important, underperformed compared to the
students who were given information that the task was highly important. This study
demonstrated that a task must be highly valued in order for social compensation to occur.
Further, the authors postulated that in groups where individualized feedback is not provided,
group members may turn to the overall success of the group to determine their individualized
success. In this case, social compensation may occur as high-ability members may increase their

16

effort in order to make up for a weaker member in order to gain a positive group evaluation.
Thus, although swimming relays in competitive meets are likely to meet the criteria of task
importance and feedback for social compensation effects to occur, individual performance times
are identifiable and may attenuate a social compensation effect.
The Köhler Motivation Gain Effect
In this dissertation, I focused on the Köhler effect because it allows for a focus on the
weakest member of the group, which is of primary interest. The Köhler effect occurs when less
capable individuals perform better in a group context compared to performing the task
individually (Baron & Kerr, 2003). The effect was first discovered by Otto Köhler in the 1920s.
He studied rowing teams and had members perform an arm curling task until exhaustion. Rowers
performed the task both individually and in a dyad condition (rowers used a weighted bar to curl
and were therefore dependent on one another). Köhler found that individuals who were in the
dyad condition persisted longer at the task compared to performing the task individually (Kerr et
al., 2007).
The Köhler effect has been studied in a variety of physical tasks. Some of these physical
tasks have been similar to the original Köhler studies where participants were asked to perform a
motor persistence task. Participants held a metal bar at arm’s length for as long as possible. In
the group condition, partners held the bar over a trip wire, so when one partner stopped, the other
partner had to stop as well (Gockel et al., 2008; Hertel et al., 2000; Kerr et al., 2005). Other
physical tasks have also been used including an air-blowing task (Kerr & Bruun, 1983) as well as
an abdominal plank exercise (Feltz, Irwin, & Kerr, 2011). Most physical tasks that have been
used in these studies minimize coordination between partners so coordination losses do not
interfere with the resulting motivation gain effects (Hertel et al., 2000).

17

The Köhler effect has not been limited to physical tasks but has also been studied in
cognitive computer-mediated tasks where face to face contact is eliminated. Researchers had
used varying tasks like an idea-generating task (Hertel et al., 2003) or a simulated retail game
(Hertel et al., 2008). These computer-mediated tasks have demonstrated the Köhler motivation
gain effect particularly when indispensability of group members is increased. Despite the fact
that computer-mediated work has demonstrated these motivation gains, Hertel et al. (2008)
showed that face to face groups still show larger motivation gains when compared to the
motivation gains of computer-mediated tasks. In this dissertation, the focus is on a face to face
group participating in a physical task.
Köhler motivation gain effects have been seen to be task dependent. Steiner’s (1972) task
taxonomy for groups has provided three task demands that have been studied in the Köhler effect
literature: additive, co-active, and conjunctive. An additive task is one in which the group result
is the sum of the individual member parts. A co-active task demand is one where individuals
perform a task at the same time however their outcomes are independent of one another. A
conjunctive task demand is one where the performance of the group is dependent on the weakest
group member. All three of these task conditions have been used to demonstrate motivation
gains; however, conjunctive task demands result in the largest motivation gain for the weakest
group member (Weber & Hertel, 2007).
Additionally there are two processes by which the Köhler effect is thought to occur:
indispensability and social comparison. The first suggests that it is a group member’s feelings of
indispensability to the group that elicits a motivation gain. In this process, motivation gains
should be the greatest in conjunctive tasks where the group as a whole can only do as well as
their weakest member (Hertel et al., 2008; Kerr et al., 2007). Inferior members in this case

18

should demonstrate the greatest motivation gain because the success of the group is dependent on
their performance. In social comparison, group members either make social comparisons to
stronger individuals in the group, revising their performance goals upwards, or they can make a
successful performance their goal (Kerr et al., 2007). It is likely that both indispensability and
social comparison occur simultaneously in group tasks; however, the strongest motivation effects
have been demonstrated in conjunctive tasks (Kerr et al., 2007; Weber & Hertel, 2007).
There are some boundary conditions concerning the Köhler motivation gain effect
including gender, feedback, and discrepancy in ability between partners, however in this
dissertation the focus was solely on gender, discrepancy in ability, and self-efficacy as
moderators of the Köhler effect. Gender has been found to moderate motivation gains. Women
have shown motivation gains in all three task conditions. Conjunctive task demands have yielded
the greatest motivation gains, with coactive and then additive being increasingly lower. (Kerr et
al., 2007; Weber & Hertel, 2007). One reason why women might show greater motivation gains
in conjunctive tasks is due to the finding that women are more focused on the relational aspects
of groups compared to men. This may cause women’s feelings of indispensability to be
additionally heightened. Men have been shown to exhibit motivation gains when information is
provided about a superior coworker regardless of whether they are in a group task or performing
individually. These gender findings indicate that social comparison may be more salient for men;
whereas, indispensability may be more salient for women (Weber & Hertel, 2007).
Another important consideration is the discrepancy in ability between group members.
The greatest increases in motivation occur when the weaker group member’s performance is
roughly 1.4 times lower than other group members’ (Messé, Hertel, Kerr, Lount, & Park, 2002).
If the discrepancy between the weakest and most capable members is too large, motivation losses

19

(or free-riding) have been found to occur (Hertel et al., 2008). In addition, Feltz et al. (2011)
have found that motivation attenuates as a result of too high or too low of a discrepancy between
partners. Weaker partners may assume that to keep up with, or even beat a stronger partner, may
be unachievable. Additionally, those weaker partners may feel as if their performance will not
significantly contribute to the group’s performance. However, outside of laboratory conditions,
such as in swimming relay performance, discrepancies may be considerably less than 40% and
still be viewed as comparably relevant. In a 200 yard freestyle relay for example, the range of
times within a relay might be 1-3 seconds, which in a 20-something second race, might equate to
a discrepancy of 10-12%.
One additional moderator of the Köhler motivation gain effect is self-efficacy. Seok
(2004) replicated the Köhler effect physical task paradigm by having female dyadic participants
hold a bar over a trip wire. He additionally gave participants false self-efficacy feedback by
showing participants their likelihood of doing well on the next task as based upon their
performance from the first trial. All individuals in the study demonstrated a motivation gain in a
conjunctive task condition where the participant was the weakest group member. However, this
was moderated by self-efficacy in that those individuals in the low self-efficacy conditions
demonstrated the highest motivation gains particularly when partnered with a moderately more
capable partner.
Seok (2004) provided some rationales for these findings. The first was that the high selfefficacy individuals may have been overconfident after their original performance so they may
not have increased their effort on subsequent performances. Another reason may have been that
the high self-efficacy individuals may not have had to prove themselves to their partners or the
experimenter, whereas the low self-efficacy individuals may have viewed subsequent

20

performances as a second chance, therefore initiating a self-presentation effect for those weaker
members. This rationale was supported statistically in that the low self-efficacy individuals
perceived it to be significantly more important to perform well in the last trial compared to the
first two trials. Additionally the participants were all female and according to other Köhler effect
literature (Weber & Hertel, 2007), indispensability may be more salient for women in these types
of tasks.
While the Köhler effect has been studied in many physical tasks, it has only recently been
studied in sports teams. Hüffmeier & Hertel (2011) looked at the indispensability of group
members in swimming relays. They used Olympic archival data of these relays to compare
individual times that were swum at the Olympic meet to those times swum in the relay condition.
They hypothesized that swimmers would swim faster in the relay condition compared to their
individual swims due to more intergroup competition compared to inter-individual competition
in the individual condition. Their analysis confirmed this hypothesis and swimmers did swim
faster in the relay condition compared to the individual condition. Further, they serial ordered the
relay (first, second, third, and fourth) and examined how motivation changed as a result of one’s
position on the relay. They found that those individuals who swam in the latter positions of the
relay, specifically those who swam last, had greater motivation gains than swimmers in the
earlier legs of the relay. The authors credited these findings to a greater sense of indispensability
in latter relay positions.
Huffmeier and Hertel’s results were supported by an additional study using archival data
by Huffmeier, Krumm, Kanthak, and Hertel (2012). Here, they used archival data to verify that
one’s serial position in a relay does increase a member’s indispensability. This increase in
indispensability was compared to other explanations of motivation gains based on the differences

21

between an individual start and a relay start. Swimmers were faster in the relay condition
compared to the individual condition when the swimmer’s performance was instrumental for the
team (i.e., later serial position in the relay) and when the group’s performance was important (i.e.
the relay had a higher chance of winning a medal).
Osborn, et al. (2012) also looked at archival data of National Collegiate Athletic
Association (NCAA) swimming relays. In this study, they compared an individual’s fastest
individual race time during the season to their relay split time at the NCAA championship meet.
The authors found similar results to Hüffmeier & Hertel (2011) that individuals swam faster in
the relay conditions than they did in the individual conditions. Further, the findings in Osborn et
al. (2012) indicated that these motivation gains increased as the task became more important, for
example in a scoring relay situation compared to a preliminary race.
In Osborn et al. (2012), gender also modified this task importance relationship with
females showing motivation gains when the task was less important (i.e. the preliminary race)
compared to men who only demonstrated this motivation gain during the scoring relay race.
These gender findings support previous Köhler effect research (Weber & Hertel, 2007), which
indicates that indispensability may be more salient for women where women show motivation
gains in less important task settings. While men may be more sensitive to social comparison and
may only demonstrate motivation gains when the task is very important and they can positively
compare themselves to stronger others.
One other important finding from the Osborn et al. (2012) study was that they rank
ordered the individuals in the relay according to ability. Those who were ranked first on the relay
were the fastest in the individual event. Similarly, those who were ranked fourth, were the
slowest individuals in the relay with the second and third ranked individuals in the middle with

22

respective speeds. The researchers found that Köhler motivation gains were greatest in the fourth
ranked member. This member swam significantly faster in the relay condition compared to the
individual condition. Further there was a non-significant social loafing trend in the first ranked
member, where the first ranked member swam slightly slower in the relay condition compared to
the individual condition.
Osborn et al. (2012) then compared the rank order effects (based on time) to those of
st

nd

rd

th

Hüffmeier & Hertel’s (2011) serial ranks, based on relay position (1 , 2 , 3 , and 4 ). Osborn
et al. (2012) found that similar to Hüffmeier & Hertel (2012), those individuals who swam in the
rd

th

later positions of the relay (3 and 4 ) had greater motivation gains than those individuals who
swam first on the relay. However there were an equal percentage of individuals who were ranked
rd

th

according to their time 3 and 4 who swam in the last position of the relay. These results
indicated that rank ordering individuals based on time might provided stronger motivation gain
effects compared to serial ordering individuals based on relay position.
Sports teams may be an ideal place to examine the Köhler effect given that motivation
should be sufficiently high in a competitive setting. Similarly, sports teams have set groups that
should be accustomed to both practicing and competing with one another and therefore should be
familiar with other teammate’s abilities. Sports settings also provide feedback as to how both
individuals and the team are doing, so this feedback can be used to evaluate a group’s
performance.
Relays, in particular, are one type of additive task. However, Kerr et al. (2005) specified
that relays have certain characteristics that allow them to function more like quasi-conjunctive
tasks. Relays are divisible in that each member of the relay must adequately complete his or her

23

leg of the relay for the team to be successful overall. Similarly relays are sequential. Each
member of the relay must adequately complete his or her leg of the relay before the next leg can
be started. Hertel et al. (2003) found motivation gains in a divisible computer-mediated cognitive
task in which additive task groups had slightly better performance than those groups performing
under individual conditions. These studies suggest that additive tasks when performed divisibly
may increase identifiability which may lead to greater motivation games. Additionally, when an
additive task is performed sequentially (and one cannot start until the previous teammate
finishes), this may add an extra layer of indispensability.
Self-Efficacy and Performance
Self-efficacy is defined as one’s beliefs in his or her abilities to succeed in a certain
situation or task (Bandura, 1997). In the sports field, self-efficacy has commonly been examined
in accordance with performance and particularly performance over time (Feltz, 1982, 1988;
Feltz, Landers, & Reader, 1979; Feltz & Mungo, 1983). There has been a variety of
methodologies, measurements, sports and other motor tasks used in self-efficacy research. A
meta-analysis (Moritz et al., 2000) has demonstrated that there is a moderate correlation between
self-efficacy and performance. Generally this performance is strengthened when measurements
between the self-efficacy belief and the performance measure are concordant and also when the
measures are taken closer together. Self-efficacy and performance has also demonstrated a
reciprocal relationship. Self-efficacy is seen to enhance performance, while successful
performance is seen to increase one’s self-efficacy beliefs (Feltz, 1982; McAuley, 1985).
There have been some studies that have found a negative relationship between selfefficacy and performance. Vancouver and colleagues (Vancouver, Thompson, Tischner, &
Putka, 2002) argued that a better performance would increase self-efficacy, which would in turn

24

lead to overconfidence and a subsequent performance decrease. Low self-efficacy, in turn, would
lead to more practice at the task and increased effort, which would help performances in the
future. Vancouver and his colleagues did indeed find that self-efficacy was negatively related to
performance in a within-person analysis over time but positively related to performance at the
between-person level (Vancouver, Thompson, & Williams, 2001; Vancouver et al., 2002;
Vancouver & Kendall, 2006). However, these original studies were conducted on ambiguous
cognitive tasks, using either computer game outcomes or academic performance as the
performance measure, and where participants had little incentive to act on their efficacy beliefs
(Feltz et al., 2008).
In recent research in sport, there have been multiple studies that have examined selfefficacy’s positive or negative relationship with performance. In one study, Beattie, Lief,
Adamoulas, & Oliver (2011) used a golf putting task and found that overall there was a
significant and positive correlation between self-efficacy and performance. But, self-efficacy had
a slight negative, yet non-significant, effect on subsequent performance. Other research by
Gilson, Chow, and Feltz (2012) found that using a squat-lifting task, self-efficacy was positively
related to performance at both the between-person and within-person levels even while
controlling for the participants’ past performance. This study lends support for the positive
relationship between self-efficacy and performance in sports tasks. Overall, the literature
indicates that self-efficacy demonstrates a strong, positive relationship with sport performance.
Other Efficacy
Lent and Lopez (2002) proposed a tripartite model of efficacy beliefs, which includes
self-efficacy, other efficacy, and relation-inferred self-efficacy (RISE). The rationale for this
model was that self-efficacy theory commonly ignores interpersonal relations and the effects of

25

these relationships on self-efficacy. In this model, relationships are close because those
individuals have a mutual impact on each other. Further, these relationships are high in
interdependence. Sports teams are a good place to study these interpersonal relations because
they are an interdependent group. Swimming is a sport where most of the competition takes
place at an individual level. However, relays are an integral part of a swimming competition and
are both interdependent and offer a starting point to study the interpersonal relations among
teammates. In social cognitive theory, the social environment has a large impact on efficacy
beliefs. Teammates and particularly relay members may hold beliefs about themselves or their
other relay members, which in turn constitute the social environment and may therefore have an
effect on efficacy beliefs.
Lent and Lopez (2002) define other-efficacy beliefs as an individual’s beliefs about
his/her partner’s abilities to perform a specific behavior. However, this belief is also reciprocal in
that the second partner will also have other-efficacy beliefs concerning the first partner as well.
Sources of these beliefs can come from a variety of outside factors and include perceptions of the
partner’s accomplishments in past situations, experiences with the partner, or information
conveyed by third party observers (e.g., coach’s beliefs about the partner). In turn, other-efficacy
beliefs can also influence behavior including the type and amount of effort employed in a joint
task or persistence intentions.
Lent and Lopez (2002) also define RISE beliefs as an individual’s beliefs regarding how
significant others view the individual’s efficacy at a particular task. This is a generally a measure
of an individual’s reflection on how their partner sees them. RISE beliefs may not be similar to
the individual’s own self-efficacy beliefs and may help determine other group outcomes like
persistence and relationship satisfaction. RISE may function as a filter through which group

26

behavior is interpreted. Feedback, either positive or negative, may be interpreted differently
depending on an individual’s RISE beliefs. For example, a comment about how well the group
can do may be interpreted as supportive if an individual’s RISE beliefs are high. However the
same comment may be interpreted as shallow and insincere if RISE beliefs are low. These
interpretations of RISE may affect subsequent group behavior. Lent and Lopez postulate that
RISE beliefs are most influential when individuals have other limited sources for self-efficacy
beliefs.
Jackson, Beauchamp, and Knapp (2007) did not study efficacy and performance directly,
but rather looked at the relationships between the three aspects of the tripartite model of efficacy.
The authors used tennis doubles to study the relationships between the three types of efficacy
beliefs and both satisfaction and commitment to their partnership. They also looked at actor and
partner effects within this relationship. Actor effects are when the predictor and the outcome
variable occur within an individual (e.g., self-efficacy influences an individual’s own
performance), while partner effects are when the predictor variable within one person influences
the outcome variable in the partner (e.g. other-efficacy beliefs that individuals hold about their
partners affect how their partners perform). The three forms of efficacy that one partner holds
may influence both the motives and behaviors not only toward their partner but also the motives
and behaviors of that partner.
Results of Jackson et al. (2007) showed that both self- and other efficacy were both
significantly correlated with satisfaction and commitment. There were also significant actor
effects for RISE and other efficacy in relation to self-efficacy. The more that individuals
believed in their partners the greater were the self-efficacy beliefs for those individuals
respectively. There were no partner effects found in relation to the different aspects of efficacy;

27

however, there was a partner effect between self-efficacy and commitment. These results
indicate that as self-efficacy increases in an individual, his or her partner should feel increased
commitment to the relationship. This study demonstrates that relational beliefs may have an
impact on both the individual as well as other group members.
Jackson and colleagues (Jackson et al., 2008) furthered the literature concerning the
tripartite model of efficacy beliefs by conducting qualitative surveys of elite international-level
athlete dyads. Here they wanted to evaluate the antecedents and consequences of self-efficacy,
other efficacy, and RISE as a way to further support Lent and Lopez’s (2002) original model.
Results indicated that there were four higher-order themes across the efficacy measures: oneself,
one’s partner, the dyad, and external factors. Perceptions regarding oneself were most frequently
cited as sources for self-efficacy. These included past performance mastery achievements, sport
experience, and pre-competition preparation. Verbal persuasion as well as other-efficacy and
RISE beliefs also were found to be sources of self-efficacy. Sources of other efficacy included
the partner’s past performances individually, past performance as a dyad, comparisons with past
partners and with other athletes in general, and the partner’s motivation (e.g. desire to succeed,
achieve goals, willingness to work hard, etc.). Sources for RISE included ones’ own self-efficacy
and mastery experiences, past mastery experiences as a dyad as well as dyad experience,
motivation, ones’ partner’s verbal and nonverbal behavior, and physiological factors.
There was a distinction between intrapersonal and interpersonal concepts with respect to
the consequences of each efficacy belief (Jackson et al., 2008). Self-efficacy consequences
included improved individual performance, greater effort and motivation, greater ability to
concentrate on the task as well as improved affect. Other efficacy outcomes included greater
responsiveness to the partner, more open and positive verbal behavior towards the partner, and

28

increased satisfaction in the relationship. However some athletes thought that higher levels of
other efficacy could lead to negative affective responses and the possible breakdowns of
relationships. The consequences of RISE included enhanced self-efficacy beliefs, increased
motivation in a relational context, greater relationship persistence intentions and overall higher
relationship satisfaction. These results provide support for the existence of these relational
efficacy sources and consequences and align with those originally proposed by Lent and Lopez’s
(2002) model.
One specific study that supports the tripartite model and these relationships between self,
other-efficacy and RISE beliefs was done by Jackson et al. (2010). They studied relational
efficacy beliefs in a coach-athlete dyad. Here they found that high other-efficacy beliefs
predicted enhanced commitment for both dyad members, which supports Lent and Lopez’s
(2002) original theory. RISE beliefs were seen to exert different effects for coaches and athletes.
When coaches believed that their athletes were confident in their coaching abilities, commitment
from both the coach and athlete increased. However when athletes thought their coach was
confident in their abilities, this led to lowered commitment from both coaches and athletes. The
authors explanation for these findings were that athletes may receive behavioral feedback that
reinforces their own self-efficacy beliefs which may lead to athletes feeling confident in their
abilities to compete at higher levels and therefore feeling less commitment to this specific coach.
This study demonstrates that the tripartite model has a practical application for commitment in
groups. However, these results may not be replicated in other studies in which the partnership or
group is not based upon a hierarchy as with a coach-athlete relationship.
The research concerning other efficacy in sport has been limited and has consisted of a
few recent studies that primarily focus on the relationship between other efficacy and

