3 O... u .0.

 

 

 

 

.
z A.
z a.
..........mu:. .1
1, 4 .
.3! ‘

.8“?

3:11.24!
1. 1 iv!
:

he... "5

u.

, Elva“.

v.5 47‘

a .1

 

Thigh

This is to certify that the
thesis entitled

The Effects of Competitiveness of Task Performance: An
Interactional Approach

presented by
Eric J. Sambolec

has been accepted towards fulfillment
of the requirements for the

M. A. degree in Psychology

N5” .2... 2/2/W7 2.2 .. 2%24

Major Professor’s Signature

/ZI//3//o 9/ /5//;':/04

Date

 

 

 

MSU Is an AMnnatlvo Action/Equal Opportunity Institution

 

_ LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIFlC/DateDue.p65-p. 15

 

THE EFFECTS OF COMPETITIVENESS ON TASK PERFORMANCE: AN
INTERACTIONAL APPROACH

By

Eric J. Sambolec

A THESIS

Submitted to
Michigan State University
in partial fiJlfillment of the requirements
for the degree of

MASTERS OF ARTS
Department of Psychology

2002

 

ABSTRACT

THE EFFECTS OF COMPETITIVENESS ON TASK PERFORMANCE: AN
INTERACTIONAL APPROACH

By

Eric J. Sambolec

In previous research, competitiveness has been studied as a trait, a state, or
interactionally. Research has shown that competitiveness is related to performance
outcomes. However, gaps remain in the literature on the effects of competitiveness on
task performance across situations. This paper explores literature from sport, social, and
personality psychology. Several themes are examined. Based on this review, the author
hypothesized that competitiveness can be primed to influence performance at tasks.
Other hypotheses involve gender difl‘erences and situational differences. Differences in
interpersonal competitiveness as a trait were examined in relation to group performance.
Competitiveness was manipulated through supraliminal priming. Results showed that
this priming influenced performance at a group persistence task. Results also partially
replicated past research on group motivation gains (Kohler, 1926; 1927; Hertel, Kerr,
Messé, 1999). However, some findings were counter to previous research in this area.
Explanations about gender differences are proposed. These explanations take into
account the competing hypotheses of Instrumentality X Value motives versus Social
Comparison / Goal Setting motives to perform well in groups. Implications for group

performance in social, athletic, and work domains are discussed.

Copyright by
Eric J. Sambolec

2002

To all those who supported and all those who stood in the way of my dreams.

ACKNOWLEDGMENTS

The list of people I need to thank for getting me to this point in my life is far too
long to mention here. However, I will try to thank the most influential. First of all I am
permanently indebted to my parents who brought me into the world and raised me well
and with love. I would also like to thank all of my friends and family who were there for
me when I needed them.

A round of thank you’s also goes out to my committee: Larry Messé, Norb Kerr,
and Isis Settles who gave me plenty of feedback and research ideas throughout this long
process. When I first started graduate school, I wanted to examine how competitiveness
and interpersonal liking influenced group performance on sports teams. Larry and Norb
helped me to refine these ideas into an experimental social psychological format. Before
and during the time it took to conduct and write this thesis I have been fortunate to guide
many talented athletes on two very special cross country and track teams. The first is Mt.
St. Mary Academy in Buffalo and the second is Lansing Community College. Thank you
ladies for allowing me to use my somewhat unusual approach to coach you. This time
has been very influential to help me develop and apply my research ideas.