29

performance. Beauchamp & Whinton (2005) studied self- and other efficacy between rider and
horse in equestrian dyads. Participants were recruited from a 1-day competition where they
competed in three different equestrian events (dressage, cross-country, and show-jumping).
Before each event, self and other (horse) efficacy were evaluated by the participants. The
authors wanted to look specifically at the relationship between self-efficacy, other efficacy, and
performance. They hypothesized that other efficacy would be able to explain significant
variation in performance beyond self-efficacy alone. The authors’ hypothesis was supported in
the dressage category where other-efficacy explained 4% more of the variance in performance
than just self-efficacy alone. However, there were no significant effects found in the other two
events of cross-country or show-jumping. The authors mentioned this result could be due to the
minimal variation in the scores within these two events. While the authors address the limitations
of using a human-horse dyad, the findings do suggest that other efficacy does make a significant
contribution towards performance beyond that of self-efficacy alone.
Two more recent studies have also studied this relationship between other efficacy and
performance. Dithurbide (2011) looked at high-school aged women’s volleyball teams and the
relationship between other efficacy, collective efficacy, trust, and performance over the course of
a club volleyball season. However the results indicated that other efficacy did not have a
relationship with performance for the season. The author provided a few rationales for why there
was no effect found. The first was that these athletes were high school aged and the other studies
concerning other efficacy have been conducted with adults. Trust and other efficacy were highly
correlated in this study and therefore these participants may have had difficulty distinguishing
between the two constructs; whereas adults may better understand the nuances between the two
concepts. Another reason may have been the nature of this team task. To date, other efficacy has

30

been exclusively studied in dyads. Dithurbide postulated that it may be easier to establish a sense
of “me and other” at the beginning of the season when individuals are first creating a team.
Further, it may be more difficult as the season goes on to distinguish between the self and others
in a highly interdependent team sport and establish the necessary feeling of “other” for otherefficacy to be sufficiently studied compared to a dyadic relationship.
One other study also studied dyadic relationships and the relationship between self-,
other-efficacy, and performance (Dunlop et al., 2011). Female dyads performed a series of
dance-based videogames for practice and were given false efficacy feedback concerning their
coordination abilities. Dyads performed one final dance trial after this feedback was given and
performance was measured compared to the performance on the trials. Results indicated a main
effect for other-efficacy, where those in the high other-efficacy condition outperformed those in
the low other-efficacy condition, regardless of their own self-efficacy beliefs. However, there
was no effect for self-efficacy on this task. These results suggest that the effects of other-efficacy
may be stronger than those of self-efficacy in relation to performance.
Dunlop et al. (2011) explained these findings through Karau and William’s (1993)
collective effort model. The model proposes that the level of engagement one portrays in an
interdependent activity is determined by both personal effort and the possibility of the attainment
of the desired outcome. When an individual perceives that he or she can contribute to the
outcome, he or she becomes engaged in the task. However, if that individual perceives that his or
her actions will not sufficiently affect the outcome, disengagement occurs. In the Dunlop et al.
study, those partners in the high other-efficacy condition may have thought a successful
performance was more attainable given their beliefs in the abilities of their partner and therefore
increased performance. Similarly, those in the low other-efficacy condition may have thought

31

that a successful performance was unattainable given their partner’s low abilities. While this may
be a reasonable explanation, the performance measure in this study was based upon the success
of the dyad as a whole rather than the individuals making a unique contribution to the group.
Research has yet to demonstrate whether a unique contribution has an effect on the relationship
between other-efficacy and performance.
Summary
Both Köhler’s experimental work and subsequent research have demonstrated motivation
gains in physical tasks where weaker individuals perform better on tasks when paired with a
moderately more capable partner (Weber & Hertel, 2007). The Köhler effect has two separate
processes, social comparison and indispensability. When partnered with a more capable partner,
individuals may revise their performance goals upward (social comparison) or individuals may
feel that their effort is highly instrumental for the team to be successful overall (indispensability).
While the Köhler effect has been strongest in conjunctive task demands, relays function as a
quasi-conjunctive demand in which the relay is both divisible and sequential, so each relay
member must perform adequately on his or her own performance in order for the team to do well
overall.
Only one study to date has looked at the relationship between the Köhler motivation gain
effect and self-efficacy. Seok (2004) found that those in a low-efficacy condition demonstrated
the largest motivation gain in a conjunctive task demand. Meanwhile, research in sport has
demonstrated a strong, reciprocal relationship between self-efficacy and performance (Feltz,
1982; McAuley, 1985). However, research on efficacy beliefs has indicated that other-efficacy or
one’s beliefs about their partner’s abilities to perform a specific behavior, may also be important
to performance (Dunlop et al., 2011; Lent & Lopez, 2002). With limited information on sport

32

performance and the Köhler effect, it is necessary to look at how both self-efficacy and otherefficacy may moderate the Köhler motivation gain effect in real sports teams. Further, this
research may lead to more information regarding motivation in sports teams.

33

CHAPTER 3
METHOD
The purpose of this dissertation study was to replicate the previous Köhler effect research
with real, intact relay teams and investigate the potential moderating effects of self- and other
efficacy on performance. First, a pilot study was conducted to determine if the Köhler motivation
gain effect could be replicated in real-time relays based on findings from the analysis of archival
data. Based on results from the pilot data, a full-scale study was designed to test the hypotheses
put forth in Chapter 1.
PILOT STUDY
Participants
Participants in the pilot study were 44 high-school aged swimmers (28 women, 16 men)
recruited from three large swim teams in the state of Wisconsin. In order to participate in the
study, participants had to be high school age or older. Those individuals who were starting high
school at the end of the summer were also invited to participate in the study. The average age of
study participants was 15.9 years (SD = 1.5). Swimmers’ average experience participating in
swimming was 7.3 years (SD = 2.7). And, swimmers swam on their respective teams for an
average of 5.6 years (SD = 3.1). Participants were organized into 11 same-sex relay teams each
containing four members.
Measures
Demographics. A demographics questionnaire was used to obtain background
information from the participants. Items included age, years of competitive swimming
experience, years swimming on current team, long or short course swimming preference, top
three events, personal best times for those best events, personal best times in the 100 freestyle,

34

and the participant’s personal and team goals for the season. The questionnaire also asked about
how much participants enjoyed being on relays, in general, as well as how important it was to
each participant to be on the 400 yard freestyle relay at his or her championship meet at the end
of the season. The demographics questionnaire also asked if participants ever swam on a
freestyle championship relay before and if so, what their favorite position is and why.
Descriptive statistics for the swimmers are presented in Table 1.
Motivation Gain. Individual times of the participants were collected at two points in the
season. These were then averaged to get a most consistent measure of what the participants swim
time was throughout the season. These averaged individual times were used to place participants
into relays. Relays were created so that the final times of the relays (i.e., sum of individual times)
would be similar to one another, and therefore, more competitive as each relay team would have
an equal chance of winning. Participants swam an individual and relay condition swim at one
time toward the end of the season. These times, respectively, were compared to determine if
motivation gains occurred.
Procedure
Participants were recruited from among three large swim teams in the State of Wisconsin.
Both parental consent and participant assent were obtained from participants. Swimmers swam
two baseline 100 yard race-pace swims on two separate practice days. The times were hand
timed by coaches and were recorded by the researcher. Participants’ times were averaged
together to make a more consistent measure of how these swimmers usually swim in practice.
The averaged swim times were then used to place the swimmers into ad-hoc relay groups
for a third practice. At the third practice, swimmers swam both an individual 100-yard race and
400-yard freestyle relay (relay condition of the event). Times were collected on both of the

35

swims and were once again hand timed by the coaches and recorded by the researcher.
Participants had about 15 min. for rest between each swim. In swimming, it is common practice
for coaches to conduct timed swims during a practice setting. During most of these race-pace
efforts, there is less time to rest in between swimming repeats and 15 min. was deemed an
appropriate time for swimmers to effectively “warm down” by doing easy swimming in between
the individual and the relay swim.
Standardization of Swimming Times
In swimming, there are differences between a regular start and a relay exchange start that
greatly affect one’s time in leaving the block. A relay exchange is usually much faster,
sometimes being as low as 0.2 to 0.3 s while a in a regular start, a good reaction time can be
around 0.6 to 0.7. These differences account for the fact that swimmers in a relay can see when
they should leave the block as they wait for the swimmer in the water to finish, leaving them
with a “reaction time” in which they must anticipate when to start. These differences can lead to
relay swim split times to be slightly faster due to the nature of the start. In order to correct for
these faster times, the reaction time can be added to the split time in order to obtain a
standardized split time (Osborn et al., 2012). An example of this would be the average reaction
time of 0.3 s added to a split time of 58.42. This would result in the standardized split time of
58.72 s.
In this pilot study, due to the fact that these times were collected in practice, relay
exchange technology was not available and was, therefore, not used. In order to standardize these
split times, the average reaction time from the archival data set used in Osborn et al. (2012) was
used. This split time was 0.3 and was added to each individual’s time who used a relay exchange
start in order to obtain the standardized split times for the rest of the data analysis.

36

Results
A repeated measures ANOVA was used to determine performance differences between
individual and relay swims. Ranking of the individuals in the relay groups was determined by
their life-time best swim performances. Individuals with the fastest life-time best performance in
st

the relay were ranked 1 , while individuals with the slowest life-time best performance were
th

ranked 4 . The 2

nd

rd

and 3 place rankings were also given respectively by life-time best time, so

that within one relay group, there would be one person with each rank status. Life-time best
performance times were used as these may be a more consistent measure of performance. Other
research within swimming has found that faster swimmers are more consistent in the same event
between competitions compared to slower swimmers (Stewart & Hopkins, 2000). Due to this
finding, slower swimmers may have been less consistent on the 2 days in practice that times
were collected for the pilot study, which may have resulted in a skewed individual average of
performance. By basing individual performance off life-time best times, this kept individual
performance more consistent across both fast and slow swimmers.
In order to check the rankings that were given to relay members, a manipulation check
was performed so that swimmers who were ranked higher should swim faster overall than lower
ranked swimmers A 4 x 2 (Rank x Condition: individual, relay) analysis of variance (ANOVA)
with repeated measures on the last factor was conducted, using swim performance as the
2

dependent variable. There was a significant rank main effect, F(3, 40) = 2.85, p = .049, ηp =
0.176. A Tukey post hoc analysis was used to perform pair-wise comparisons between the
ranked members. This analysis showed that the fourth ranked members performed worse (M =

37

71.27, SD = 7.45) than the first ranked members (M = 65.57, SD = 3.00). Thus the ranking
manipulation was successful. There was no significant main effect for condition.
In the 4 x 2 repeated measures ANOVA there was a marginal Rank x Condition
2

interaction, F(3,40) = 2.32, p = .09, ηp = .148. This trend was investigated further by
conducting a 2 x 2 (Rank x Condition: individual, relay) ANOVA with repeated measures on the
last measure. There was a significant Rank x Condition interaction, F(1, 20) = 5.64, p = .02, ηp

2

= .220. According to 95% confidence intervals, the second ranked individual performed better in
the relay condition, M = 66.49, CI [64.20, 68.77] than in the individual condition, M =68.50, CI
[66.07, 71.01], although this was not significant. Means for the fourth ranked swimmers
individual condition, M = 71.39, CI [68.09, 74.69] compared to the relay condition, M = 71.15,
CI [68.27, 74.03] demonstrated no change. There was also no change for the first ranked
members’ individual performance, M = 65.57, CI [62.27, 68.86] compared to their relay
performance, M = 65.45, CI [62.57, 68.33]. Swim performance time data is presented in Table 2.
Discussion
The manipulation check results indicated that the rank ordering used to place individuals
into relays was accurate. Individuals who swam faster in practice were also the faster members
of the relay team.
While the slowest individual on the relay did not demonstrate the greatest motivation
gain, interesting trends emerged. Second ranked swimmers were seen to perform almost 2 s
better in the relay performance compared to their individual performance. The calculated power
was only .27, thus there was not much power to find a significant effect with only 11 swimmers
at each rank. While a power analysis was performed before the pilot study was conducted, the
study was limited by both the number of swimmers who showed up to practice on the day of the
38

timed relay event as well as swimmers who were still participating in practice in the late part of
the season.
However, while not statistically significant, 2 s can make a large practical impact on the
final time in a swimming relay. Further, this trend supports previous research concerning the
Köhler effect. Previous research has shown that the Köhler effect is more pronounced when one
is partnered with a moderately more capable partner (Messé et al., 2002). In these practice relays,
the range of times was quite large (i.e., 24 s). The first ranked individual, though most likely
moderately more capable than the second individual was much more capable than the third or
fourth ranked members. This may be why the second ranked member was the only person to
demonstrate a trend toward performance increase during the relay, because these individuals
were the only ones who may have been able to successfully compare themselves to the first
ranked individual or to think they could make a significant contribution to the relay performance.
The first ranked individuals’ performance was also not significantly worse in the relay condition
compared to the individual condition. Again, the power was low to find any significant effect.
The pilot study provided several suggestions for the dissertation study. First, in order for
the Köhler motivation gain effect to occur where the least capable individual demonstrates a
greater motivation gain in the group task, relay members should be closer in ability. The range
between the first and fourth ranked member in the pilot study was much larger than may be
typical for relays in a competitive setting. Performance differences may occur if participants in
the relay are closer in ability and feel as if they can (a) socially compare themselves to the others
in the group and (b) feel like their performance will make a substantial contribution to the relay
as a whole. In this study, relay members’ times may have been too far apart for them to
realistically feel that they could have an impact on the relay’s overall result.

39

Further, while these pilot study results indicate that there are performance differences in a
practice environment, a competitive environment may be a stronger place to study these
motivation effects as swimmers will already be highly motivated. Lastly, a larger sample size
and more relay teams are recommended to have greater power to detect differences. A power
analysis following f index recommendations suggests that a sample size of 140 is needed to
observe a moderate (f=.30) Köhler effect with probability > .80.
DISSERTATION
Participants
Six Division II and III NCAA swimming teams were recruited to participate in the study.
Of the six teams, five consisted of both a men and women’s team, while the last team comprised
only women. Participants were 199 NCAA swimmers (111 females, 88 males). In terms of
academic grade in college, 66 = freshman, 34 = sophomore, 37 = junior, 30 = senior. The other
32 participants did not indicate a grade. Swimmers ranged in age from 18 to 22 years (M = 19.3,
SD = 1.2) and had been swimming competitively from 1 to 18 years (M = 10.2, SD = 3.35). In
order for participants to be included in the study, they had to participate in at least one or more
freestyle relay events (200, 400, or 800 yards) at their fall invitation meet during the 2012-2013
season. See Table 3 for descriptive statistics.
Measures
Demographics. A demographics questionnaire was used to obtain background
information from the participants. Items included age, year in school (freshman, sophomore,
junior, senior, fifth year), years of competitive swimming experience, top three events, personal
best times for those best events, personal best times for the 50, 100, and 200 yard freestyle, and
the participant’s personal and team goals for the season. The questionnaire also asked how much
participants enjoyed being on relays in general as well as how important it was to each
40

participant to be on the 200, 400, and 800 yard freestyle relays at the conference championship
meet at the end of the season. See Appendix A for Demographics questionnaires.
Self-Efficacy. The self-efficacy measure comprised nine items. It asked participants how
confident they were in their ability to swim 50, 100, or 200 yards in under a given time in
seconds (e.g. for females: 50: 28-24, 100: 66-50, 200: 126-110; males: 50: 25-21, 100: 59-43,
200: 117-101). The self-efficacy measure specifically asked how confident participants were in
their ability to swim a given distance under a certain time at the fall invitational meet during the
season. Items were assessed on an 11-point scale, ranging from 0 (not at all confident) to 10
(totally confident). Self-efficacy was assessed just prior to a team’s fall invitational meet. See
Appendix for self-efficacy questionnaires at all distances for each gender.
Other Efficacy. The other-efficacy measure also comprised nine items and was used for
individuals who swam on a relay at the fall invitational meet. The other-efficacy measure asked
each participant how confident he or she was in the abilities of his or her relay team’s other
members to swim a given distance in under a certain time. The measure used the same times as
used in the self-efficacy questionnaire. The measure concerns an individual’s belief about his/her
relay team as a whole team, rather than each other member separately. Other-efficacy measures
were collected before relays during each team’s fall invitational meet. Items were assessed on an
11 point scale, with 0 (not at all confident) to 10 (totally confident). See Appendix for other
efficacy questionnaires at all distances for each gender.
RISE. RISE also comprised nine items and was used for individuals who swam on a
relay at a fall invitational meet. The RISE questionnaire asked each participant how confident
he/she thought his/her relay members felt about that participant’s own ability to swim a given
distance under a certain time. Once again these were the same times as used in both the self- and

41

other efficacy questionnaires. This measure concerns the individual’s belief about his/her relay
teams’ belief as a whole team rather than each member’s belief in the individual separately.
RISE measures were collected before relays during the team’s fall invitational meets. Items were
assessed on an 11-point scale, ranging from 0 (not at all confident) to 10 (totally confident). See
Appendix for RISE questionnaires at all distances for each gender.
RPE. Perceived exertion was measured using the Borg RPE scale (Borg, 1998). The
scale ranges from 6 - 20 where 6 means “no exertion at all” and 20 means “maximal exertion.”
Participants were asked to rate their total feeling of exertion after their relay performance. See
Appendix AA for Post Relay Questionnaire.
Motivation Gain. Both an individual and relay measure of performance is needed to
indicate performance gains or losses. Relay times of the participants were collected at each
team’s fall invitational meet for the season. The individual time used, was the individual’s lifetime best individual performance time in that event. Individual life-time best times were used
because not all individuals competed in both the individual and relay freestyle events at these
meets. The use of life-time best performances allowed for the individual measures to be kept
consistent across the participants. The individual life-time best time was used to rank participants
from fastest to slowest. If there was a tie between individual times (i.e., two or more swimmers
swam the exact same time), times were verified from the online database
www.collegeswimming.org as well as self-reported individual best times. Any differences
between these two times were then used to rank participants accordingly. Individual and relay
times were then compared to determine if motivation gains occurred.
Procedure

42

Once IRB approval was obtained, six NCAA Division II and III coaches were contacted
about their participation in the study. Once coaches consented to participate, data collection was
arranged to take place at each school’s fall invitational meet during November and December.
During fall invitational meets, swimmers are often rested in order to get more experience
swimming races that they might swim in the conference championship at the end of the season.
Fall invitational meets function as trial runs prior to the end of the season where most teams will
rest in order to get some experience swimming fast and in prelims/finals format similar to their
conference championship meet. However, these fall invitational meets may not be as important
as the conference championship meet and therefore motivation in relays may still be lower than
at the end of the season. Swimmers were informed about the study prior to the first day of the
meet. If participants consented to participate, they completed a demographics questionnaire
along with a baseline self-efficacy survey concerning the 50, 100, and 200-yard freestyle events
prior to the start of the fall invitational meet.
A total of 199 swimmers participated in the study and composed 91 4-person swim
relays. When divided by gender, there were 54 female relay teams and 37 male teams. While
swimmers could only participate in one relay event in each length, some swimmers swam in
multiple relays with varying lengths (i.e., a swimmer could swim in the 200, 400, and 800
freestyle relay, but could not participate in the 200 freestyle relay multiple times). Swimmers
were repeated across relays in the following manner; 68 swimmers swam in one relay event, 98
swam in two relay events, and 33 swam in all three relay events. Throughout the meet and prior
to any relay swim, participants completed questionnaires concerning their self-efficacy beliefs
about their own swim, other-efficacy beliefs and their RISE beliefs of their fellow relay
members. They also answered a pre-relay question asking how indispensible they felt their

43

contribution would be to the relay. See Appendix Z for Pre-Relay Questionnaire. After the relay
swim, they answered a post-relay questionnaire which asked questions concerning social
comparison, their contribution to the team, as well as RPE using the Borg scale.
The pre-competition questionnaires were administered on deck prior to warm-ups. This
provided a time close enough to the meet to accurately reflect efficacy beliefs but did not
interfere with competition routines. Post-relay surveys were administered either on-deck after the
meet session had concluded or during team lunch or dinner meals immediately after the
conclusion of the swimming session. The researcher distributed and collected the surveys during
these times. Individual times for each participant were collected from
www.collegeswimming.org and relay times were collected from the results of the fall invitational
meet of each team.
Treatment of the Data
Prior to data analysis, individuals were ranked within their relay groups based upon their
life-time best individual time. Individuals with the fastest life-time best performance were ranked
st

th

1 , while individuals with the slowest life-time best performance were ranked 4 . The 2
rd

3

nd

and

place rankings were also given respectively by life-time best time, so that within 1 relay

group, there would be one person with each rank status. Life-time best performance times were
used as these may be a more consistent measure of performance across both slow and fast
swimmers. Relay splits were standardized by adding the reaction time (i.e., time between when
the swimmer in the water touches the wall, and starting swimmer’s feet leave the block) to the
relay time of each swimmer as was performed in the pilot study as well as Osborn et al. (2012).
An example of this would be the average reaction time of 0.3 s added to a split time of 58.42.
This would result in the standardized split time of 58.72 s. In one meet in particular, the reaction
44

times were not accurately measured for all the relays, which can occur when the timing system is
not working correctly. Some of the reaction times from this meet were missing, so the average
reaction time from Osborn et al. (2012) was used in place of these missing reaction times to
create the standardized relay times.
Descriptive statistics (i.e. means, standard deviations) were calculated for all variables.
Bivariate correlations were calculated between self-efficacy, other efficacy, and RISE beliefs as
well as between efficacy scores (self-efficacy, other-efficacy, and RISE) and performance
(Hypothesis 3). High and low self- and other efficacy were used as predictors in the analyses.
The average of the raw efficacy scores was calculated by summing the raw scores and dividing
by the number of questions within the scale. These raw efficacy scores were then converted to zscores. In order to accomplish the high and low groupings, raw efficacy scores were converted to
z-scores and the top 33% was deemed high self- or other efficacy while the bottom 33% of
scores were categorized as low self- and other efficacy. For brevity of the results, all swimming
distances will be reported in yards. These will be referred to as the 50, 100, and 200 individual
distances and relate to relay distances of 200, 400, and 800 respectively.
Due to the nested nature of the data, with swimmers grouped into relays, multi-level
modeling was used. Multilevel statistical techniques such as hierarchical linear modeling (HLM)
account for interdependence of the data and allow for an analysis of data at multiple levels
(Raudenbush & Bryk, 2002). For purely hierarchical multi-level models, individuals are assumed
to belong to only one group. However, in this case, swimmers could belong to multiple relay
groups and hence a cross-classified HLM was used. A cross-classified model allows for level-1
units (i.e. swimmers) to be classified in more than one classification (i.e. relays). One benefit of
using this type of HLM model is that is allows for researchers to estimate components of

45

variance between levels of analyses as well as estimate random effects within the data
(Raudenbush & Bush, 2002). Ignoring the cross-classified nature of the data and using a purely
hierarchical data model can lead to bias in standard errors and variance components (Garson,
2013).
The first step was to develop an unconditional model in which no individual predictors
are entered into the model. This step determines the variance of the model and confirms that
HLM is an appropriate statistic to analyze the data. A one way ANOVA was performed
indicating heterogeneity of variance in performance times, F(1, 726) = 11845.65, p < .001.
Raycov and Marcoulides (2008) recommend that for variables with more positive skewness a
logarithmic transformation is an appropriate transformation. Thus, a natural log transformation
was conducted on the dependent variable, Time.
Further, in cross-classified models, variables can be treated as random if the researcher
desires to generalize that effect to other populations (Raudenbush & Bryk, 2002). In this
dissertation it is desirable to use both high and low self- and other efficacy as random effects to
be able to generalize to other populations. However in preliminary data analysis, when both these
variables and their interaction terms (i.e, Relay by high/low efficacy) were treated as random
effects, the models would not converge most likely due to a lack of data and degrees of freedom.
However, deviance comparisons between using only high or low efficacy as a random effect vs.
using these variables as a fixed effect, indicated that there were no differences between the
2

2

models for self-efficacy, χ (5) = 9.12, p = .10, or other efficacy, χ (5) = 4.83, p > .5, therefore
these variables were treated as fixed effects.
Hypothesis 1 was tested through a series of correlations between efficacy measures and
performance. HLM was used to test Hypothesis 2 through 6. Details about the specific models
46

employed are described in Chapter 4. HLM was also used to test research questions 1 through 4.
Means and standard deviations were used to evaluate research questions 5 through 7. All HLM
models were tested using HLM v.7 (Raudenbush, Bryk, & Congdon, 2012).

47

CHAPTER 4
RESULTS
Descriptives
Descriptive statistics for all athletes are presented in Table 3. Relay statistics including
efficacy measures as well as pre- and post-relay measures are presented in Table 4. The
swimmers’ best performance measures varied according to distance raced: 50 (M = 24.24, SD =
1.70), 100 (M = 52.89, SD = 4.11), and 200 (M = 116.05, SD = 8.22). Swimmers reported that
they liked being on relays (M = 8.99, SD = 1.42) and that it was important to be on a
championship relay (M = 7.18, SD = 2.84).
Baseline self-efficacy scores for each distance were moderate; 50 (M = 5.89, SD = 2.29),
100 (M = 6.38, SD = 1.38), and 200 (M = 4.41, SD = 2.50). Relay self-efficacy scores were also
moderate and slightly higher than baseline scores; 50 (M = 6.47, SD = 2.38), 100 (M = 6.73, SD
= 1.13), and 200 (M = 5.11, SD = 2.53). Other-efficacy scores for the relay were also moderate;
50 (M = 6.70, SD = 2.16), 100 (M = 6.83, SD = 1.21), and 200 (M = 5.40, SD = 2.19).
Additionally RISE beliefs for the relay were moderate; 50 (M = 6.52, SD = 2.38), 100 (M = 6.78,
SD = 1.19), and 200 (M = 5.45, SD = 2.42).
Standard deviations for efficacy scores indicate that ranges of efficacy scores were large.
Approximately 50% of the sample indicated that at least one of the events in which efficacy
beliefs were measured was one of their top three events. This indicates that for the rest of the
sample, freestyle, and especially freestyle at these distances may not have been their strengths
which may account for the large range within the scores.

48

Table 1
Pilot Study Swimmer Demographics
Variable

M

SD

Age

15.9

1.50

Years of Competitive Swimming

7.30

2.70

Years on Team

5.60

3.19

Best 100 Freestyle Time

62.56

5.27

Do you like being on relays?*

8.67

1.19

Importance of being on a championship
freestyle relay?*

6.14

2.92

* Likert scale from 0 = Do Not Like to 10 = Absolutely Like
** Likert scale from 0 = Not important to 10 = Very Important

49

Table 2
Estimated marginal performance means and confidence intervals between ranks across
conditions
95% CI
Rank Condition
M
SE
Lower Bound
Upper Bound
1st
ranked

1

65.573

1.631

62.277

68.869

2

65.458

1.425

62.578

68.339

2nd
ranked

1

68.545

1.631

65.248

71.841

2

66.489

1.425

63.608

69.370

3rd
ranked

1

69.591

1.631

66.295

72.887

2

69.878

1.425

66.998

72.759

4th
ranked

1

71.395

1.631

68.099

74.692

2

71.151

1.425

68.270

74.032

Note: Condition 1 = Individual, Condition 2 = Relay

50

Table 3
Dissertation Study Swimmer Demographics
Variable

M

SD

Age

19.39

1.20

Years of Competitive Swimming

10.2

3.35

Best 50 Freestyle Time

24.24

1.70

Best 100 Freestyle Time

52.89

4.11

Best 200 Freestyle Time

116.05

8.22

Do you like being on relays?*

8.99

1.42

Importance of being on a championship
freestyle relay?*

7.18

2.84

* Likert scale from 0 = Do Not Like to 10 = Absolutely Like
** Likert scale from 0 = Not important to 10 = Very Important

51

Table 4
Dissertation Study Relay Variables
Variable

200 Yard Relay

400 Yard Relay

800 Yard Relay

M

SD

M

SD

M

SD

Baseline Self-efficacy

5.89

2.29

6.38

1.38

4.41

2.50

Relay Self-efficacy

6.47

2.38

6.73

1.13

5.11

2.53

Relay Other efficacy

6.70

2.16

6.83

1.21

5.40

2.19

Relay RISE

6.52

2.38

6.78

1.19

5.45

2.42

Pre-Relay Indispensability

7.78

2.15

8.25

1.60

7.91

1.70

Post-Relay Indispensability

8.37

1.60

8.46

1.41

7.59

1.92

Relay Social Comparison

7.73

2.39

7.68

2.27

7.79

2.19

RPE

18.29

2.05

18.08

1.77

17.80

1.98

52

Bivariate Correlations
Bivariate correlations between efficacy and pre and post performance measures for the
200, 400, and 800 relays are presented in Tables 5, 6, and 7 respectively. Baseline self-efficacy
and relay self-efficacy were highly and significantly correlated for all relay distances; 200 (r =
.95), 400 (r = .80), and 800 (r = .93). Additionally, relay self-efficacy and both other-efficacy
and RISE beliefs showed significant relationships at all relay distances respectively; 200 (r = .76
and r = .94), 400 (r = .86 and r = .94), and 800 (r = .89 and r = .95). There were significant
correlations between self- and other efficacy beliefs and performance at both the individual and
relay level however these relationships are discussed in more detail as they relate to Hypothesis
3.
Hypothesis Testing
Hypothesis 1 stated that there would be a negative relationship between both self-efficacy
and other efficacy and performance times. Bivariate correlations between the variables supported
this hypothesis. Both baseline self-efficacy and relay self-efficacy were negatively and
moderately related to individual performance at the 200 (r = -.48 and r = -.52), 400 (r = -.41 and
r = -.43), and 800 distance (r = -.67 and r = -.68). For the relay performance, the same
relationship was found at the 200 (r = -.59 and r = -.63), 400 (r = -.47 and r = -.50), and 800
distance (r = -.30 and r = -.31). Efficacy beliefs were stronger at the relay level compared to the
individual level with the exception of the 800 distance. The relationship between other efficacy
beliefs and performance also demonstrated significantly moderate and negative correlations at
the individual and relay levels for the 200 (r = -.38 and r = -.52), 400 (r = -.33 and r = -.41) and
800 (r = -.59 and r = -.29). Thus Hypothesis 1 was supported.

53

Table 5
Correlations between self-efficacy, other efficacy, RISE beliefs and individual and relay performances for the 200 yard freestyle relay
Variables

1

2

3

4

5

6

7

8

9

1. Baseline Self-efficacy
2. Relay Self-efficacy

.95**

3. Relay Other efficacy

.71**

.76**

4. Relay RISE

.91**

.94**

.81**

5. Individual Time

-.48**

-.52**

-.38**

-.46**

6. Relay Time

-.59**

-.63**

-.52**

-.60**

.88**

7. Pre-Relay Indispensability

.53**

.51**

.38**

.56**

-.32**

-.40**

8. Post Relay Indispensability

.53**

.52**

.42**

.57**

-.24*

-.40**

.65**

9. Post Social Comparison

.17

.23

.20

.27*

-.12

-.26*

.26

.47**

10. RPE

.21

.19

.30*

.23

-.06

-.20

.30*

.35**

Note ** significant at p < .01, * significant at p < .05

54

.32**

Table 6

Correlations between self-efficacy, other efficacy, RISE beliefs and individual and relay performances for the 400 yard freestyle relay
Variables

1

2

3

4

5

6

7

8

9

1. Baseline Self-efficacy
2.Relay Self-efficacy

.80**

3.Relay Other efficacy

.86**

.86**

4.Relay RISE

.79**

.94**

.88**

5.Individual Time

-.41**

-.43**

-.33**

-.44**

6.Relay Time

-.47**

-.50**

-.41**

-.52**

.89**

7.Pre-Relay Indispensability

.42**

.38**

.38**

.39**

-.19

-.20*

8.Post Relay Indispensability

.31**

.28*

.25*

.27*

-.09

-.17

.42**

9.Post Social Comparison

.24*

.20

.15

.16

-.21

-.17

.24

.30*

10. RPE

.05

.22

.20

.18

-.01

-.18

.03

.24*

Note ** significant at p < .01, * significant at p < .05

55

.08

Table 7
Correlations between self-efficacy, other efficacy, RISE beliefs and individual and relay performances for the 800 yard freestyle relay
Variables

1

2

3

4

5

6

7

8

9

1.Baseline Self-efficacy
2.Relay Self-efficacy

.93**

3.Relay Other efficacy

.81**

.89**

4.Relay RISE

.90**

.95**

.87**

5.Individual Time

-.67**

-.68**

-.59**

-.68**

6.Relay Time

-.30**

-.31**

-.29**

-.27**

.46**

7.Pre-Relay Indispensability

.43**

.40**

.42**

.43**

-.26*

-.05

8.Post Relay Indispensability

.49**

.48**

.45**

.45**

-.19

-.13

.39**

9.Post Social Comparison

.25*

.15

.12

.13

-.15

.06

.20

.43**

10. RPE

.12

.26*

.26*

.24*

-.09

-.10

.08

.36**

Note ** significant at p < .01, * significant at p < .0

56

.11

In order to use HLM for testing the hypotheses, an unconditional means model was run to
examine the variance of the times between swimmers (i.e. within relays) and between relays
within the model. For clarity within the results section, between swimmer variance is referred to
as “within relay” variance. The unconditional means model entered and run in HLM appears
below:
LnTIME = θ0 + b00j + c00k + eijk
In this model and the following models, the subscript j, represents the swimmer while the
subscript k, represents the relay. In this model, LnTime was the outcome variable, θ0 was the
overall mean of Time, b00j represents the variance between swimmers, within relays, c00k
represents the variance between relays k, and eijk represents the random error of swimmer j in
relay k. The variance interpretation differs slightly in a cross-classified model as it accounts for
variance between individuals, between relays, and also provides a residual variance. In most
cross-classified models, the within-unit variance is difficult to calculate as the number of data
points within each cell is low due to the cross-classification of units within classifications. This
within-unit variance is calculated as part of the error or residual variance (Hox & Rogers, 2010).
2

In the unconditional model, Time varied significantly between relays, χ (84, N = 85) =
2

258585.73, p < .001, but not within relay, χ (155, N = 156) = 157.96, p = .41. The intraclass
correlation coefficient (ICC) was calculated and indicated that 99% of the variance in Time is
explained through differences between relays (ρc00k = .99). While the variance between
individuals is not significant, Goldstein (2003) cautions that ICC values for cross-classified
models can be affected by the number of units within a classification. In this dissertation, each
57

classification (i.e., relay) has only four units (i.e., swimmers), so within each relay, variability in
times is smaller. This may be a reason as to why in the unconditional model, variability between
individuals within each relay is not significant.
In order to use HLM to analyze data, variance should be significant both between
classifications as well as within classification, however the cross-classified structure of the data
(i.e., units repeated within groups) encourages the use of HLM to analyze these data despite the
fact that variance in the unconditional model was only significant between relays. The next step
in the HLM process was to add the variables of interests Hypothesis 2 stated that swimmers will
swim faster in the relay condition compared to the individual condition. Additionally Hypothesis
3 stated that the weakest relay member would demonstrate a motivation gain in the relay
compared to the individual condition, while Hypothesis 4 stated that the strongest relay member
would demonstrate a motivation loss in the relay compared to the individual condition. To test
these hypotheses the following model was built (Model 1):
Yijk = π0jk + π1jk*(TEAMijk) + π2jk*(D400ijk) + π3jk*(D800ijk) + π4jk*(RANK2ijk) +
π5jk*(RANK3ijk) + π6jk*(RANK4ijk) + π7jk*(TEAMD400ijk) +π8jk*(TEAMD800ijk)
+ π9jk*(TEAMR2ijk) + π10jk*(TEAMR3ijk) + π11jk*(TEAMR4ijk) + eijk
In Model 1, Yijk is the performance time of swimmer j in relay k, π0jk is the Level 1
intercept, π1jk is the effect of performing individually or in the relay, π2jk is the effect of the 400
distance individually, π3jk is the effect of the 800 distance individually, π4jk is the effect of the
second ranked individual, π5jk is the is the effect of the third ranked individual, π6jk is the effect
of the fourth ranked individual, π7jk is the interaction effect of the relay performance at the 400
58

relay distance, π8jk is the interaction effect of the relay performance at the 800 relay distance,
π9jk is the interaction effect of the second ranked individual on the relay, π10jk is the interaction
effect of the third ranked individual on the relay, π11jk is the interaction effect of the fourth
ranked person on the relay, and eijk is the error associated with each swimmer j in relay k.
The results indicated that individually, the second ranked individual, t(279) = 5.52, p <
.001, third ranked individual t(279) = 9,28, p < .001, and fourth ranked individual t(279) = 13.98,
p < .001 swam slower than the first ranked individual. The first ranked individual also swam
slower in the relay condition compared to the individual condition, t(279) = 8.51, p < .001 A
series of contrasts were performed to determine the differences between performance at the
individual level and relay level for other distances and ranks. Results of the contrasts as well as
mean times indicated that the first ranked individual swam slower in the relay compared to the
2

2

individual condition in the 400, χ (1) = 91.97, p < .001, and 800 distance, χ (1) = 123.03, p <
.001. Thus Hypothesis 4 was supported.
The second ranked individual also swam slower in the relay condition compared to the
2

2

individual condition in the 200, χ (1) = 14.61, p < .001, 400, χ (1) = 21.95, p < .001, and 800
2

distance, χ (1) = 38.04, p < .001. The third ranked individual showed the same pattern as the
2

2

first two individuals in the 200, χ (1) = 4.24, p < .05, 400, χ (1) = 7.21, p < .05, and 800
2

distance, χ (1) = 18.66, p < .001. The fourth ranked individual performed significantly faster on
2

the relay performance compared to the individual performance in the 200, χ (1) = 4.22, p < .05,
2

and tended to perform faster on the 400, χ (1) = 2.87, p = .08. Thus Hypothesis 3 regarding the
59

fourth ranked member was supported and Hypothesis 2 concerning a main effect for faster
performance times in the relay was not supported.
2

For Model 1, Time varied significantly between relays, χ (84, N = 85) = 5363.98, p <
2

.001 but not within relays, χ (155, N = 156) = 102.71, p > .5. The ICC for relays, (ρc00k = .90),
demonstrated that 90% of the variance in Time in Model 1 was explained between relays.
Deviance between models was compared and Model 1 was significantly different from the
2

unconditional model, χ (11, N = 15) = 621.32, p < .001. Complete results for this model can be
found in Table 8, while estimated mean times calculated from this model can be found in Table 9
as well as Figure 1 through 3.
Hypothesis 5 stated that self-efficacy would moderate the motivation gain relationship so
that individuals with low self-efficacy beliefs would show a motivation gain in the relay
condition compared to the individual condition. Further, high self-efficacy beliefs would result in
slower times in the relay compared to the individual condition, resulting in a motivation loss.
HLM was used again to test this hypothesis. High and Low Self-Efficacy measures as well as
Relay/Efficacy interaction effects were added to Model 1 allowing for the creation of Model 2:
Yijk = π0jk + π1jk*(TEAMijk) + π2jk*(D400ijk) + π3jk*(D800ijk) + π4jk*(RANK2ijk)
+ π5jk*(RANK3ijk) + π6jk*(RANK4ijk) + π7jk*(TEAMD400ijk) + π8jk*(TEAMD800ijk)
+ π9jk*(TEAMR2ijk) + π10jk*(TEAMR3ijk) + π11jk*(TEAMR4ijk) + π12jk*(LOWSEijk)
+ π13jk*(HIGHSEijk) + π14jk*(TEAMHSEijk) + π15jk*(TEAMLSEijk) + eijk
In Model 2, Yijk is the performance time of swimmer j in relay k, π0jk is the Level 1
intercept, π1jk is the effect of performing individually or in the relay, π2jk is the effect of the 400
60

Table 8
Model 1 for performance time
Fixed Effect

Parameter

Coefficient

SE

df

t

Intercept

Β00

3.15

.013

279

228.41**

Relay

Β10

.03

.004

279

8.51**

Distance 400

Β20

.78

.017

279

43.93**

Distance 800

Β30

1.55

.018

279

82.75**

Rank 2

Β40

.02

.003

279

5.52**

Rank 3

Β50

.03

.003

279

9.28**

Rank 4

Β60

.05

.003

279

13.98**

Relay*Distance
400
Relay*Distance
800

Β70

.00

.004

279

.49

Β80

.00

.004

279

2.05*

Relay*Rank 2

Β90

-.02

.005

279

-4.08**

Relay*Rank 3

Β100

-.02

.005

279

-5.70**

Relay*Rank 4

Β110

-.04

.005

279

-9.28**

Random Effect

Parameter

Variance
Component

SD

df

Χ

Individual

b00j

.000

.000

155

102.71

Relay

c00k

.004

.065

84

5363.98**

Error

e

.000

.020

*p < .05, **p < .001

61

2

Table 9
Model 1 Estimated mean times and percent changes between relay and individual performance
200 Yard
Relay
Individual

Rank 2

Rank 3

Rank 4

Relay

24.43

Individual

23.99
24.42

Individual

24.80

Relay

24.55

Individual

24.80

Relay

24.55

400 Yard
Relay

23.50

Relay

Rank 1

%
Change

%
Change

51.48
3.94**

53.63

53.62

4.18**

53.91

2.04**

53.91

*p < .05, **p < .001

62

4.94**

116.37

2.79**

114.78
1.19*

54.33
-1.00*

116.40
113.21

53.27
.96*

%
Change

110.91

52.55
1.81**

800 Yard
Relay

117.01

1.94**

117.06
-.77

117.01

-.04

200 Distance
Performance (Sec)

25
24.5
Rank 1

24

Rank 2

23.5

Rank 3
Rank 4

23
Individual

Relay
Condition

Performance (Sec)

Figure 1. Individual and relay performance by rank at the 200 distance

56
55.5
55
54.5
54
53.5
53
52.5
52
51.5
51
50.5
50

400 Distance

Rank 1
Rank 2
Rank 3
Rank 4

Individual

Condition

Relay

Figure 2. Individual and relay performance by rank at the 400 distance

63

800 Distance

119

Performance (Sec)

118
117
116
115

Rank 1

114

Rank 2

113

Rank 3

112

Rank 4

111
110
Individual

Condition

Relay

Figure 3. Individual and relay performance by rank at the 800 distance

64

distance individually, π3jk is the effect of the 800 distance individually, π4jk is the effect of the
second ranked individual, π5jk is the effect of the third ranked individual, π6jk is the effect of the
fourth ranked individual, π7jk is the interaction effect of the relay performance at the 400 relay
distance, π8jk is the interaction effect of the relay performance at the 800 distance, π9jk is the
interaction effect of the second ranked individual on the relay, π10jk is the interaction effect of
the third ranked individual on the relay, π11jk is the interaction effect of the fourth ranked person
on the relay, π12jk is the effect of low self-efficacy individually, π13jk is the effect of high selfefficacy individually, π14jk is the interaction effect between relay and high self-efficacy, π15jk is
the interaction effect between relay and low self-efficacy, and eijk is the error associated with
each swimmer j in relay k.
Significant predictors were the same as found in Model 1. The only other significant
predictor was the interaction effect between Relay x High Self-Efficacy t(279) = -5.47, p < .001,
indicating that individuals with high self-efficacy decreased their times from the individual event
to the relay, implying that performance improved.
Estimated means were calculated and contrasts were run in order to determine the effects
of High or Low Self-Efficacy on performance Time from individual to relay within specific
ranks. In general, individuals with low self-efficacy performed worse (increased their times)
2

from the individual to relay condition compared to individuals with high self-efficacy, χ (1) =
23.68, p < .001. However, fourth ranked members with low self-efficacy show no performance