My interest in group performance enhancement comes out of my love for running,
particularly team cross country running. Thank you to all of those who have guided me
in my rumring endeavors. Thank you to all who have helped me through this difficult
time during which I lost my beloved ability to run and one very dear to me. Without all

of you I never would have finished this thesis. Woo hool

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. viii
LIST OF FIGURES ................................................................................. ix
CHAPTER 1
INTRODUCTION .................................................................................... 1

Trait Competitiveness ....................................................................... 2

State Competitiveness ....................................................................... 5

Person-Situation Interaction Approaches to Competitiveness ......................... 8

Gender Differences ........................................................................ 11

Tasks ......................................................................................... l 5

Hypotheses .................................................................................. 18
CHAPTER 2

STUDY 1 ............................................................................................. 22

Method ....................................................................................... 22

Participants ......................................................................... 22

Measures ........................................................................... 22

Procedure ........................................................................... 25

Results ....................................................................................... 27

Malleability of Social Values .................................................... 27

Posttest Measures of Competitiveness .......................................... 29

Discussion ................................................................................... 34
CHAPTER 3

STUDY 2 ............................................................................................. 39

Method ....................................................................................... 39

Participants ......................................................................... 39

Apparatus and Measures .......................................................... 40

Procedure ........................................................................... 42

Results ....................................................................................... 50

Characteristics of Pretest and Experimental Samples ........................ 50

Task Performance and Priming Effects ........................................ 54

Posttest Measures of Trait Competitiveness ................................... 61

Mood Effects ....................................................................... 62

Discussion ................................................................................... 62

Priming and Task Demands in Trial 3 .......................................... 62

Priming and Task Demands in Trial 4 .......................................... 65

Trait Competitiveness ............................................................ 66

Malleability of Trait Competitiveness .......................................... 67

Mood Measures .................................................................... 68

vi

CHAPTER 4

GENERAL DISCUSSION ........................................................................ 70
Measures .................................................................................... 7O
Implications for Motivation Gains and Kohler Effect ................................. 71
Limitations .................................................................................. 73
Directions for Future Research ........................................................... 7 5
CHAPTER 5
CONCLUSIONS AND IMPLICATIONS ....................................................... 78
APPENDIX A
Measure of Social Values .......................................................................... 81
APPENDIX B
Competitiveness Index .............................................................................. 84
APPENDD( C
Sports Orientation Questionnaire .................................................................. 86
APPENDIX D
Competitive Word Task Prime .................................................................... 89
APPENDIX E
Cooperative Word Task Prime ..................................................................... 92
APPENDIX F
Neutral Word Task Prime .......................................................................... 95
APPENDIX G
Competitive Story Prime ........................................................................... 98
APPENDIX H
Cooperative Story Prime ........................................................................ 100
APPENDIX 1
Neutral Story Prime ............................................................................... 102
APPENDIX J
Supplemental Analyses ........................................................................... 104
REFERENCES .................................................................................... 107

vii

LIST OF TABLES

Table 2.1 - Factor Loading for the SOQ in Study 1 ........................................... 30
Table 2.2 — Scores on Competitiveness Measures After Priming by Condition ............ 33
Table 3.1 — Scores on SOQ and Subscales by Gender on Pretest Sample .................. 51
Table 3.2 — SOQ Pretest Factor Loadings ....................................................... 52
Table 3.3 — Correlations Between Pretest Competitiveness and Control Trials ............ 60

viii

LIST OF FIGURES

Figure 2.1 -— Presentation of Priming Stimuli in Study 1 ...................................... 27
Figure 2.2 — Pretest Competitiveness Scores on Social Values Measures for Males
in Study 1 ............................................................................. 28
Figure 2.3 - Pretest Competitiveness Scores on Social Values Measures for Females
in Study 1 ............................................................................. 29
Figure 3.1 — Motivation Gain for Trial 3 by Condition ....................................... 57
Figure 3.2 — Motivation Gain for Trial 4 by Condition ....................................... 58

Chapter 1

INTRODUCTION

Competitiveness is a strong motivational force that can guide behavior.
Competition is present in our society across a broad spectrum of domains. As a
motivation that energizes and channels behavior, it can be a detemiining factor in
people’s level of achievement in life (Jackson, Ahmed, & Heapy, 1976). Historically,
this concept has been studied as both a personality variable and as a temporary
psychological state in persons that is triggered by the demands of certain situations. In
this thesis, I examined competitiveness in both of these ways. There has been
surprisingly little research on the motivational properties of competitiveness. Some of
the reasons for these gaps in the literature are discussed below.