65

2

difference from individual to relay performance for the 200, χ (1) = .00, p >.5, and 400
2

distances, χ (1) = .63, p > .5.
Individuals with high self-efficacy ranked second, showed no significant difference in
2

2

performance in the 200, χ (1) = .09, p > .5, and 400, χ (1) = 1.13, p = .28. While third ranked
2

individuals with high self-efficacy showed no difference in performance in the 200, χ (1) =
2

2

2.54, p = .1, 400, χ (1) = .99, p >.5, and 800, χ (1) = .05, p >.5. Fourth ranked members with
2

high self-efficacy still performed faster from individual to relay performance in the 200, χ (1) =
2

2

29.96, p < .001, 400, χ (1) = 26.33, p < .001, and 800, χ (1) = 14.81, p < .001.
2

With respect to variance, Time varied significantly between relays, χ (84, N = 85) =
51.29, p < .001. ICCs were calculated for relays (ρc00k = .91), demonstrating that 91% of the
variance in Time was explained between relays. Deviance comparisons between this model
2

(Model 2) and Model 1 indicated that Model 2 is significantly different, χ (4) = 66.74, p < .001.
Hypothesis 5 was not supported. Complete results for Model 2 can be found in Table 10.
Estimated mean times and percent changes from individual to relay performance can be found in
Table 11. Mean times are also displayed in Figure 4 through 6.
Hypothesis 6 predicted that other efficacy would moderate the motivation gain
relationship, in that those with higher other efficacy would demonstrate a motivation gain from
the individual to the relay condition. HLM was used to analyze this hypothesis. Model 1 was
used as a starting point and both High and Low Other Efficacy as well as an interaction effect

66

Table 10
Model 2 for performance time, including the moderating effects of self-efficacy
Fixed Effect

Parameter

Coefficient

SE

df

t

Intercept

Β00

3.15

.013

279

234.88**

Relay

Β10

.047

.005

279

9.43**

Distance 400

Β20

.784

.017

279

45.94**

Distance 800

Β30

1.55

.017

279

86.57**

Rank 2

Β40

.020

.003

279

5.90**

Rank 3

Β50

.034

.003

279

9.93**

Rank 4

Β60

.053

.003

279

14.48**

Low Self-Efficacy

Β70

.001

.003

279

.52

High Self-Efficacy

Β80

.001

.004

279

.30

Relay*Distance 400

Β90

.003

.004

279

.76

Relay*Distance 800

Β100

.009

.004

279

2.09*

Relay*Rank 2

Β110

-.022

.004

279

-4.71**

Relay*Rank 3

Β120

-.031

.004

279

-6.68**

Relay*Rank 4

Β130

-.052

.004

279

-10.55**

Low Self-Efficacy

Β140

.001

.003

279

.52

High Self-Efficacy

Β150

.001

.004

279

.30

Parameter

Variance
Component

SD

df

Χ

Individual

b00j

.000

.001

155

117.78

Relay

c00k

.003

.062

84

5359.74**

Error

e

.000

.019

Random Effect

*p < .05, **p < .001

67

2

Table 11
Model 2 Estimated mean times and percent changes between relay and individual performance
200
Yard
Relay
Low
SelfEfficacy

Individual

23.52

Relay

24.78

Individual

24.01

Relay

24.74

Individual

24.34

Relay

24.85

Individual

24.80

Relay

24.82

Individual

23.50

Relay

24.07

Individual

23.99

Relay

24.03

Individual

24.33

Relay

24.13

Individual

24.79

Relay

24.10

%
Change

400
Yard
Relay

%
Change

51.51

800
Yard
Relay

%
Change

111.01

Rank 1

Rank 2

Rank 3

5.39**

54.47

5.74**

52.58
3.06**

54.37

54.62

6.35**

113.33
3.40**

53.32
2.08**

118.07

117.87

4.00**

114.92
2.42**

54.33

118.39

3.02**

117.09

Rank 4
.04

54.54

.37

118.22

.95*

High
SelfEfficacy
Rank 1

Rank 2

Rank 3

Rank 4

51.48
2.41**

52.90

110.94
2.75**

52.55
.15

52.80

53.04

.48

*p < .05, **p < .001

68

52.96

114.47

1.07*

114.84
-.46

54.30
-2.78**

3.35**

113.25

53.29
-.79

114.66

114.97

.11

117.02
-2.45**

114.80

-1.89**

Low Self-Efficacy
Performance (Sec)

25.5
25
24.5
Rank 1

24

Rank 2

23.5

Rank 3

23

Rank 4

22.5
Individual

Relay
Condition

High Self-Efficacy
Performance (Sec)

25.5
25
24.5
Rank 1

24

Rank 2

23.5

Rank 3

23

Rank 4

22.5
Individual

Relay
Condition

Figure 4. Individual and relay performance by rank with low and high self-efficacy beliefs at the
200 distance.

69

Low Self-Efficacy
Performance (Sec)

55
54
53

Rank 1

52

Rank 2
Rank 3

51

Rank 4

50
Individual

Relay
Condition

Performance (Sec)

High Self-Efficacy
55
54.5
54
53.5
53
52.5
52
51.5
51
50.5
50

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

Figure 5. Individual and relay performance by rank with low and high self-efficacy beliefs at the
400 distance

70

Performance (Sec)

Low Self-Efficacy
119
118
117
116
115
114
113
112
111
110
109

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

Performance (Sec)

High Self-Efficacy
119
118
117
116
115
114
113
112
111
110
109

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

Figure 6. Individual and relay performance by rank with low and high self-efficacy beliefs at the
800 distance

71

between Relay by both High and Low Efficacy were added. The completed Model 3 is as
follows:
Yijk = π0jk + π1jk*(TEAMijk) + π2jk*(D400ijk) + π3jk*(D800ijk) + π4jk*(RANK2ijk)
+ π5jk*(RANK3ijk) + π6jk*(RANK4ijk) + π7jk*(TEAMD400ijk) + π8jk*(TEAMD800ijk)
+ π9jk*(TEAMR2ijk) + π10jk*(TEAMR3ijk) + π11jk*(TEAMR4ijk) + π12jk*(LOWOEijk)
+ π13jk*(HIGHOEijk) + π14jk*(TEAMLOEijk) + π15jk*(TEAMHOEijk) + eijk
In Model 3, Yijk is the performance time of swimmer j in relay k, π0jk is the Level 1
intercept, π1jk is the effect of performing individually or in the relay, π2jk is the effect of the 400
distance individually, π3jk is the effect of the 800 distance individually, π4jk is the effect of the
second ranked individual, π5jk is the is the effect of the third ranked individual, π6jk is the effect
of the fourth ranked individual, π7jk is the interaction effect of the relay performance at the 400
relay distance, π8jk is the interaction effect of the relay performance at the 800 distance, π9jk is
the interaction effect of the second ranked individual on the relay, π10jk is the interaction effect
of the third ranked individual on the relay, π11jk is the interaction effect of the fourth ranked
person on the relay, π12jk is the effect of low other efficacy individually, π13jk is the effect of
high other efficacy individually, π14jk is the interaction effect between relay and low other

72

efficacy, π15jk is the interaction effect between relay and high other efficacy, and eijk is the error
associated with each swimmer j in relay k.
Significant predictors were the same as found in Model 1, with the addition of High
Other Efficacy, t(279) = 2.44, p < .05. However this was superseded by a Relay x High Other
Efficacy interaction, t(279) = -5.10, p < .001 indicating that those with high other efficacy in the
relay decreased their time from individual to relay performance. Model 3 was also used to
answer research Question 3, which asked if low other efficacy would lead to a social
compensation effect, however, the Relay x Low Other Efficacy interaction was not significant,
t(279) = 1.06, p = .29. The Relay x Rank 2, t(279) = -4.30, p < .001, Relay x Rank 3, t(279) = 6.21, p < .001, and Relay x Rank 4, t(279) =-9.85, p < .001, interactions were also significant.
To investigate Model 3 further, estimated means were calculated and contrasts were run
to determine the effects of High and Low Other Efficacy on differences in performance from the
individual to the relay condition as it pertained to specific ranks. In general, individuals with low
other efficacy performed worse in the relay condition compared to the individuals with high
2

other efficacy, χ (1) = 11.05, p < .01. The results of the contrasts were very similar to the results
of Self-Efficacy in that second ranked individuals with high other efficacy had no difference in
2

their performance from the individual to the relay condition in the 200, χ (1) = .27, p > .5, and
2

400 distance, χ (1) = .86, p > .5. While the third ranked member with high other efficacy also
2

showed no performance change from the individual to relay performance in the 200, χ (1) =
2

2

1.73, p = .18, 400, χ (1) = .86, p > .5, and 800, χ (1) = .51, p > .5. The fourth ranked member

73

with high other efficacy increased performance from the individual to relay swim in the 200, χ
2

2

2

(1) = 24.70, p < .001, 400, χ (1) = 23.33, p < .001, and 800, χ (1) = 10.13, p < .01.
2

With respect to variance, Time varied significantly between relays, χ (84, N = 85) =
5668.57, p < .001. The ICC for relays was calculated, relays (ρc00k = .91), demonstrating that
91% of the variance in Time was explained between relays. Deviance in this model was
2

compared against Model 1 indicating that the models are significantly different, χ (4) = 42.80, p
2

< .001, and against Model 2, χ (0) = 23.48, p > .5, indicating there was no significant difference
between the models with Self-Efficacy or Other Efficacy as moderators. Hypothesis 6 was
supported only for the fourth ranked member. Full Model 3 results can be found in Table 12.
Estimated mean times and percent changes between individual and relay performance can be
found in Table 13. Mean times are also displayed in Figure 7-9. Model comparisons for all
hypothesis tested models can be found in Table 14 while ICC calculations for hypotheses models
can be found in Table 15.
Research Questions
There were seven additional research questions that were investigated in this dissertation.
The first research question asked if there would be a performance-change difference from
individual to relay performance depending on the serial position of the relay. This research
question was also tested using an HLM model. In this model, Rank was replaced with Position
(2, 3, 4) as well as a Relay/Position interaction effect. The model used for this research question
(Model 4) is below:

74

Table 12
Model 3 for performance time, including the moderating effects of other efficacy
Fixed Effect

Parameter

Coefficient

SE

df

t

Intercept

Β00

3.15

.013

279

226.70**

Relay

Β10

.045

.005

279

9.05**

Distance 400

Β20

.784

.017

279

44.10**

Distance 800

Β30

1.55

.018

279

83.02**

Rank 2

Β40

.020

.003

279

5.78**

Rank 3

Β50

.034

.003

279

9.86**

Rank 4

Β60

.054

.003

279

14.65**

Relay*Distance 400

Β70

.001

.004

279

.40

Relay*Distance 800

Β80

.010

.004

279

2.30*

Relay*Rank 2

Β90

-.020

.004

279

-4.30**

Relay*Rank 3

Β100

-.030

.004

279

-6.21**

Relay*Rank 4

Β110

-.049

.005

279

-9.85**

Low Other Efficacy
High Other
Efficacy
Relay*Low Other
Efficacy
Relay*High Other
Efficacy

Β120

-.002

.003

279

-.66

Β130

.009

.003

279

2.44*

Β140

.004

.004

279

1.06

Β150

-.021

.004

279

-5.10**

Random Effect

Parameter

Variance
Component

SD

df

Χ

Individual

b00j

.000

.001

155

116.62

Relay

c00k

.004

.065

84

5668.57**

Error

E

.000

.019

*p < .05, **p < .001
75

2

Table 13
Model 3 Estimated mean times and percent changes between relay and individual performance
200
Yard
Relay
Low
Other
Efficacy

Individual

23.37

Relay

24.57

Individual

23.86

Relay

24.56

Individual

24.20

Relay

24.69

Individual

24.68

Relay

24.69

Individual

23.66

Relay

24.22

Individual

24.15

Relay

24.22

Individual

24.50

Relay

24.33

Individual

24.98

Relay

24.34

%
Change

400
Yard
Relay

%
Change

51.22

800
Yard
Relay

%
Change

110.30

Rank 1

Rank 2

Rank 3

5.11**

53.93

5.20**

52.28
2.96**

53.92

54.19

6.19**

112.58
3.14**

53.03
2.00**

117.13

117.10

4.01**

114.20
2.18**

54.07

117.68

3.04**

116.44

Rank 4
.06

54.20

.23

117.71

1.08*

High
Other
Efficacy
Rank 1

Rank 2

Rank 3

Rank 4

51.85
2.36**

53.17

111.66
2.54**

52.92
.26

53.16

53.42

.44

*p < .05, **p < .001

76

53.43

115.44

1.29*

115.61
-.49

54.74
-2.55**

3.41**

113.97

53.68
-.66

115.47

116.01

.34

117.88
-2.38**

116.04

-1.55*

Low Other Efficacy
Performance (Sec)

25
24.5
24

Rank 1

23.5

Rank 2
Rank 3

23

Rank 4

22.5
Individual

Relay
Condition

High Other Efficacy
Performance (Sec)

25.5
25
24.5

Rank 1

24

Rank 2
Rank 3

23.5

Rank 4

23
Individual

Relay
Condition

Figure 7. Individual and relay performance by rank with low and high other efficacy beliefs at
the 200 distance

77

Low Other Efficacy
Performance (Sec)

55
54
53

Rank 1

52

Rank 2
Rank 3

51

Rank 4

50
Individual

Relay
Condition

High Other Efficacy
Performance (Sec)

55
54
53

Rank 1

52

Rank 2
Rank 3

51

Rank 4

50
Individual

Relay
Condition

Figure 8. Individual and relay performance by rank with low and high other efficacy beliefs at
the 400 distance

78

Performance (Sec)

Low Other Efficacy
119
118
117
116
115
114
113
112
111
110
109

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

High Other Efficacy
Performance (Sec)

119
117
115

Rank 1

113

Rank 2
Rank 3

111

Rank 4

109
Individual

Relay
Condition

Figure 9. Individual and relay performance by rank with low and high other efficacy beliefs at
the 800 distance

79

Table 14
Deviance comparisons between hypotheses models
χ

2

df

p

Unconditional vs. 1

621.32

11

< .001

1 vs. 2

66.29

4

< .001

1 vs. 3

42.80

4

< .001

2 vs. 3

23.48

0

> .5

Comparison

Table 15
ICCs of random effects pertaining to hypotheses models
Model

Swimmer

Relay

Error

Unconditional

.000

.997

.002

1

.000

.907

.092

2

.000

.912

.087

3

.000

.914

.085

80

Yijk = π0jk + π1jk*(TEAMijk) + π2jk*(D400ijk) + π3jk*(D800ijk) + π4jk*(TEAMD400ijk)
+ π5jk*(TEAMD800ijk) + π6jk*(TEAMP2ijk) + π7jk*(TEAMP3ijk) + π8jk*(TEAMP4ijk)
+ π9jk*(POS2ijk) + π10jk*(POS3ijk) + π11jk*(POS4ijk) + eijk
In Model 4, Yijk is the performance time of swimmer j in relay k, π0jk is the Level 1
intercept, π1jk is the effect of performing individually or in the relay, π2jk is the effect of the 400
distance individually, π3jk is the effect of the 800 distance individually, π4jk is the interaction
effect of swimming on the relay at the 400 distance, π5jk, is the interaction of swimming on the
relay at the 800 distance, π6jk is the interaction effect of the second positioned individual on the
relay, π7jk is the interaction effect of the third positioned individual on the relay, π8jk is the
interaction effect of the fourth positioned individual on the relay, π9jk is the effect of the second
positioned individual, π10jk is the effect of the third position individual, π11jk is the effect of the
fourth position individual, and eijk is the error associated with each swimmer j in relay k.
In Model 4, the only significant predictors were the Relay effect, t(279) = 3.68, p < .001,
in which the relay time was slower in the 200 than the individual time for the individual in the
first position. Unsurprisingly, the 400, t(279) = 44.03, p < .001, and 800 distance, t(279) = 83.05,
p < .001, were significantly slower than the 200 distance. There were no significant interactions
between relay performance and position.
2

For Model 4, Time varied significantly between relays, χ (84, N = 85) = 2713.23, p <
2

.001 and within relays, χ (155, N = 156) = 206.63, p < .01. The ICCs were calculated for relays
81

(ρc00k = .86), and individuals, (ρb00j = .01), demonstrating that 86% of the variance in Time in
this Model 4 was explained between relays while 1% of the variance was explained within
relays, between individuals. The full model details can be found in Table 16 and means and
percent changes between individual and relay condition can be found in Table 17. Mean times
are also displayed in Figure 10-12.
The second research question asked if there are differences with respect to gender and the
moderating effects of self- and other efficacy. HLM was used to test this research question by
creating two additional models. For self-efficacy, Model 2 was used as a start and the variable
Gender, an interaction between Relay x Gender, an interaction between Gender x High/Low
Self-Efficacy, an interaction between Rank x Gender, and an interaction effect between Relay x
Gender x High/Low Self-Efficacy were added. The model used for gender and self-efficacy is
below (Model 5):
Yijk = π0jk + π1jk*(TEAMijk) + π2jk*(D400ijk) + π3jk*(D800ijk) + π4jk*(RANK2ijk)
+ π5jk*(RANK3ijk) + π6jk*(RANK4ijk) + π7jk*(GENDERijk) + π8jk*(LOWSEijk)
+ π9jk*(HIGHSEijk) + π10jk*(TEAMD400ijk) + π11jk*(TEAMD800ijk) +
π12jk*(TEAMR2ijk) + π13jk*(TEAMR3ijk) + π14jk*(TEAMR4ijk) +
π15jk*(TEAMLSEijk) + π16jk*(TEAMHSEijk) + π17jk*(TGENDERijk) +
π18jk*(GENLSEijk) + π19jk*(GENHSEijk) + π20jk*(R2GENijk) + π21jk*(R3GENijk) +
π22jk*(R4GENijk) + π23jk*(TGENLSEijk) + π24jk*(TGENHSEijk) + eijk

82

Table 16
Model 4 for performance time according to position
Fixed Effect

Parameter

Coefficient

SE

df

t

Intercept

Β00

3.17

.013

279

229.27**

Relay

Β10

.005

.005

279

2.68**

Distance 400

Β20

.78

.017

279

44.03**

Distance 800

Β30

1.55

.018

279

83.05**

Position 2

Β40

.009

.004

279

2.09*

Position 3

Β50

.005

.004

279

1.25

Position 4

Β60

.001

.004

279

.35

Relay*Distance 400

Β70

.005

.005

279

.41

Relay*Distance 800

Β80

.005

.005

279

1.61

Relay*Position 2

Β90

-.005

.006

279

-.94

Relay*Position 3

Β100

-.002

.006

279

-.41

Relay*Position 4

Β110

-.008

.006

279

-1.36

Parameter

Variance
Component

SD

df

Χ

Individual

b00j

.006

.000

155

206.63*

Relay

c00k

.064

.064

84

2713.23**

Error

e

.024

.024

Random Effect

*p < .05, **p < .001

83

2

Table 17

Model 4 Estimated mean times and percent changes between relay and individual performance
200 Yard
Relay
24.49
24.24

Relay

24.58

Individual

24.15

Relay

24.57

Individual

24.05

Relay

Position 3

24.10

Individual
Position 2

Individual
Relay

Position 1

24.32

%
Change

400 Yard
Relay

%
Change

52.58
1.98**

53.74

53.94

2.21**

53.92

116.64

2.88*

114.45
1.63**

52.67
1.72*

%
Change

113.38

53.08
1.40*

800 Yard
Relay

117.08

2.30*

114.02
1.95**

52.67

117.02

2.62**

113.56

Position 4
1.13*

53.38

*p < .01, **p < .001

84

1.36*

115.87

2.03**

200 Distance
Performance (Sec)

24.9
24.7
24.5

Position 1

24.3

Position 2

24.1

Position 3

23.9

Position 4

23.7
Individual
Relay
Condition
Figure 10. Individual and relay performance by position at the 200 distance

400 Distance
Performance (Sec)

54.5
54
53.5

Position 1

53

Position 2
Position 3

52.5

Position 4

52
Individual
Relay
Condition
Figure 11. Individual and relay performance by position at the 400 distance

85

800 Distance
Performance (Sec)

118
117
116
Position 1

115

Position 2

114

Position 3

113

Position 4

112
Individual
Relay
Condition

Figure 12. Individual and relay performance by position at the 800 distance

86

In Model 5, Yijk is the performance time of swimmer j in relay k, π0jk is the Level 1
intercept, π1jk is the effect of relay performance at the 200 distance, π2jk is the effect of the 400
distance individually, π3jk is the effect of the 800 distance individually, π4jk is the effect for
second ranked females, π5jk is the is the effect for third ranked females, π6jk is the effect for
fourth ranked females, π7jk is the effect for males individually, π8jk is the effect of low selfefficacy for females individually, π9jk is the effect of high self-efficacy for females individually,
π10jk is the interaction effect of the relay performance at the 400 relay distance, π11jk is the
interaction effect of the relay performance at the 800 distance, π12jk is the interaction effect of
the second ranked individual on the relay, π13jk is the interaction effect of the third ranked
individual on the relay, π14jk is the interaction effect of the fourth ranked individual on the relay,
π15jk is the interaction effect between relay and low self-efficacy beliefs for females, π16jk is the
interaction effect between relay and high self-efficacy beliefs for females, π17jk is the interaction
effect of relay performance for males, π18jk is the interaction effect of low self-efficacy beliefs
for males, π19jk is the interaction effect of high self-efficacy beliefs for males, π20jk is the
interaction effect of second ranked males individually, π21jk is the interaction effect of third
ranked males individually, π22jk is the interaction effect of fourth ranked males individually,
π23jk is the interaction effect of low self-efficacy beliefs and being on the relay for males, π24jk

87

is the interaction effect of high self-efficacy beliefs and being on the relay for females, and eijk is
the error associated with each swimmer j in relay k.
Results demonstrated that there were significant interaction effects between Relay x High
Self-Efficacy, t(279) = -2.03, p < .05 and Relay x Low Self-Efficacy, t(279) = 2.02, p < .05
indicating that individuals performed differently in the relay condition. There were also
significant interaction effects for Relay x Gender, t(279) = 2.73, p < .01 indicating that males
performed differently in the relay condition, and Gender x Low Self-Efficacy, t(279) = 2.78, p <
.01, indicating that males performed differently with low self-efficacy beliefs. Further there were
significant Rank x Gender interactions between the second ranked individual x gender, t(279) = 2.06, p < .05, and the third ranked individual x gender, t(279) = -3.50, p < .05. Finally, there was
a significant interaction between Relay x Gender x High Self-Efficacy, t(279) = -3.71, p < .001.
Contrast and mean times were calculated to demonstrate the differences between specific
Rank according to Gender and High or Low Self-Efficacy. The results indicated that for females,
the fourth ranked member demonstrated a motivation gain from the individual to relay
2

2

performance with high self-efficacy in the 200, χ (1) = 15.94, p < .001, 400, χ (1) = 13.02, p <
2

.001, and 800, χ (1) = 5.05, p < .05. For males with high self-efficacy, the third ranked member
2

performed better from the individual to relay condition in the 200, χ (1) = 7.88, p < .01, and
2

400, χ (1) = 5.19, p < .05. For males with high self-efficacy, the fourth ranked member also
2

showed a motivation gain from the individual to relay performance in the 200, χ (1) = 40.30, p <
2

2

.001, 400, χ (1) = 35.09, p < .001, and 800, χ (1) = 23.71, p < .001.

88

2

For Model 5, Time varied significantly between relays, χ (84, N = 85) = 693.38, p <
.001. The ICC was calculated for relays, (ρc00k = .59), demonstrating that 59% of the variance in
Time in this model is explained between relays. The full model details can be found in Table 18
and means and percent changes between individual and relay condition can be found in Table 19
and Table 20. Mean times are also displayed in Figure 13 through 15.
The second research question also asked about the relationship between other efficacy and
gender. Model 3 was used as a base. Gender, an interaction between Relay x Gender, an
interaction between Gender x High/Low Other Efficacy, and an interaction effect between Relay
x Gender x High/Low Other Efficacy were added. Below is the model used (Model 6):
Yijk = π0jk + π1jk*(TEAMijk) + π2jk*(D400ijk) + π3jk*(D800ijk) + π4jk*(RANK2ijk)
+ π5jk*(RANK3ijk) + π6jk*(RANK4ijk) + π7jk*(GENDERijk) + π8jk*(LOWOEijk)
+ π9jk*(HIGHOEijk) + π10jk*(TEAMD400ijk) + π11jk*(TEAMD800ijk +
π12jk*(TEAMR2ijk) + π13jk*(TEAMR3ijk) + π14jk*(TEAMR4ijk) + π15jk*(TEAMLOEijk) +
π16jk*(TEAMHOEijk) + π17jk*(TGENDERijk) + π18jk*(GENLSEijk) + π19jk*(GENHSEijk) +
π20jk*(R2GENijk) + π21jk*(R3GENijk) + π22jk*(R4GENijk) + π23jk*(TGENLOEijk) +
π24jk*(TGENHOEijk) + eijk
In Model 6, Yijk is the performance time of swimmer j in relay k, π0jk is the Level 1
intercept, π1jk is the effect of relay performance at the 200 distance, π2jk is the effect of the 400
distance individually, π3jk is the effect of the 800 distance individually, π4jk is the effect for

89

Table 18
Model 5 for performance, including the moderating effects of self-efficacy and gender
Fixed Effect

Parameter

Coefficient

SE

df

t

Intercept

Β00

3.20

.006

279

458.40**

Relay

Β10

.040

.005

279

7.46**

Distance 400

Β20

.784

.077

279

111.11**

Distance 800

Β30

1.55

.007

279

210.61**

Rank 2

Β40

.023

.003

279

5.84**

Rank 3

Β50

.039

.003

279

9.99**

Rank 4

Β60

.051

.004

279

12.02**

Gender

Β70

-.111

.008

279

-12.87**

Low Self-Efficacy

Β80

.001

.004

279

.31

High Self-Efficacy

Β90

-.007

.005

279

-1.53

Relay*Distance 400

Β100

.003

.004

279

.82

Relay*Distance 800

Β110

.009

.004

279

2.35*

Relay*Rank 2

Β120

-.020

.004

279

-4.54**

Relay*Rank 3

Β130

-.032

.004

279

-7.01**

Relay*Rank 4

Β140

-.052

.004

279

-10.95**

Relay*Low Self-Efficacy

Β150

-.010

.005

279

-2.02*

Relay*High Self-Efficacy

Β160

-.011

.005

279

-2.03*

Relay*Gender

Β170

.018

.006

279

2.74*

Gender*Low Self-Efficacy

Β180

.022

.008

279

2.74*

Gender*High Self-Efficacy

Β190

-.000

.008

279

-.10

Rank 2*Gender

Β200

-.011

.005

279

-2.11*

Rank 3*Gender

Β210

-.019

.005

279

-3.46**

Rank 4*Gender
Relay*Gender*Low SelfEfficacy
Relay*Gender*High SelfEfficacy

Β220

-0.00

.005

279

-.12

Β230

-.016

.009

279

-1.80

Β240

-.032

.008

279

-3.71**

90

Table 18 (cont’d)
Parameter

Variance
Component

SD

df

Χ

Individual

b00j

.000

.004

155

138.69

Relay

c00k

.000

.023

84

773.29**

Error

e

.000

.018

Random Effect

*p < .05, **p < .001

91

2

Table 19
Model 5 Estimated mean times and percent changes between relay and individual performance
for females
200
Yard
Relay
Low
SelfEfficacy

Individual

24.64

Relay

25.94

Individual

25.23

Relay

26.01

Individual

25.63

Relay

26.12

Individual

25.93

Relay

25.91

Individual

24.42

Relay

25.15

Individual

24.99

Relay

25.22

Individual

25.39

Relay

25.33

Individual

25.69

Relay

25.12

%
Change

400
Yard
Relay

%
Change

54.02

800
Yard
Relay

%
Change

116.48

Rank 1

Rank 2

Rank 3

5.28**

57.07

5.64**

55.30
3.10**

57.21

57.46

6.33**

119.24
3.45**

56.18
1.92**

123.86

124.17

4.13**

121.14
2.27**

56.85

124.71

2.94**

122.58

Rank 4
-.08

56.99

.25

123.70

.90

High
SelfEfficacy
Rank 1

Rank 2

Rank 3

Rank 4

53.52
3.01**

55.33

115.41
3.37**

54.79
.88

55.47

55.71

1.23*

*p < .05, **p < .001

92

55.26

120.38

1.89**

120.03
.07

56.33
-2.23**

4.04**

118.14

55.66
-.26

120.08

120.91

.73

121.45
-1.90**

119.92

-1.26*

Table 20
Model 5 Estimated mean times and percent changes between relay and individual performance
for males
200
Yard
Relay
Low
SelfEfficacy

23.77

Individual

22.80

Relay

23.56

Individual

23.00
23.49

Individual

23.70

Relay

23.73

Individual

21.82

Relay

22.17

Individual

22.08

Relay

21.97

Individual

22.26

Relay

21.90

Individual

22.94

Relay

Rank 3

22.54

Relay

Rank 2

Individual
Relay

Rank 1

22.13

%
Change

400
Yard
Relay

%
Change

49.40
5.49**

52.30

51.83

5.86**

51.66

113.50

6.55**

107.80
3.67**

50.41
2.13**

%
Change

106.52

49.99
3.31**

800
Yard
Relay

112.48

4.34**

108.70
2.48**

51.95

112.13

3.15**

112.02

Rank 4
.11

52.19

.46

113.27

1.11

High
SelfEfficacy
Rank 1

Rank 2

Rank 3

Rank 4

47.83
1.61*

48.77

103.13
1.96**

48.40
-.48

48.33

48.18

-.14

*p < .05, **p < .001

93

48.67

104.89

.51*

105.23
-1.28*

50.30
-3.56**

2.63**

104.36

49.80
-1.61*

105.84

104.56

-.63

108.45
-3.23**

105.63

-2.60**

Female
Performance (Sec)

26
25.5
25

Rank 1

24.5

Rank 2

24

Rank 3

23.5

Rank 4

23
Individual

Relay
Condition

Male
Performance (Sec)

24
23.5
23
Rank 1

22.5

Rank 2

22

Rank 3

21.5

Rank 4

21
Individual

Relay
Condition

Figure 13. Individual and relay performance by rank and gender with high self-efficacy beliefs at
the 200 distance

94

Performance (Sec)

Female
56.5
56
55.5
55
54.5
54
53.5
53
52.5

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

Performance (Sec)

Male
50.5
50
49.5
49
48.5
48
47.5
47
46.5

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

Figure 14. Individual and relay performance by rank and gender with high self-efficacy beliefs at
the 400 distance

95

Female
Performance (Sec)

123
121
119

Rank 1

117

Rank 2
Rank 3

115

Rank 4

113
Individual

Relay
Condition

Male
Performance (Sec)

110
108
106

Rank 1

104

Rank 2
Rank 3

102

Rank 4

100
Individual

Relay
Condition

Figure 15. Individual and relay performance by rank and gender with high self-efficacy beliefs at
the 800 distance

96

second ranked females, π5jk is the is the effect for third ranked females, π6jk is the effect for
fourth ranked females, π7jk is the effect for males individually, π8jk is the effect of low other
efficacy for females individually, π9jk is the effect of high other efficacy for females
individually, π10jk is the interaction effect of the relay performance at the 400 relay distance,
π11jk is the interaction effect of the relay performance at the 800 distance, π12jk is the interaction
effect of the second ranked individual on the relay, π13jk is the interaction effect of the third
ranked individual on the relay, π14jk is the interaction effect of the fourth ranked individual on
the relay, π15jk is the interaction effect between relay and low other efficacy beliefs for females,
π16jk is the interaction effect between relay and high other efficacy beliefs for females, π17jk is
the interaction effect of relay performance for males, π18jk is the interaction effect of low other
efficacy beliefs for males, π19jk is the interaction effect of high other efficacy beliefs for males,
π20jk is the interaction effect of second ranked males individually, π21jk is the interaction effect
of third ranked males individually, π22jk is the interaction effect of fourth ranked males
individually, π23jk is the interaction effect of low other efficacy beliefs and being on the relay for
males, π24jk is the interaction effect of high other efficacy beliefs and being on the relay for
females, and eijk is the error associated with each swimmer j in relay k.
Results of Model 6 indicated that there was a significant Gender main effect, t(279) = 11.86, p < .001. Further, a Relay x Rank 2 interaction, t(279) = -4.38, p < .001, a Relay x Rank 3
97

interaction, t(279) = -6.46, p < .001, and a Relay x Rank 4 interaction, t(279) = -9.91, p < .001,
indicated that these ranks perform differently at the relay level. Further, there were significant
Rank 2 x Gender, t(279) = -2.08, p < .001, and Rank 3 x Gender, t(279) = -2.21, p < .001,
interactions indicating that males at rank 2 and 3 performed differently than females. There was
also a Relay x High Other Efficacy interaction, t(279) = -2.37, p < .05, however this was
superseded by a Relay x Gender x High Other Efficacy interaction, t(279) = -2.54, p < .05.
Contrasts and means were calculated to determine the relationship between Gender and
Other Efficacy within the different ranks. Females ranked fourth with high other efficacy
2

performed better in the relay condition compared to the individual condition in the 200, χ (1) =
2

9.96, p < .01, and 400, χ (1) = 9.81, p < .01. For males with high other efficacy, performance in
2

the relay was also better than the individual condition for the third ranked member in the 200, χ
2

(1) = 7.39, p < .01, and 400, χ (1) = 6.24, p < .05, as well as the fourth ranked member in the
2

2

2

200, χ (1) = 35.11, p < .001, 400, χ (1) = 33.00, p < .001, and 800, χ (1) = 18.39, p < .001.
2

Time varied significantly between relays, χ (84, N = 85) = 1078.42, p < .001. The ICC
was calculated for relays, (ρc00k = .68), demonstrating that 68% of the variance in Time in
Model 6 was explained between relays. The full model details can be found in Table 21 and
means and percent changes between individual and relay condition according to rank and gender
can be found in Table 22 and Table 23. Mean times are also displayed in Figure 16 through 18.
The third research question pertaining to a social compensation effect was addressed
above with Hypothesis 6. The fourth research question asked if there would be a relationship
between RISE beliefs and performance. HLM was also used to test this question and the model
98

Table 21
Model 6 for performance, including the moderating effects of other efficacy and gender
Fixed Effect

Parameter

Coefficient

SE

df

t

Intercept

Β00

3.19

.007

279

408.13**

Relay

Β10

.043

.005

279

8.17**

Distance 400

Β20

.783

.008

279

92.37**

Distance 800

Β30

1.55

.009

279

175.06**

Rank 2

Β40

.023

.003

279

5.96**

Rank 3

Β50

.038

.004

279

9.66**

Rank 4

Β60

.051

.004

279

12.05**

Gender

Β70

-.109

.009

279

-11.86**

Low Other Efficacy
High Other
Efficacy
Relay*Distance 400

Β80

.001

.004

279

.39

Β90

.004

.004

279

.87

Β100

.001

.004

279

.26

Relay*Distance 800

Β110

.010

.004

279

2.38*

Relay*Rank 2

Β120

-.020

.004

279

-4.38**

Relay*Rank 3

Β130

-.030

.004

279

-6.46**

Relay*Rank 4
Relay*Low Other
Efficacy
Relay*High Other
Efficacy
Relay*Gender
Gender*Low Other
Efficacy
Gender*High Other
Efficacy
Rank 2*Gender

Β140

-.048

.004

279

-9.91**

Β150

.003

.005

279

.60

Β160

-.012

.005

279

-2.37*

Β170

.004

.006

279

.76

Β180

.004

.008

279

.49

Β190

-.002

.008

279

-.26

Β200

-.011

.005

279

-2.08*

Rank 3*Gender

Β210

-.012

.005

279

-2.21*

Rank 4*Gender
Relay*Gender*
Low Other Efficacy

Β220

.003

.005

279

.64

Β230

.004

.009

279

.50

99

Table 21 (cont’d)

Relay*Gender*
High Other
Efficacy

Β240

-.021

.008

279

-2.54*

Parameter

Variance
Component

SD

df

Χ

Individual

b00j

.000

.004

155

137.71

Relay

c00k

.000

.029

84

1082.50**

Error

E

.000

.019

Random Effect

*p < .05, **p < .001

100

2

Table 22
Model 6 Estimated mean times and percent changes between relay and individual performance
for females
200
Yard
Relay
Low
Other
Efficacy

Individual

24.57

Relay

25.75

Individual

25.16

Relay

25.83

Individual

25.54

Relay

25.96

Individual

25.87

Relay

25.83

Individual

24.63

Relay

25.40

Individual

25.22

Relay

25.48

Individual

25.60

Relay

25.61

Individual

25.93

Relay

25.48

%
Change

400
Yard
Relay

%
Change

53.78

800
Yard
Relay

%
Change

116.03

Rank 1

Rank 2

Rank 3

4.79**

56.43

4.91**

55.08
2.66**

56.61

56.89

5.88**

118.83
2.78**

55.91
1.64*

122.85

123.26

3.72**

120.61
1.76*

56.64

123.87

2.69**

122.19

Rank 4
-.15

56.62

-.03

123.26

.87

High
Other
Efficacy
Rank 1

Rank 2

Rank 3

Rank 4

53.91
3.14**

55.67

116.31
3.32**

55.21
1.04

55.86

56.13

1.61*

*p < .05, **p < .001

101

55.86

121.61

2.09**

120.90
.16

56.77
-1.72*

4.21**

119.12

56.04
.04

121.21

122.21

1.08

122.48
-1.61*

122.61

-.70

Table 23
Model 6 Estimated mean times and percent changes between relay and individual performance
for males
200
Yard
Relay
Low
Other
Efficacy

Individual

22.11

Relay

23.39

Individual

22.38

Relay

23.20

Individual

22.70

Relay

23.30

Individual

23.37

Relay

23.55

Individual

22.02

Relay

22.33

Individual

22.29

Relay

22.14

Individual

22.61

Relay

22.24

Individual

23.27

Relay

22.48

%
Change

400
Yard
Relay

%
Change

48.40

800
Yard
Relay

%
Change

104.42

Rank 1

Rank 2

Rank 3

5.79**

51.27

5.91**

49.00
3.63**

50.84

51.06

6.88**

105.71
3.75**

49.71
2.61**

111.62

110.69

4.71**

107.23
2.73**

51.16

111.17

3.67**

110.36

Rank 4
.79

51.62

.91

112.39

1.83*

High
Other
Efficacy
Rank 1

Rank 2

Rank 3

Rank 4

48.20
1.40*

48.94

103.99
1.52*

48.80
-.65

48.53

48.74

-.54

*p < .05, **p < .001

102

49.28

105.66

.37

106.79
-1.52*

50.95
-3.38**

2.45**

105.27

49.50
-1.63*

106.55

106.12

-.62

109.91
-3.27**

107.29

-2.38**

Female
Performance (Sec)

26.5
26
25.5

Rank 1

25

Rank 2
Rank 3

24.5

Rank 4

24
Individual

Relay
Condition

Male
Performance (Sec)

24
23.5
23

Rank 1

22.5

Rank 2
Rank 3

22

Rank 4

21.5
Individual

Relay
Condition

Figure 16. Individual and relay performance by rank and gender with high other efficacy beliefs
at the 200 distance

103

Performance (Sec)

Female
57
56.5
56
55.5
55
54.5
54
53.5
53

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

Performance (Sec)

Male
51.5
51
50.5
50
49.5
49
48.5
48
47.5

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

Figure 17. Individual and relay performance by rank and gender with high other efficacy beliefs
at the 400 distance

104

Female
Performance (Sec)

124
122
120

Rank 1

118

Rank 2
Rank 3

116

Rank 4

114
Individual

Relay
Condition

Male
Performance (Sec)

112
110
108
Rank 1

106

Rank 2

104

Rank 3

102

Rank 4

100
Individual

Relay
Condition

Figure 18. Individual and relay performance by rank and gender with high other efficacy beliefs
at the 800 distance

105

consisted of Model 1, with the addition of Low and High RISE beliefs and interaction terms
between RISE beliefs x Relay. The model used (Model 7) is below:
Yijk = π0jk + π1jk*(TEAMijk) + π2jk*(D400ijk) + π3jk*(D800ijk) + π4jk*(RANK2ijk)
+ π5jk*(RANK3ijk) + π6jk*(RANK4ijk) + π7jk*(TEAMD400ijk) + π8jk*(TEAMD800ijk)
+ π9jk*(TEAMR2ijk) + π10jk*(TEAMR3ijk) + π11jk*(TEAMR4ijk) + π12jk*(LOWRISEijk)
+ π13jk*(HIGHRISEijk) + π14jk*(TLRISEijk) + π15jk*(THRISEijk) + eijk
In Model 7, Yijk is the performance time of swimmer j in relay k, π0jk is the Level 1
intercept, π1jk is the effect of performing individually or in the relay, π2jk is the effect of the 400
distance individually, π3jk is the effect of the 800 distance individually, π4jk is the effect of the
second ranked individual, π5jk is the is the effect of the third ranked individual, π6jk is the effect
of the fourth ranked individual, π7jk is the interaction effect of the relay performance at the 400
relay distance, π8jk is the interaction effect of the relay performance at the 800 distance, π9jk is
the interaction effect of the second ranked individual on the relay, π10jk is the interaction effect
of the third ranked individual on the relay, π11jk is the interaction effect of the fourth ranked
person on the relay, π12jk is the effect of low RISE individually, π13jk is the effect of high RISE
individually, π14jk is the interaction effect between relay and high RISE, π15jk is the interaction
effect between relay and low RISE, and eijk is the error associated with each swimmer j in relay
k.