One particular area of psychology in which competitiveness is likely to exert
significant effects is task performance. Research on task performance historically has
looked at what aspects of the situation are related to performance on particular tasks.
This tradition in social psychology dates all the way back to Triplett’s research on social
facilitation (Triplett, 1898). While there have been many investigations of personality
factors that influence task performance, few have examined competitiveness in particular.

Before we research competitiveness, it is necessary to first define it. The first
notable definition of competitiveness in the psychology literature is, “an individual’s
tendency to approach or avoid a competitive situation” (Martens, 1975). Martens went
on further to describe competitiveness as “a social comparison process and a situation

specific form of achievement motivation. Competitiveness originates in the intrinsic

 

motivation to be competent but competitiveness may also be extrinsically motivated”
(Martens, 1976). So we see by these definitions that Martens defined competitiveness as
a cognitive process that is activated in particular situations. Although achievement
motivation need not be conceptualized as a personality variable, Martens described it as
such. Competitiveness traverses the domains of cognitive, social, and personality
psychology. Because of the broad characterization of this concept, it has been
operationalized and studied in difl‘erent ways.

Trait Competitiveness

Studies that treat competitiveness exclusively as a trait have been sparse in social
and personality psychology. However, some work from this perspective has been done in
sport psychology. For instance, Jones, Neuman, Altmann, and Dreschler (2001)
developed the Sports Performance Inventory (SP1), which attempted to measure athletic
potential. Based on an exploratory factor analysis of 83 self-report questionnaire items,
six factors were extracted. Competitiveness was among them. This subscale exhibited
acceptable reliability, Cronbach’s alpha = .95. It was also able to discriminate between
college varsity athletes and novice athletes. Since college athletes scored significantly
higher than novice athletes did on this subscale, there is good evidence of concurrent
validity.

Recent research by Hellandsig (1998) has shown that competitiveness can predict
outcomes of interest in sport psychology. In this research a competitiveness subscale of a
measure called the Sport Orientation Questionnaire (Gill, Dzewaltowski, & Deeter, 198 8)
was used to measure differences in adolescent athletes. The Sports Orientation

Questionnaire (SOQ) is a self-report instrument that measures, among other things, trait

competitiveness. High levels of competitiveness, as measured by this method, predicted
high levels of performance in a variety of sports. This evidence supports the predictive
validity of the scale. Physical competence attributed to self and teammates also predicted
high performance in all the sports measured. So while competitiveness is not the only
predictor of performance at sports, it did have a significant connection to achievement in
this domain. These results as a whole support the hypothesis that competitiveness can be
operationalized as a trait or individual difl‘erence variable that has implications for
outcomes in sports.

As shown by the studies presented above trait competitiveness, as assessed by
self-report, is related to performance at team sports. However, to truly capture a
personality trait, it is necessary to measure it via variety of methods. Some research has
conceptualized competitiveness as a personality trait but measured it by different
methods than above. For example, Higgs (1972a) measured competitiveness of
participants by observer report. Observers classified participants as good or average
competitors based on their behavior in three game situations. Conditions were controlled
such that participants ran side-by-side with another person on a treadmill whose previous
performance was approximately equal to their own. In another condition participants ran
alone on a treadmill. Those judged to be high in competitiveness by observer ratings,
persisted longer than those judged to be average in competitiveness. This was true across
conditions. Persistence was measured in relation to the participant’s own ability on an
oxygen uptake trial conducted before the experimental conditions were imposed. The
results of this study support the construct validity of the author’s measure. That is, those

who are judged to be more competitive should outperform those judged to be less so and

the evidence was consistent with this prediction. However, the author did not include a
detailed description of the observer ratings used to judge competitiveness. In order to
assess the validity of this measure and the other self-report measures, they should be
compared in future research. Another issue with this study is that competitiveness was
also treated as a psychological state induced by the demands of the situation.
Competitiveness conceptualized as a state and as a person-situation interaction is
discussed in subsequent sections of this paper.