106

All predictors were significant in Model 7 similar to Model 2. In addition, the interaction
between relay x high RISE was a significant predictor of Time, t(279) = -5.59, p < .001. Means
and contrasts were once again performed and followed a similar trend to both Self- and Other
Efficacy. In general, swimmers with low RISE beliefs performed worse from the individual to
2

relay condition than individuals with high RISE beliefs, χ (1) = 41.00, p <.001. Further the only
individuals who performed better in the relay condition compared to the individual condition
2

2

were the fourth ranked members with high RISE beliefs, 200, χ (1) = 28.86, p < .001, 400, χ
2

(1) = 139.78, p < .001, and 800, χ (1) = 13.18, p < .001. Time varied significantly between
2

relays, χ (84, N = 85) = 5443.61, p < .001. The ICC was calculated for relays, (ρc00k = .91),
demonstrating that 91% of the variance in time in this model was explained between relays. Full
model results are presented in Table 24, while estimated means and percent changes between
individual and relay performance are presented in Table 25. Mean times are also displayed in
Figure 19 through 21.
The fifth research question asked if individuals socially compare themselves to others
within their relay team. Means for each relay distance (200: M = 7.73, SD = 2.39; 400: M = 7.68,
SD = 2.27; 800: M = 7.79, SD = 2.19) indicated that relay members compared their own swim
time to others within the relay. In swimming, it is common for coaches to record split times in
one place during the relay event and then swimmers can immediately see their time after the
event. If this was the case, looking or “comparing” ones’ time may not have been intentional but
rather the result of looking at the relay’s split times as a whole.
The sixth research question addressed whether relay members think they make a
contribution to the relay team and if this differs according to whether the assessment is before or
107

Table 24
Model 7 for performance, including the moderating effects of RISE beliefs
Fixed Effect

Parameter

Coefficient

SE

df

t

Intercept

Β00

3.15

.013

279

236.44**

Relay

Β10

.046

.004

279

9.57**

Distance 400

Β20

.784

.017

279

45.93**

Distance 800

Β30

1.55

.017

279

86.50**

Rank 2

Β40

.020

.003

279

5.79**

Rank 3

Β50

.034

.003

279

9.88**

Rank 4

Β60

.053

.003

279

14.47**

Relay*Distance 400

Β70

.002

.004

279

.68

Relay*Distance 800

Β80

.009

.004

279

2.28*

Relay*Rank 2

Β90

-.021

.004

279

-4.62**

Relay*Rank 3

Β100

-.031

.004

279

-6.53**

Relay*Rank 4

Β110

-.050

.004

279

-10.23**

Low RISE

Β120

.002

.003

279

.67

High RISE

Β130

.002

.003

279

.59

Relay*Low RISE

Β140

.005

.004

279

1.24

Relay*High RISE

Β150

-.023

.004

279

-5.59**

Parameter

Variance
Component

SD

df

Χ

Individual

b00j

.000

.001

155

5443.61**

Relay

c00k

.003

.062

84

112.58

Error

E

.000

.019

Random Effect

*p < .05, **p < .001

108

2

Table 25
Model 7 Estimated mean times and percent changes between relay and individual performance
200
Yard
Relay
Low
RISE

Individual

23.52

Relay

24.77

Individual

24.00

Relay

24.73

Individual

24.35

Relay

24.85

Individual

24.81

Relay

24.83

Individual

23.52

Relay

24.06

Individual

24.00

Relay

24.02

Individual

24.34

Relay

24.14

Individual

24.80

Relay

24.12

%
Change

400
Yard
Relay

%
Change

51.54

800
Yard
Relay

%
Change

111.03

Rank 1

Rank 2

Rank 3

5.29**

54.43

5.60**

52.59
3.00**

54.33

54.60

6.35**

113.30
3.31**

53.34
2.05**

118.03

117.88

4.04**

114.92
2.35**

54.36

118.46

3.07**

117.11

Rank 4
.07

54.56

.36

118.38

1.07*

High
RISE
Rank 1

Rank 2

Rank 3

Rank 4

51.52
2.31**

52.87

110.99
2.62**

52.57
.09

52.78

53.04

.38

*p < .05, **p < .001

109

53.00

114.50

1.09*

114.88
-.53

54.34
-2.75**

3.34**

113.26

53.32
-.83

114.70

115.07

.16

117.07
-2.46**

114.99

-1.77*

Low RISE
Performance (Sec)

25
24.5
24

Rank 1

23.5

Rank 2
Rank 3

23

Rank 4

22.5
Individual

Relay
Condition

High RISE
Performance (Sec)

25
24.5
24

Rank 1

23.5

Rank 2
Rank 3

23

Rank 4

22.5
Individual

Relay
Condition

Figure 19. Individual and relay performance by rank with low and high RISE beliefs at the 200
distance

110

Low RISE
Performance (Sec)

55
54
53

Rank 1

52

Rank 2
Rank 3

51

Rank 4

50
Individual

Relay
Condition

High RISE
Performance (Sec)

55
54
53

Rank 1

52

Rank 2
Rank 3

51

Rank 4

50
Individual

Relay
Condition

Figure 20. Individual and relay performance by rank with low and high RISE beliefs at the 400
distance

111

Performance (Sec)

Low RISE
119
118
117
116
115
114
113
112
111
110
109

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

Performance (Sec)

High RISE
119
118
117
116
115
114
113
112
111
110
109

Rank 1
Rank 2
Rank 3
Rank 4
Individual

Relay
Condition

Figure 21. Individual and relay performance by rank with low and high RISE beliefs at the 800
distance

112

after the actual performance. Means of member’s feelings of their contribution before the relay
(200: M = 7.78, SD = 2.15; 400: M = 8.25, SD = 1.60; 800: M = 7.91, SD = 1.70) were slightly
lower than means after the relay (200: M = 8.37, SD = 1.60; 400: M = 8.46, SD = 1.41) with the
exception of the 800 distance, which was slightly lower compared to the before measure (M =
7.59, SD = 1.92). A series of paired sample t tests were conducted and there were no significant
differences at any of the distances between the pre and post measures.
The last research question asked how much effort relay members believed they gave on
their performance. Means for all relay distances (200: M = 18.29, SD = 2.05; 400: M = 18.08,
1.77; 800: M = 17.80, SD = 1.98) indicated swimmers perceived their effort as between “very
hard” and “extremely hard.”
In Summary, Hypothesis 1 was supported in which self- and other efficacy beliefs were
negatively related to performance. Hypothesis 2 was not supported; swimmers did not swim
faster in the relay condition compared to the individual condition. Hypothesis 3 and 4 were both
supported in that the weakest member demonstrated a motivation gain in the relay while the
strongest member demonstrated a motivation loss. Hypothesis 5 was not supported with respect
to self-efficacy while Hypothesis 6 was supported but only for the fourth ranked member.

113

CHAPTER 5
DISCUSSION
Motivation losses between individual and group performance have been frequently
studied within a sport context. However, motivation gains within a group setting have been less
frequently studied. Swimming is a sport in which researchers can easily compare the differences
between individual and group performance. Swimmers in a competitive setting should hold
beliefs about their own capabilities to perform and beliefs about the capabilities of their
teammates in a relay task. These beliefs should also influence the performance motivation of a
swimmer in a competitive performance environment. This dissertation aimed to replicate the
Köhler motivation gain effect in real-time swimming relays rather than through the use of
archival data. Further this dissertation investigated the moderating effects of self- and other
efficacy on relay performance. This chapter discusses the findings of this dissertation, discusses
strengths and limitations of the study, identifies implications of the study, and presents future
research directions.
Performance Results
The major findings in this dissertation demonstrated that overall, swimmers’
performances from best individual times to the relay performance times were dependent on their
rank within the relay. In general, across distances, the first, second, and third ranked individuals
swam significantly slower in the relay condition compared to their individual condition. The
fourth ranked individual swam faster in the relay condition compared to their individual
condition. These findings show support for the Kohler effect (Hypothesis 3).
Previous Köhler effect literature has demonstrated that within dyads, the weakest
individual performs better when paired with a moderately more capable partner compared to how

114

they perform on an individual task (Gockel et al., 2008; Hertel et al., 2000; Kerr et al., 2005).
This performance increase from the weakest member is emphasized under conjunctive task
demands rather than additive task demands (Weber & Hertel, 2007). However, Kerr et al. (2005)
indicated that relays have certain characteristics that allow them to function as quasi-conjunctive
task demands. First, relays are divisible in that each member of the relay must adequately
complete his or her leg of the relay for the team to be successful overall. Similarly, relays are
sequential in that each member of the relay must adequately complete his or her leg of the relay
before the next leg can be started. These characteristics support the finding that within an
additive relay task, performance gains may be more likely to occur.
These findings support other motivation gain literature within swimming relays
specifically, Osborn et al. (2012). Osborn et al. was one of the first studies to demonstrate that
these motivation gains occur from the weakest group member in groups larger than dyads
indicating a performance gain from the fourth ranked or slowest relay members. Within this
dissertation study, the fourth ranked member also demonstrated a performance gain in the relay
condition compared to the individual condition. This finding supported Hypothesis 3. This is an
even more impressive finding practically, as these fourth ranked swimmers were swimming
faster than their lifetime best performances in the relays in order to see these motivation gain
effects. Further, these motivation gains, in both Osborn et al. (2012) and this dissertation study
occurred under a relay task demand which is usually thought of as an additive task demand.
Additional findings of Osborn et al. (2012) indicated that the first ranked member
performed slower in the relay condition compared to the individual condition, however, this was
non-significant. In this dissertation study, the first ranked individual swam significantly slower in
the relay compared to the individual condition demonstrating a motivation loss or a social loafing

115

effect. This result supported Hypothesis 4. Additionally, this is the first time within a sport
context where both motivation gains and social loafing effects occur in a group task
simultaneously. This finding supports the idea that within groups larger than dyads, motivation
processes may be more complex and researchers may need to focus on more than just the
weakest or strongest member. Further, this finding may indicate that the other members within
the relay may show similar trends in performance compared to the weakest or strongest
members. Although Hypothesis 2, that predicted that swimmers would swim faster in the relay
condition compared to the individual condition, was found in a previous study (Osborn et al.,
2012), this main effect hypothesis was not supported in the present study. Only the fourth
ranked relay swimmer swam faster in the relay condition. This may be explained by the measure
for individual performance used in the study, which was the swimmers’ lifetime best
performance. While swimmers were “rested” for these fall invitational meets in which relay time
was measured, the swimmers may not have been as “rested” as they would be for a
championship meet at the end on the season in which most lifetime best performances are swum.
Research Question 1 asked about the position effects of the relay members. Prior research
has found that indispensability increases as a result of one’s position within the relay (Huffmeier
& Hertel, 2011; Huffmeier et al., 2012). However this dissertation model (Model 4) did not
indicate any significant motivation gains for anyone on the relay due to their position. Instead
only motivation losses were found to occur. This finding lends support for the stronger effects of
rank within a relay compared to position (Osborn et al., 2012).
Moderating Effects of Efficacy Beliefs
In terms of efficacy beliefs, both high self-efficacy and high other efficacy significantly
moderated the performance relationships between individual and relay times, specifically for the

116

fourth ranked individual. Hypothesis 5 predicted that self-efficacy would moderate the
performance from the individual to relay condition in that those individuals with low selfefficacy would show a motivation gain and individuals with high self-efficacy would
demonstrate a motivation loss. This hypothesis was not supported. Individuals with low selfefficacy, specifically those ranked first through third actually performed significantly slower on
the relay compared to the individual condition. For fourth ranked individuals with low selfefficacy, there was not a significant increase in time; however, their performances were still
slower. When the effect of self-efficacy was separated by gender (Research Question 2), the
effect became much more interesting.
For both males and females with low self-efficacy, the effects were similar to those
described above. However, when examining the effects of high self-efficacy, performance was
dependent on both rank and gender. For both males and females with high self-efficacy, the first
ranked member swam significantly slower in the relay condition compared to the individual
condition; although this performance loss was less than individuals with low self-efficacy (i.e.,
percent changes from individual to relay performance were less for individuals with high selfefficacy). For the second through fourth-ranked swimmers within the relay, the effects of high
self-efficacy varied according to gender. For females with high self-efficacy beliefs, only the
fourth ranked individuals displayed a motivation gain in the relay performance, while the second
ranked females were significantly slower, displaying a motivation loss. However, for males,
both the third and fourth ranked individuals demonstrated a motivation gain, while the second
ranked member showed no significant difference from individual to relay performance.
Previous research on the moderating effects of self-efficacy and motivation gains
demonstrated that females who are given false feedback about low self-efficacy, increased their

117

performance in a group task compared to high self-efficacy participants (Seok, 2004). That is,
participants who were led to believe that their individual performance was very good compared
to others, did not increase their efforts as much in the group setting as those who thought their
individual performance was sub par. However, in this dissertation low self-efficacy beliefs were
not manipulated and did not result in a significant motivation gain. The only individuals who did
not have a performance difference in the relay with low self-efficacy were the fourth ranked
individuals. The results indicated that they did not swim statistically slower and their
performance may be regarded as at least an increase in effort to swim the same performance time
in both the individual and relay conditions. However, in practical terms, even adding a few tenths
of a second during a relay event can dramatically change where a team finishes overall.
One reason the findings in this dissertation may differ from those found in Seok (2004),
is in the validity of efficacy beliefs. In Seok’s study, participants were given false feedback about
their abilities no matter how well their performance was overall. For example, individuals who
personally thought they did well, and received low self-efficacy feedback, may have still felt
confident in their abilities but just needed to try harder, resulting in a motivation gain on the next
trial. However, in this dissertation, swimmers reported their own efficacy to swim under certain
times. Their confidence may have been based on a greater history of past performances and, thus,
a more accurate assessment of their future performance. Therefore, these lower efficacy
swimmers may not have felt that their contribution was instrumental enough to make a difference
in the relay condition (Hertel et al., 2000). One additional difference could also be that the
swimmers used in this dissertation did not all specialize in freestyle or the distances measured.
The swimmers with low self-efficacy beliefs may not have swum these freestyle events on a

118

regular basis and, therefore, may not have been confident in their abilities to swim fast on a
relay, possibly making these swimmers feel even less instrumental to the relay performance.
Overall, the performance changes for swimmers with high self-efficacy indicated that
they either swam better in the relay condition, indicating a motivation gain (e.g. third and fourth
ranked individuals). Or, these swimmers did not swim as poorly on the relay as swimmers with
low self-efficacy as indicated by lower positive percent changes from individual to relay
performance. Meaning, high self-efficacy swimmers who were ranked first and second still
swam slower in the relay condition compared to swimmers with low-self-efficacy, those with
high self-efficacy added less time resulting in lower percent changes.
Previous research on the Köhler effect has indicated that females are more motivated by
feelings of indispensability while males are more motivated by socially comparing themselves to
others (Weber & Hertel, 2007). Within this dissertation study, there were no significant
differences between self-reports of these measures according to gender. This may indicate that
both of these processes occur simultaneously for both males and females. Swimmers may feel
indispensable to their relay team and, as well as, feel confident in their ability to contribute,
resulting in motivation gains. Further, these swimmers also may have been more confident in
their abilities and therefore felt more confident successfully comparing their performance with
the other faster members.
Hypothesis 6 predicted that other efficacy should also moderate the motivation gain
relationship in that those individuals with higher other efficacy would demonstrate a motivation
gain in the relay condition. This hypothesis was supported once again by the fourth ranked
position for females and both the third and fourth ranked position for males. For the most part,
the moderating effects of other-efficacy were the same with regards to gender and rank as the

119

effects of self-efficacy. However, one additional finding was for the third ranked member of
female relays. These individuals did not swim statistically differently in the relay, however they
did show a decrease in their times.
The results of other efficacy on performance can also be explained through the same
underlying processes of the Köhler effect; social comparison and indispensability. Relay
members indicated that they compared their split times to others within their relay (Research
question 5). Further they also believed that they significantly contributed to the relay and this
feeling of contribution increased slightly after the relay (Research question 6). Correlations
between other efficacy, indispensability, and social comparison measures, indicate that
swimmers with high other efficacy also have both high pre and post relay feelings of
indispensability. However, these individuals did not indicate that they socially compared
themselves to others within the relay. This indicates that individuals with high other efficacy
were more attuned to feelings of indispensability rather than social comparison.
Relay members also indicated that they gave a “hard” or “very hard” effort according to
the Borg scale. This finding makes sense within the context of the study where swimmers were
evaluated within a competitive racing environment. Further, correlations between pre-relay
indispensability and RPE indicates that those individuals who thought they were indispensable to
the relay, also perceived they worked harder on the relay.
One explanation for these findings may have to do with impression management.
Swimmers may have wanted to present a favorable impression of themselves to their other relay
members (Kerr et al., 2005). Impression management may actually complement feelings of
indispensability, in that weaker relay members may feel indispensable to their relay team and
they may try to avoid forming an unfavorable impression of themselves if their performance is

120

poor. This may encourage weaker members to increase their performance in order to appear
favorable to their relay members especially under conditions of high indispensability. However,
the fastest members of the relay may not have felt indispensable to the relay team as their
performance would still be faster than those of the weaker individuals. They may not have felt an
increased need to appear favorable to their relay members and therefore did not increase their
performance. This may explain how there was not a social compensation effect from the fastest
member with feelings of low other efficacy (Research Question 3), but rather a social loafing
effect.
Hypothesis 1 predicted that both self-efficacy and other efficacy would be negatively
related to performance times. This hypothesis was supported as demonstrated by negative and
moderate correlations between these variables. This finding also supports other research on the
moderate relationship between efficacy beliefs and performance (Moritz et al., 2000). However,
research on efficacy beliefs has been mixed in that there is support for both efficacy beliefs being
negatively related to performance (Vancouver et al., 2002) as well as being positively related to
performance (Gilson et al., 2012) at within-subjects levels. In this dissertation, the correlation
was negative; however in the context of swimming relays, better performances were actually
represented as lower times. So in this case, if efficacy beliefs were high at the between-subjects
level, performance was also better as indicated by a lower performance time. The fact that
Vancouver et al.’s within-subjects studies have found a slight negative relationship between selfefficacy and performance does not mean that high self-efficacy is detrimental to performance.
For self-efficacy to be detrimental to performance, researchers would have to show that
individuals with high self-efficacy are consistently outperformed by individuals with low selfefficacy. This is not what has been observed in the extant research (Moritz et al., 2000). For

121

example, the individual regression slopes found in Beattie et al., 2011, clearly show that
individuals with high-efficacy outperform individuals with low self-efficacy regardless of
whether individual optimal putting occurred. The results in the present dissertation support the
research on self-efficacy in sport, in which efficacy beliefs are positively related to performance
(i.e., higher efficacy beliefs are correlated with better performance).
Research question 4 asked about the effects of RISE beliefs on performance. The results
indicated that these were similar to the results for both self-efficacy and other efficacy. Once
again these may be explained through the high correlations between these three variables
indicating that those individuals with high self-efficacy and other efficacy often had high RISE
beliefs as well. The same explanations for impression management may also have an effect here
(Kerr et al., 2005). Those individuals with high RISE beliefs (e.g., an individual was confident
that their other relay members were confident in that individual’s ability to swim fast), may have
felt like their relay members expected them to swim well and if they did not perform well, this
might leave a unfavorable impression.
Despite the fact that both self-efficacy and other efficacy demonstrate the same pattern
with respect to motivation gains within relays, these relays are the first time in which self- and
other efficacy have been studied in regards to the Köhler effect. Further, by using the tripartite
model of efficacy beliefs (Lent & Lopez, 2002), this study allows a better understanding of how
efficacy beliefs are impacted both by group member’s perceptions, and in turn, how these
efficacy beliefs influence performance. This study also helps to clarify how performance changes
in groups larger than dyads, in that both social loafing and motivation gains occur
simultaneously. Further this study can provide researchers with a better understanding of how
different ranks perform with regard to both self and other efficacy.

122

Strengths and Limitations
A strength of this dissertation is that it was conducted in the field with competitive
swimming relay teams in which swimmers should have been already motivated to perform. This
gives the phenomenon of the Kohler effect some ecological validity. Further, prior research on
motivation gains in swimming relays has used archival data rather than live relays (Huffmeier &
Hertel, 2011; Huffmeier et al, 2012; Osborn et al., 2012). Using live relays allowed the collection
of subjective measures with regards to efficacy beliefs and the underlying processes of social
comparison and indispensability.
Additionally, prior research of the Kohler effect, conducted with laboratory tasks, has
used mostly conjunctive tasks (e.g., Gockel, et al., 2008; Hertel et al., 2000; Kerr et al., 2005).
In Steiner’s (1972) taxonomy of group tasks, swimming relays are considered an additive task
but function in a quasi-conjunctive manner (Kerr et al., 2005). Although, some motivation gains
have been found using coactive and additive tasks (e.g., Feltz et al., 2012; Hertel et al., 2003),
the present dissertation found motivation gain effects for inferior group members in a team task
that is divisible and sequential. Thus, the present study’s finding adds to the applicability of
group tasks that are not purely conjunctive.
Another advantage of this dissertation is that it used groups larger than dyads, which
allows for better understanding of how multiple individuals perform in a group task. Further, it
organized these groups by ability into ranks. This allowed the analysis to demonstrate how
performance changes as a result of ability within the relay and how other variables like efficacy
beliefs, distance of race, and gender moderate this relationship.
Lastly, the use of HLM for the data analysis is also a strength of the study. Within meets,
swimmers can swim in multiple relays. Depending on their abilities in certain events, swimmers

123

might have different levels of confidence for each race distance. By using HLM to model the
data, and therefore take into account these dependencies, it allows the researcher to get a more
accurate picture of how self-efficacy, other efficacy, distance of the race, and swimmer’s rank
within the relay impacts overall performance time. Without the use of HLM, the data analysis
would split each relay distance up and treat both the race itself as well as efficacy as an
independent event when in reality swimmers are connected to their other performances and
efficacy beliefs.
Of course there were also some limitations in this dissertation. While the sample size was
large enough to produce 86 separate relays, in regards to the HLM analysis, there were not
enough relays to allow self- and other efficacy to act as random effects. This limited the ability to
generalize the effects of efficacy beliefs beyond this population (Raudenbush & Bryk, 2002).
Another limitation was the high correlations between self-efficacy, other efficacy, and
RISE beliefs. While instructions for the swimmers were presented both orally before they took
the surveys as well as written on the surveys, swimmers may not have been aware of the subtle
differences between efficacy scales which may be one reason why the scores were very similar.
Because of the similarity between scores, it was difficult to see direct relationships between
changes in performance and changes in efficacy beliefs. This may be modified in future studies
by further clarifying the differences in each scale.
Further, because of the use of three distinctly different relay distances, the variance
between the relays was much greater than the variance between the swimmers (i.e., within the
relays). This was problematic in the explanations of variance from a HLM standpoint as the
results demonstrated that most of the variance was between relays and there was no significant
variation explained between swimmers. While the use of multiple relay distances may have

124

given a more accurate picture of how swimmers swim at a meet overall, it may have also limited
the explanations of the statistical analysis.
One other limitation was the amount of tapering among teams for each meet in which the
study was conducted. While the study did not assess the state of rest of each team directly,
coaches may use different definitions of “rest” and this can cause validly issues within self-report
data. While each team “rested” for each meet, some teams were more rested than others, based
upon conversations with coaches, which also can impact performance. The teams that were more
rested were most likely more confident in their abilities to swim fast as well as more confident in
their teammates’ abilities to swim fast, which may have enhanced both self- and other efficacy
beliefs.
Similarly, the individual performance times used were the lifetime best performance from
different meets. These lifetime best times were used because not each individual swimmer swam
the individual condition of each event. These individual performance times may have been swum
last year or even further back which also may change how likely it is that swimmers were going
to swim near, close to, or faster than these times. A more accurate measure would be to use
individual times swum at that specific meet even if it limits the sample. This would be a more
accurate and comparable time to relay performance.
Implications
This study has shown that both motivation gains as well as losses occur within swimming
relays. Further, high self- and other efficacy moderate these performance differences for different
ranked members. Specifically the fourth ranked members demonstrate motivation gains with
high self-efficacy or high other efficacy. Third ranked relay members also demonstrate
motivation gains for males and perform faster albeit not significantly faster for females. This

125

may demonstrate that both these lower ranked individuals within the relay have similar views
about their abilities within the relay and through either feelings of indispensability or social
comparison, increase their performance in the relay condition. Within groups larger than dyads,
the third and fourth ranked members may function similarly while the first and second ranked
members may function similarly. Coaches should promote both self-efficacy as well as other
efficacy among their slowest relay members to try to enhance these motivation gain effects.
Further, coaches should try to emphasize the importance of the strongest two members’
contributions as this may reduce the effects of social loafing (Baron & Kerr, 2003). Coaches can
also emphasize feelings of indispensability for every member as well as the importance of
leaving favorable impressions for one’s other relay members. This may be especially true in
teams where performance has occurred before and swimmers have a concrete knowledge of how
other members perform (Karau & Williams, 1993).
Future Directions
There are a few directions for future research that can be conducted with respect to
motivation gains and the moderating effects of self- and other efficacy. First, due to the nature of
the study and how relay data were collected in the field, it was not possible to make an
individualized assessment of other efficacy for each swimmer within each relay. One follow-up
would be to use an aggregated measure of other efficacy for individual’s other relay members.
This might be a better measure of other efficacy within this context as it gives a direct measure
of how confident an individual is in each one of his or her teammates instead of generalizing
about the group.
Further, using individual performance and relay performance measured from the same
meet should allow a more accurate comparison. This would allow for researchers to generalize

126

how the swimmer is swimming in one point in time and then how that performance changes as a
result of the either the individual or relay context.
Another area to expand on this research is to ask more in-depth follow up questions
concerning the underlying processes of the Köhler effect. The questions asked in this dissertation
were limited in the amount of information they provided and also might have been confounded
given the context in which they were asked. Formal and more in-depth questioning might give
researchers a better understanding of how the processes work and whether swimmers are aware
of these processes during their performance.
While motivation gains have been demonstrated in multiple relay situations (Huffmeier
& Hertel, 2011; Huffmeier et al., 2012; Osborn et al., 2012), this dissertation research provides
an understanding of how each relay member performs based upon their rank. This research
indicates that performance in a group larger than dyads is different according to rank. This
research could be expanded by looking at these motivation gains in other sports besides
swimming. While similar results have been found in field relays (Osborn et al., 2012), expanding
the research into more group sport tasks in which performance can be directly measured would
allow for these findings to be generalized to sports beyond swimming. Some of these sports may
include track relays, team golf scores, and bowling.
Conclusion
Overall, this dissertation provides insight into motivation within groups. The majority of
motivation theories define motivation based upon a cognitive antecedent. An individual’s
thoughts determine an individual’s motivational intensity and direction (Roberts, 1992). Due to
the cognitive nature of motivation, it is difficult to measure motivation directly. However the
literature commonly infers that performance is an accurate measure of motivation. These

127

performance differences are inferred from effort, which is then inferred to be a result of
motivation.
This dissertation supports the cognitive nature of motivation and infers that motivation
occurs as a result of performance differences between individual and relay performance. Further,
self- and other efficacy beliefs modify this performance relationship and can be viewed as
moderators to performance and subsequently motivation.
The Köhler effect has been studied primarily in regards to dyads; however, often groups
are larger than dyads, and having a better understanding of how individuals function within a
group performance is important. Specifically, this is the first time in which both social loafing
effects and motivation gains were demonstrated within the same group within a sport context.
These findings allow for a better understanding of both motivation and how rank and ability
within a group task, specifically, may contribute to performance. With regards to self efficacy,
this dissertation adds to the literature supporting a positive relationship between self-efficacy and
performance. This dissertation also allows for a better understanding about how both high selfand other efficacy enhance performance at the group level for weaker members within the group.

128

APPENDICES

129

APPENDIX A
Demographic Questionnaire
Name:
Age:
Year in School:

Freshman

Sophomore

Junior

Senior

Fifth Year

Years of Competitive Swimming:
Top 3 Events:
Personal best times for those top 3 events (SCY)
Personal best time in the 50, 100, and 200 freestyle (SCY):
Personal Goals for this season:
Team goals for this season:
On the following questions answer the questions as they relate to the scale
1. Do you like swimming on relays?
Do not Like
1
2
3
Why do you feel that way?

4

5

6

7

8

Absolutely Like
9
10

2. How important is it to you to be on a championship relay this season?
Not important
1
2
3
Why do you feel this way?

4

5

6

7

8

3. Have you been on a freestyle championship relay team before?
YES
NO
If yes, which one:
What is your favorite position?
First

Second

Third

Why?

130

Fourth

Very Important
9
10

APPENDIX B
Self-Efficacy Scale, Individual, Female 50 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 50 yard freestyle today.
Rate your degree of confidence that YOU can attain the different levels of performance by circling a
number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 50 free in under :28.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 50 free in under :27.5 sec
0%

10%

20%

30%

40%

3. Swim the 50 free in under :27.0 sec
0%

10%

20%

30%

40%

4. Swim the 50 free in under :26.5 sec
0%

10%

20%

30%

40%

5. Swim the 50 free in under :26.0 sec
0%

10%

20%

30%

40%

6. Swim the 50 free in under :25.5 sec
0%

10%

20%

30%

40%

7. Swim the 50 free in under :25.0 sec
0%

10%

20%

30%

40%

8. Swim the 50 free in under :24.5 sec
0%

10%

20%

30%

40%

9. Swim the 50 free in under :24.0 sec
0%

10%

20%

30%

40%

131

Appendix C
Self-Efficacy Scale, Individual, Male 50 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 50 yard freestyle today.
Rate your degree of confidence that YOU can attain the different levels of performance by circling a
number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 50 free in under :25.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 50 free in under :24.5 sec
0%

10%

20%

30%

40%

3. Swim the 50 free in under :24.0 sec
0%

10%

20%

30%

40%

4. Swim the 50 free in under :23.5 sec
0%

10%

20%

30%

40%

5. Swim the 50 free in under :23.0 sec
0%

10%

20%

30%

40%

6. Swim the 50 free in under :22.5 sec
0%

10%

20%

30%

40%

7. Swim the 50 free in under :22.0 sec
0%

10%

20%

30%

40%

8. Swim the 50 free in under :21.5 sec
0%

10%

20%

30%

40%

9. Swim the 50 free in under :21.0 sec
0%

10%

20%

30%

40%

132

Appendix D
Self-Efficacy Scale, Individual, Female 100 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 100 yard freestyle today.
Rate your degree of confidence that YOU can attain the different levels of performance by circling a
number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 100 free in under 1:06 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 100 free in under 1:04 min
0%

10%

20%

30%

40%

3. Swim the 100 free in under 1:02 min
0%

10%

20%

30%

40%

4. Swim the 100 free in under 1:00 min
0%

10%

20%

30%

40%

5. Swim the 100 free in under :58 sec
0%

10%

20%

30%

40%

6. Swim the 100 free in under :56 sec
0%

10%

20%

30%

40%

7. Swim the 100 free in under :54 sec
0%

10%

20%

30%

40%

8. Swim the 100 free in under :52 sec
0%

10%

20%

30%

40%

9. Swim the 100 free in under :50 sec
0%

10%

20%

30%

40%

133

Appendix E
Self-Efficacy Scale, Individual, Male 100 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 100 yard freestyle today.
Rate your degree of confidence that YOU can attain the different levels of performance by circling a
number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 100 free in under :59 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 100 free in under :57 sec
0%

10%

20%

30%

40%

3. Swim the 100 free in under :55 sec
0%

10%

20%

30%

40%

4. Swim the 100 free in under :53 sec
0%

10%

20%

30%

40%

5. Swim the 100 free in under :51 sec
0%

10%

20%

30%

40%

6. Swim the 100 free in under :49 sec
0%

10%

20%

30%

40%

7. Swim the 100 free in under :47 sec
0%

10%

20%

30%

40%

8. Swim the 100 free in under :45 sec
0%

10%

20%

30%

40%

9. Swim the 100 free in under :43 sec
0%

10%

20%

30%

40%

134

Appendix F
Self-Efficacy Scale, Individual, Female 200 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 200 yard freestyle today.
Rate your degree of confidence that YOU can attain the different levels of performance by circling a
number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 200 free in under 2:06 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 200 free in under 2:04 min
0%

10%

20%

30%

40%

3. Swim the 200 free in under 2:02 min
0%

10%

20%

30%

40%

4. Swim the 200 free in under 2:00 min
0%

10%

20%

30%

40%

5. Swim the 200 free in under 1:58 min
0%

10%

20%

30%

40%

6. Swim the 200 free in under 1:56 min
0%

10%

20%

30%

40%

7. Swim the 200 free in under 1:54 min
0%

10%

20%

30%

40%

8. Swim the 200 free in under 1:52 min
0%

10%

20%

30%

40%

9. Swim the 200 free in under 1:50 min
0%

10%

20%

30%

40%

135

Appendix G
Self-Efficacy Scale, Individual, Male 200 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 200 yard freestyle today.
Rate your degree of confidence that YOU can attain the different levels of performance by circling a
number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 200 free in under 1:57 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 200 free in under 1:55 min
0%

10%

20%

30%

40%

3. Swim the 200 free in under 1:53 min
0%

10%

20%

30%

40%

4. Swim the 200 free in under 1:51 min
0%

10%

20%

30%

40%

5. Swim the 200 free in under 1:49 min
0%

10%

20%

30%

40%

6. Swim the 200 free in under 1:47 min
0%

10%

20%

30%

40%

7. Swim the 200 free in under 1:45 min
0%

10%

20%

30%

40%

8. Swim the 200 free in under 1:43 min
0%

10%

20%

30%

40%

9. Swim the 200 free in under 1:41 min
0%

10%

20%

30%

40%

136

Appendix H
Self-Efficacy Scale, Relay, Female 50 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 50 yard freestyle today on the RELAY.
Rate your degree of confidence that YOU can attain the different levels of performance on your RELAY
SWIM by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 50 free in under :28.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 50 free in under :27.5 sec
0%

10%

20%

30%

40%

3. Swim the 50 free in under :27.0 sec
0%

10%

20%

30%

40%

4. Swim the 50 free in under :26.5 sec
0%

10%

20%

30%

40%

5. Swim the 50 free in under :26.0 sec
0%

10%

20%

30%

40%

6. Swim the 50 free in under :25.5 sec
0%

10%

20%

30%

40%

7. Swim the 50 free in under :25.0 sec
0%

10%

20%

30%

40%

8. Swim the 50 free in under :24.5 sec
0%

10%

20%

30%

40%

9. Swim the 50 free in under :24.0 sec
0%

10%

20%

30%

40%

137

Appendix I
Self-Efficacy Scale, Relay, Male 50 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 50 yard freestyle today on the RELAY.
Rate your degree of confidence that YOU can attain the different levels of performance on your RELAY
SWIM by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 50 free in under :25.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 50 free in under :24.5 sec
0%

10%

20%

30%

40%

3. Swim the 50 free in under :24.0 sec
0%

10%

20%

30%

40%

4. Swim the 50 free in under :23.5 sec
0%

10%

20%

30%

40%

5. Swim the 50 free in under :23.0 sec
0%

10%

20%

30%

40%

6. Swim the 50 free in under :22.5 sec
0%

10%

20%

30%

40%

7. Swim the 50 free in under :22.0 sec
0%

10%

20%

30%

40%

8. Swim the 50 free in under :21.5 sec
0%

10%

20%

30%

40%

9. Swim the 50 free in under :21.0 sec
0%

10%

20%

30%

40%

138

Appendix J
Self-Efficacy Scale, Relay, Female 100 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 100 yard freestyle today on the RELAY.
Rate your degree of confidence that YOU can attain the different levels of performance on your RELAY
SWIM by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 100 free in under 1:06 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 100 free in under 1:04 min
0%

10%

20%

30%

40%

3. Swim the 100 free in under 1:02 min
0%

10%

20%

30%

40%

4. Swim the 100 free in under 1:00 min
0%

10%

20%

30%

40%

5. Swim the 100 free in under :58 sec
0%

10%

20%

30%

40%

6. Swim the 100 free in under :56 sec
0%

10%

20%

30%

40%

7. Swim the 100 free in under :54 sec
0%

10%

20%

30%

40%

8. Swim the 100 free in under :52 sec
0%

10%

20%

30%

40%

9. Swim the 100 free in under :50 sec
0%

10%

20%

30%

40%

139

Appendix K
Self-Efficacy Scale, Relay, Male 100 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 100 yard freestyle today on the RELAY.
Rate your degree of confidence that YOU can attain the different levels of performance on your RELAY
SWIM by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 100 free in under :59 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 100 free in under :57 sec
0%

10%

20%

30%

40%

3. Swim the 100 free in under :55 sec
0%

10%

20%

30%

40%

4. Swim the 100 free in under :53 sec
0%

10%

20%

30%

40%

5. Swim the 100 free in under :51 sec
0%

10%

20%

30%

40%

6. Swim the 100 free in under :49 sec
0%

10%

20%

30%

40%

7. Swim the 100 free in under :47 sec
0%

10%

20%

30%

40%

8. Swim the 100 free in under :45 sec
0%

10%

20%

30%

40%

9. Swim the 100 free in under :43 sec
0%

10%

20%

30%

40%

140

Appendix L
Self-Efficacy Scale, Relay, Female 200 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 200 yard freestyle today on the RELAY.
Rate your degree of confidence that YOU can attain the different levels of performance on your RELAY
SWIM by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 200 free in under 2:06 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 200 free in under 2:04 min
0%

10%

20%

30%

40%

3. Swim the 200 free in under 2:02 min
0%

10%

20%

30%

40%

4. Swim the 200 free in under 2:00 min
0%

10%

20%

30%

40%

5. Swim the 200 free in under 1:58 min
0%

10%

20%

30%

40%

6. Swim the 200 free in under 1:56 min
0%

10%

20%

30%

40%

7. Swim the 200 free in under 1:54 min
0%

10%

20%

30%

40%

8. Swim the 200 free in under 1:52 min
0%

10%

20%

30%

40%

9. Swim the 200 free in under 1:50 min
0%

10%

20%

30%

40%

141

Appendix M
Self-Efficacy Scale, Relay, Male 200 Freestyle
Please rate how confident you are that YOU can attain the different levels of performance when
swimming the 200 yard freestyle today in the RELAY.
Rate your degree of confidence that YOU can attain the different levels of performance on your RELAY
SWIM by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below

1. Swim the 200 free in under 1:57 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

50%

60%

70%

80%

90%

100%

2. Swim the 200 free in under 1:55 min
0%

10%

20%

30%

40%

3. Swim the 200 free in under 1:53 min
0%

10%

20%

30%

40%

4. Swim the 200 free in under 1:51 min
0%

10%

20%

30%

40%

5. Swim the 200 free in under 1:49 min
0%

10%

20%

30%

40%

6. Swim the 200 free in under 1:47 min
0%

10%

20%

30%

40%

7. Swim the 200 free in under 1:45 min
0%

10%

20%

30%

40%

8. Swim the 200 free in under 1:43 min
0%

10%

20%

30%

40%

9. Swim the 200 free in under 1:41 min
0%

10%

20%

30%

40%

142

Appendix N
Other Efficacy Scale, Female 200 Freestyle Relay
Today you will swim in the 200 yard freestyle relay

Please rate how confident you are in the OTHER MEMBERS of your relay team in attaining the
different levels of performance.
Rate your degree of confidence that the OTHER MEMBERS of your relay team can attain the different
levels of average split times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 200 free relay average split time under :28.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

2. Swim the 200 free relay average split time under :27.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 200 free relay average split time under :27.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 200 free relay average split time under :26.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

5. Swim the 200 free relay average split time under :26.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

6. Swim the 200 free relay average split time under :25.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

7. Swim the 200 free relay average split time under :25.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

8. Swim the 200 free relay average split time under :24.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

9. Swim the 200 free relay average split time under :24.0 sec
0%

10%

20%

30%

40%

50%

143

60%

70%

Appendix O
Other Efficacy Scale, Male 200 Freestyle Relay
Today you will swim in the 200 yard freestyle relay

Please rate how confident you are in the OTHER MEMBERS of your relay team in attaining the
different levels of performance.
Rate your degree of confidence that the OTHER MEMBERS of your relay team can attain the different
levels of average split times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 200 free relay average split time under :25.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

2. Swim the 200 free relay average split time under :24.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 200 free relay average split time under :24.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 200 free relay average split time under :23.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

5. Swim the 200 free relay average split time under :23.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

6. Swim the 200 free relay average split time under :22.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

7. Swim the 200 free relay average split time under :22.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

8. Swim the 200 free relay average split time under :21.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

9. Swim the 200 free relay average split time under :21.0 sec
0%

10%

20%

30%

40%

50%

144

60%

70%

Appendix P
Other Efficacy Scale, Female 400 Freestyle Relay
Today you will swim in the 400 yard freestyle relay

Please rate how confident you are in the OTHER MEMBERS of your relay team in attaining the
different levels of performance.
Rate your degree of confidence that the OTHER MEMBERS of your relay team can attain the different
levels of average split times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 400 free relay average split time under 1:06 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

2. Swim the 400 free relay average split time under 1:04 min
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 400 free relay average split time under 1:02 min
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 400 free relay average split time under 1:00 min
0%

10%

20%

30%

40%

50%

60%

5. Swim the 400 free relay average split time under :58 sec
0%

10%

20%

30%

40%

50%

60%

6. Swim the 400 free relay average split time under :56 sec
0%

10%

20%

30%

40%

50%

60%

7. Swim the 400 free relay average split time under :54 sec
0%

10%

20%

30%

40%

50%

60%

8. Swim the 400 free relay average split time under :52 sec
0%

10%

20%

30%

40%

50%

60%

9. Swim the 400 free relay average split time under :50 sec
0%

10%

20%

30%

40%

50%

145

60%

Appendix Q
Other Efficacy Scale, Male 400 Freestyle Relay
Today you will swim in the 400 yard freestyle relay

Please rate how confident you are in the OTHER MEMBERS of your relay team in attaining the
different levels of performance.
Rate your degree of confidence that the OTHER MEMBERS of your relay team can attain the different
levels of average split times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 400 free relay average split time under :59 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

2. Swim the 400 free relay average split time under :57 sec
0%

10%

20%

30%

40%

50%

60%

3. Swim the 400 free relay average split time under :55 sec
0%

10%

20%

30%

40%

50%

60%

4. Swim the 400 free relay average split time under :53 sec
0%

10%

20%

30%

40%

50%

60%

5. Swim the 400 free relay average split time under :51 sec
0%

10%

20%

30%

40%

50%

60%

6. Swim the 400 free relay average split time under :49 sec
0%

10%

20%

30%

40%

50%

60%

7. Swim the 400 free relay average split time under :47 sec
0%

10%

20%

30%

40%

50%

60%

8. Swim the 400 free relay average split time under :45 sec
0%

10%

20%

30%

40%

50%

60%

9. Swim the 400 free relay average split time under :43 sec
0%

10%

20%

30%

40%

50%

146

60%

Appendix R
Other Efficacy Scale, Female 800 Freestyle Relay
Today you will swim in the 800 yard freestyle relay

Please rate how confident you are in the OTHER MEMBERS of your relay team in attaining the
different levels of performance.
Rate your degree of confidence that the OTHER MEMBERS of your relay team can attain the different
levels of average split times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 800 free relay average split time under 2:06 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

2. Swim the 800 free relay average split time under 2:04 min
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 800 free relay average split time under 2:02 min
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 800 free relay average split time under 2:00 min
0%

10%

20%

30%

40%

50%

60%

70%

5. Swim the 800 free relay average split time under 1:58 min
0%

10%

20%

30%

40%

50%

60%

70%

6. Swim the 800 free relay average split time under 1:56 min
0%

10%

20%

30%

40%

50%

60%

70%

7. Swim the 800 free relay average split time under 1:54 min
0%

10%

20%

30%

40%

50%

60%

70%

8. Swim the 800 free relay average split time under 1:52 min
0%

10%

20%

30%

40%

50%

60%

70%

9. Swim the 800 free relay average split time under 1:50 min
0%

10%

20%

30%

40%

50%

147

60%

70%

Appendix S
Other Efficacy Scale, Male 800 Freestyle Relay
Today you will swim in the 800 yard freestyle relay

Please rate how confident you are in the OTHER MEMBERS of your relay team in attaining the
different levels of performance.
Rate your degree of confidence that the OTHER MEMBERS of your relay team can attain the different
levels of average split times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 800 free relay average split time under 1:57 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

2. Swim the 800 free relay average split time under 1:55 min
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 800 free relay average split time under 1:53 min
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 800 free relay average split time under 1:51 min
0%

10%

20%

30%

40%

50%

60%

70%

5. Swim the 800 free relay average split time under 1:49 min
0%

10%

20%

30%

40%

50%

60%

70%

6. Swim the 800 free relay average split time under 1:47 min
0%

10%

20%

30%

40%

50%

60%

70%

7. Swim the 800 free relay average split time under 1:45 min
0%

10%

20%

30%

40%

50%

60%

70%

8. Swim the 800 free relay average split time under 1:43 min
0%

10%

20%

30%

40%

50%

60%

70%

9. Swim the 800 free relay average split time under 1:41 min
0%

10%

20%

30%

40%

50%

148

60%

70%

Appendix T
Relation-Inferred Self-Efficacy Scale, Female 200 Freestyle Relay
Today you will swim in the 200 yard freestyle relay

Please rate how confident your OTHER RELAY MEMBERS are in YOUR ability in attaining
the different levels of performance.
Rate your relay members’ degree of confidence that YOU can attain the different levels of average split
times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 200 free relay average split time under :28.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

2. Swim the 200 free relay average split time under :27.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 200 free relay average split time under :27.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 200 free relay average split time under :26.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

5. Swim the 200 free relay average split time under :26.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

6. Swim the 200 free relay average split time under :25.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

7. Swim the 200 free relay average split time under :25.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

8. Swim the 200 free relay average split time under :24.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

9. Swim the 200 free relay average split time under :24.0 sec
0%

10%

20%

30%

40%

50%

149

60%

70%

Appendix U
Relation-Inferred Self-Efficacy Scale, Female 400 Freestyle Relay
Today you will swim in the 400 yard freestyle relay

Please rate how confident your OTHER RELAY MEMBERS are in YOUR ability in attaining
the different levels of performance.
Rate your relay members’ degree of confidence that YOU can attain the different levels of average split
times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 400 free relay average split time under 1:06 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

2. Swim the 400 free relay average split time under 1:04 min
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 400 free relay average split time under 1:02 min
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 400 free relay average split time under 1:00 min
0%

10%

20%

30%

40%

50%

60%

5. Swim the 400 free relay average split time under :58 sec
0%

10%

20%

30%

40%

50%

60%

6. Swim the 400 free relay average split time under :56 sec
0%

10%

20%

30%

40%

50%

60%

7. Swim the 400 free relay average split time under :54 sec
0%

10%

20%

30%

40%

50%

60%

8. Swim the 400 free relay average split time under :52 sec
0%

10%

20%

30%

40%

50%

60%

9. Swim the 400 free relay average split time under :50 sec
0%

10%

20%

30%

40%

50%

150

60%

Appendix V
Relation-Inferred Self-Efficacy Scale, Female 800 Freestyle Relay
Today you will swim in the 800 yard freestyle relay

Please rate how confident your OTHER RELAY MEMBERS are in YOUR ability in attaining
the different levels of performance.
Rate your relay members’ degree of confidence that YOU can attain the different levels of average split
times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 800 free relay average split time under 2:06 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

2. Swim the 800 free relay average split time under 2:04 min
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 800 free relay average split time under 2:02 min
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 800 free relay average split time under 2:00 min
0%

10%

20%

30%

40%

50%

60%

70%

5. Swim the 800 free relay average split time under 1:58 min
0%

10%

20%

30%

40%

50%

60%

70%

6. Swim the 800 free relay average split time under 1:56 min
0%

10%

20%

30%

40%

50%

60%

70%

7. Swim the 800 free relay average split time under 1:54 min
0%

10%

20%

30%

40%

50%

60%

70%

8. Swim the 800 free relay average split time under 1:52 min
0%

10%

20%

30%

40%

50%

60%

70%

9. Swim the 800 free relay average split time under 1:50 min
0%

10%

20%

30%

40%

50%

151

60%

70%

Appendix W
Relation-Inferred Self-Efficacy Scale, Male 200 Freestyle Relay
Today you will swim in the 200 yard freestyle relay

Please rate how confident your OTHER RELAY MEMBERS are in YOUR ability in attaining
the different levels of performance.
Rate your relay members’ degree of confidence that YOU can attain the different levels of average split
times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 200 free relay average split time under :25.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

2. Swim the 200 free relay average split time under :24.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 200 free relay average split time under :24.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 200 free relay average split time under :23.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

5. Swim the 200 free relay average split time under :23.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

6. Swim the 200 free relay average split time under :22.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

7. Swim the 200 free relay average split time under :22.0 sec
0%

10%

20%

30%

40%

50%

60%

70%

8. Swim the 200 free relay average split time under :21.5 sec
0%

10%

20%

30%

40%

50%

60%

70%

9. Swim the 200 free relay average split time under :21.0 sec
0%

10%

20%

30%

40%

50%

152

60%

70%

Appendix X
Relation-Inferred Self-Efficacy Scale, Male 400 Freestyle Relay
Today you will swim in the 400 yard freestyle relay

Please rate how confident your OTHER RELAY MEMBERS are in YOUR ability in attaining
the different levels of performance.
Rate your relay members’ degree of confidence that YOU can attain the different levels of average split
times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 400 free relay average split time under :59 sec
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

70%

80%

90%

100%

2. Swim the 400 free relay average split time under :57 sec
0%

10%

20%

30%

40%

50%

60%

3. Swim the 400 free relay average split time under :55 sec
0%

10%

20%

30%

40%

50%

60%

4. Swim the 400 free relay average split time under :53 sec
0%

10%

20%

30%

40%

50%

60%

5. Swim the 400 free relay average split time under :51 sec
0%

10%

20%

30%

40%

50%

60%

6. Swim the 400 free relay average split time under :49 sec
0%

10%

20%

30%

40%

50%

60%

7. Swim the 400 free relay average split time under :47 sec
0%

10%

20%

30%

40%

50%

60%

8. Swim the 400 free relay average split time under :45 sec
0%

10%

20%

30%

40%

50%

60%

9. Swim the 400 free relay average split time under :43 sec
0%

10%

20%

30%

40%

50%

153

60%

Appendix Y
Relation-Inferred Self-Efficacy Scale, Male 800 Freestyle Relay
Today you will swim in the 800 yard freestyle relay

Please rate how confident your OTHER RELAY MEMBERS are in YOUR ability in attaining
the different levels of performance.
Rate your relay members’ degree of confidence that YOU can attain the different levels of average split
times by circling a number from:
0 (cannot do at all) to 100 (positive you can do) using the scale given below
1. Swim the 800 free relay average split time under 1:57 min
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

80%

90%

100%

2. Swim the 800 free relay average split time under 1:55 min
0%

10%

20%

30%

40%

50%

60%

70%

3. Swim the 800 free relay average split time under 1:53 min
0%

10%

20%

30%

40%

50%

60%

70%

4. Swim the 800 free relay average split time under 1:51 min
0%

10%

20%

30%

40%

50%

60%

70%

5. Swim the 800 free relay average split time under 1:49 min
0%

10%

20%

30%

40%

50%

60%

70%

6. Swim the 800 free relay average split time under 1:47 min
0%

10%

20%

30%

40%

50%

60%

70%

7. Swim the 800 free relay average split time under 1:45 min
0%

10%

20%

30%

40%

50%

60%

70%

8. Swim the 800 free relay average split time under 1:43 min
0%

10%

20%

30%

40%

50%

60%

70%

9. Swim the 800 free relay average split time under 1:41 min
0%

10%

20%

30%

40%

50%

154

60%

70%

Appendix Z
Pre Relay Questionnaire
1. How indispensible is your upcoming swim to the outcome of the relay?
No Contribution
Maximum Contribution
1
2
3
4
5
6
7
8
9
10

155

Appendix AA
Post Relay Questionnaire
1. Did you compare your split time to others’ split times in the relay?
No Comparison
Maximum Comparison
1 2 3
4
5
6
7
8
9
10
2. As you were preparing to start, how much did you think you would contribute to the
relay?
No Contribution
Maximum Contribution
1 2 3
4
5
6
7
8
9
10
3. Please describe your exertion on the previous swim on the relay:
6

No Exertion At All

7
8

Extremely Light

9
10
11

Light

12
13

Somewhat Hard

14
15

Hard

16
17

Very Hard

18
19

Extremely Hard

20

Maximal Exertion

156

Appendix AB
Confidentiality Assurance
While the nature of this research contains sensitive information regarding yourself and
other teammates, it is of absolute importance that you are completely honest about how you feel.
Your answers will help to determine very important information with respect to confidence and
motivation.
Your answers will be kept strictly confidential and will only be seen by the researcher
conducting the study. Your coach, parents, and other teammates will not have access to any part
of your survey. When the researchers are done with your survey, your name will be removed
from all parts of the survey and your answers will no longer be identifiable.
When you are finished with the survey, please return it to the researcher and place it
directly into the provided envelope.

157

REFERENCES

158

REFERENCES

Bandura, A. (1997). Self-Efficacy: The exercise of control. New York: W. H. Freeman.
Baron, R. S., & Kerr, N. L. (2003). Group process, group decision, group action (2nd ed.).
Buckingham, England: Open University Press.
Beattie, S., Lief, D., Adamoulas, M., & Oliver, E. (2011). Investigating the possible negative
effects of self-efficacy on golf putting performance. Psychology of Sport and Exercise,
12(4), 434–441.
Beauchamp, M. R., & Whinton, L. C. (2005). Self-efficacy and other-efficacy in dyadic
performance: Riding as one in equestrian eventing. Journal of Sport and Exercise
Psychology, 27, 245-252.
Borg, G. (1970). Perceived exertion as an indicator of somatic stress. Scandinavian journal of
rehabilitation medicine, 2, 92–98.
Buckley, J. P, & Borg, G. A. (2011). Borg’s scales in strength training; from theory to practice in
young and older adults. Applied Physiology, Nutrition, and Metabolism, 36(5), 682-692.
Dithurbide, L. (2010). Teammate efficacy and teammate trust: An examination of team dynamics
in volleyball defense. Doctoral dissertation, Michigan State University.
Dunlop, W. L., Beatty, D. J., & Beauchamp, M. R. (2011). Examining the influence of otherefficacy and self-efficacy on personal performance. Journal of Sport and Exercise
Psychology, 33, 586-593.
Everett, J. J., Smith, R. E., & Williams, K. D. (1992). Effects of team cohesion and
identifiability on social loafing in relay swimming performance. International
Journal of Sport Psychology, 23, 311-324.
Feltz, D. L. (1982). A path analysis of the causal elements in Bandura’s theory of self-efficacy
and an anxiety-based model of avoidance behavior. Journal of Personality and Social
Psychology, 42, 764-781.
Feltz, D. L. (1988). Gender differences in the causal elements of self-efficacy on a highavoidance motor task. Journal of Sport and Exercise Psychology, 10, 151-166.
Feltz, D. L., Irwin, B. C., & Kerr, N. K. (2011). Match making: An examination of discrepancy
in ability as a moderator of motivation gains in partnered exercise games. Journal of
Diabetes Science and Technology.
Feltz, D. L., Landers, D. M., & Reader, U. (1979). Enhancing self-efficacy in high-avoidance

159

motor tasks: A comparison of modeling techniques. Journal of Sport Psychology, 1, 112122.
Feltz, D. L., & Mungo, D. A. (1983). A replication of the path analysis of the causal elements in
Bandura’s theory of self-efficacy and the influence of automatic perception. Journal of
Sport Psychology, 5, 263-277.
Feltz, D.L., Short, S.E., & Sullivan, P.J. (2008). Self-efficacy in sport: Research and strategies
for working with athletes, teams, and coaches. Champaign, IL: Human Kinetics.
Garson, D. (2013). Hierarchical linear modeling: Guide and applications. Newbury Park, CA:
Sage.
Gilson, T. A., Chow, G. M., & Feltz, D. L. (2012). Self-effiacy and athletic squat performance:
Positive or negative influences at the within- and between-levels of analysis. Journal of
Applied Social Psychology, 42, 1467-1485.
Gockel, C., Kerr, N. L., Seok, D.-H., & Harris, D. W. (2008). Indispensability and group
identification as sources of task motivation. Journal of Experimental Social Psychology,
44, 1316-1321.
Goldstein, H. (2003). Multilevel statistical models. (3rd edition ed.). London: Arnold.
Hertel, G., Deter, C., & Konradt, U. (2003). Motivation gains in computer-mediated work
groups. Journal of Applied Social Psychology, 33, 2080-2105.
Hertel, G., Kerr, N. L., & Messé, L. A. (2000). Motivation gains in groups: Paradigmatic and
theoretical advances on the Köhler effect. Journal of Personality and Social Psychology,
79, 580-601.
Hertel, G., Niemeyer, G., & Clauss, A. (2008). Social indispensability or social comparison: The
why and when of motivation gains in inferior group members. Journal of Applied Social
Psychology, 38, 1329-1363.
Hox, J. & Roberts, K. (2010). Handbook of advanced multilevel analysis. New York, NY:
Routledge.
Hüffmeier, J., & Hertel, G. (2011). When the whole is more than the sum of its parts: Motivation
gains in the wild. Journal of Experimental Social Psychology, 47, 455-459.
Hüffmeier, J., Krumm, S., Kanthak, J., & Hertel, G. (2012). “Don’t let the group down”: Facets
of instrumentality moderate the motivating effects of groups in a field experiment.
European Journal of Social Psychology, 42, 533-538.
Jackson, B., Beauchamp, M. R., & Knapp, P. (2007). Relational efficacy beliefs in athlete dyads:

160

An investigation using actor-partner interdependence models. Journal of Sport and
Exercise Psychology, 29, 170-189.
Jackson, B., Grove, J. R., & Beauchamp, M. R. (2010). Relational efficacy beliefs and
relationship quality within coach-athlete dyads. Journal of Social and Personal
Relationships, 27(8), 1035-1050.
Jackson, B., Knapp, P., & Beauchamp, M. R. (2008). Origins and consequences of tripartite
efficacy beliefs within elite athlete dyads. Journal of Sport and Exercise Psychology, 30,
512-540.
Karau S. J., & Williams, K. D. (1993). Social loafing: A meta-analytic review and theoretical
integration. Journal of Personality and Social Psychology, 45, 681-706.
Karau, S. J., & Williams, K. D. (1997). The effects of group cohesiveness on social loafing and
social compensation. Group Dynamics, 1, 156-168.
Kerr, N. L., & Bruun, S. E. (1983). Dispensability of member effort and group motivation losses:
Free-rider effects. Journal of Personality and Social Psychology, 44, 78-94.
Kerr, N. L., Messé, L .A., Park, E. S., & Sambolec, E. (2005). Identifiability, performance
feedback and the Köhler effect. Group Processes and Intergroup Relations, 8, 375-390.
Kerr, N. L., Messé, L. A., Seok, D.-H., Sambolec, E. J., Lount, R. B., Jr., & Park, E. S. (2007).
Psychological mechanisms underlying the Köhler motivation gain. Personality and
Social Psychology Bulletin, 33, 828-841.
Lent, R. W., & Lopez, F. G. (2002). Cognitive ties that bind: A tripartite view of efficacy beliefs
in growth-promoting relationships. Journal of Social and Clinical Psychology, 21, 256286.
McAuley, E. (1985). Modeling and self-efficacy: A test of Bandura’s model. Journal of Sport
Psychology, 7, 283-295.
Miles, J. A., & Greenberg, J. (1993). Using punishment threats to attenuate social loafing effects
among swimmers. Organizational Behavior and Human Decision Processes, 56, 246265.
Moritz, S. E., Feltz, D. L., Fahrbach, K. R., & Mack, D. E. (2000). The relation of self-efficacy
measures to sport performance: A meta-analytic review. Research Quarterly for Exercise
and Sport, 71, 280-294.
Messé, L. A., Hertel, G., Kerr, N. L., Lount, R. B., Jr., & Park, E. S. (2002). Knowledge of
partner’s ability as a moderator of group motivation gains: An exploration of the Köhler
discrepancy effect. Journal of Personality and Social Psychology, 82, 935-946.

161

Osborn, K. A., Irwin, B. C., Skogsberg, N. J., & Feltz, D. L. (2012) The Köhler effect:
Motivation gains and losses in real sports groups. Sport, Exercise, and Performance
Psychology,
Raudenbush, S.W., & Bryk, A.S. (2002). Hierarchical linear models: Applications and data
analysis methods. Newbury Park, CA: Sage.
Raudenbush, S.W., Bryk, A. S., & Congdon, R. (2012). Hierarchial Linear and Nonlinear
Modeling (Version 7). Lincolnwood, IL: Scientific Software International.
Raycov, T. & Marcoulides, G. A. (2008). An introduction to applied multivariate analysis.
Routledge Academic.
Roberts, G. C. (1992). Motivation in sport and exercise. Champaign, IL: Human Kinetics.
Scherr, J., Wolfarth, B., Christle, J. W., Pressler, A., Wagenpfeil, S., & Halle, M. (2013).
Associations between Borg’s rating of perceived exertion and physiological measures of
exercise intensity. European Journal of Applied Physiology, 113(1), 147-155.
Seok, D.-H. (2004). Exploring self-efficacy as a possible moderator of the Köhler discrepancy
effect. Unpublished Master’s thesis, Michigan State University.
Skatrud-Mickelson, M., Benson, J., Hannon, J. C., & Askew, E. W. (2011). A comparison of
subjective and objective measures of physical exertion. Journal of Sports Sciences,
29(15), 1635-1644.
Steiner, I. D. (1972). Group process and productivity, New York: Academic Press.
Stewart, A. M., & Hopkins, W. G. (2000). Consistency of swimming performance within and
between competitions. Medicine and Science in Sports and Exercise, 32, 997-1001.
Todd, A. R., Seok, D., Kerr, N. L., & Messé, L. A. (2006). Social compensation: Fact or social
comparison artifact? Group Processes and Intergroup Relations, 9, 431-442.
Vancouver, J. B., & Kendall, L.N. (2006). When self-efficacy negatively relates to motivation
and performance in a learning context. Journal of Applied Psychology, 91, 1146-1153.
Vancouver, J. B., Thompson, C. M., Tischner, E. C., & Putka, D. J. (2002). Two studies
examining the negative effect of self-efficacy on performance. Journal of Applied
Psychology, 87, 506-516.
Vancouver, J. B., Thompson, C. M., & Williams, A. A. (2001). The changing signs in the
relationships among self-efficacy, personal goals, and performance. Journal of Applied
Psychology, 86, 605-620.
Weber, B., & Hertel, G. (2007). Motivation gains of inferior group members: A meta-analytical
162

review. Journal of Personality and Social Psychology, 93, 973-993.
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, 570-581.
Williams, K. D., Nida, S. A., Baca, L. D., & Latané, B. (1989). Social loafing and swimming:
Effects of identifiability on individual and relay performance of intercollegiate
swimmers. Basic and Applied Social Psychology, 10, 73-81.

163