Other research has not only treated competitiveness as a personality trait, but has
tried to link it to other traits. For example, Higgs (1972b) measured participants’
competitiveness by observer report and then subjected them to a battery of psychological
and motor ability tests. However, there were no significant differences in personality, as
assessed by this procedure, between those judged to be high or average in
competitiveness. The only differences were in motor ability. These findings suggest that
competitiveness may be linked to physiology rather than other personality traits.
However, the limited sample size and methods of assessing competitiveness and
personality leave the results of this study in doubt.

Frederick (2000) also tried to link competitiveness to other personality variables,
as well as to performance measures. The results of this study showed that
competitiveness was negatively correlated to both internal locus of control and GPA. In
summary there have been some attempts to link competitiveness to personality traits.
Competitiveness has predicted outcomes of interest such as performance at some athletic
tasks. However there is no evidence of a distinctive personality profile that is linked to

competitiveness. The use of single methods and small samples in this research also lead

me to question its results. In order to truly assess the affects of competitiveness on
performance, it must be studied from not only a variety of methods but across different
situations. As a reasonable next step, it appears usefirl to explore the effects of
competitiveness on performance, not only among athletes in athletic contexts but also to
other contexts as well for purposes of generalizeability.

State Competitiveness

Past research has operationalized competitiveness as temporary and changeable
aspect of particular situations. These studies have manipulated the enviromnent so that
participants would behave in a more or less competitive manner. This research has used
varied methods to determine how a competitive environment affects behavior. While
most of this research has been performed in the area of social psychology, some also
comes from sport psychology.

Some studies have measured competitiveness along with social loafing. Social
loafing is a phenomenon that occurs in small performance groups. Specifically, when
group size increases individuals tend to exert less efl‘ort. Stroope (1993) found that social
loafing was eliminated in a group rowing task when individual performance was
identifiable to the group. That is, when individual split times were publicly displayed
during a rowing task, participants did not show decrements in their performance
compared to when they rowed alone. Furthermore, participants felt more competitive
when their split times were identifiable and they were in a group.

This effect of the competitive environment on performance is not limited to one
sport or to the laboratory. A field study of swimmers obtained similar results (W illiarns,

Nida, Baca, & Latané, 1989). This study measured swimmers at a Big Ten

Championship meet. Those athletes who competed in both individual and relay events of
the same stroke and distance were examined. Once again, the results showed that social
loafing was eliminated in identifiable, group situations. However, in this case, the
swimmers not only failed to decrease performance when identifiable in a group, but
actually increased their performance under these circumstances. These swimmers also
felt significantly more competitive when in a relay situation, as rated by self-report on
questionnaires. The researchers then went on to manipulate identifiability and individual
or relay event. In this case, results show that swimmers were less likely to exhibit social
loafing on relays when their performances were publicly identifiable. Social loafing did
occur when their individual performances on a relay were not identifiable to the group.

Taken together these studies show strong support for the idea that particular
situations can make individuals feel and behave more competitively. It may be the case
that the factors listed above can eliminate social loafing in work groups, committees, and
legislative bodies. These are only speculations though, because these experiments have
not been done within those contexts.

It seems then, that the presence of others and identifiability are key components of
the social environment that lead to the elimination of social loafing. These factors, in
turn, are positively related to self-reported feelings of competitiveness. However, it is
unclear fiom the results of this research whether feelings of competitiveness led to the
elimination of social loafing or vice-versa. It may also be the case that other’s presence,
identifiability, or a combination of the two leads to both increased performance and
competitiveness. Other research by Williams, Harkins, and Latané (1981) demonstrated

that identifiability led to the elimination of social loafing at a group cheering task. So

while the role of identifiability in the elimination of social loafing has been demonstrated,
the direct role of competitiveness has not. However it has been established that self-
reported competitiveness is related to the elimination of social loafing for certain group
tasks. The key concept to take from these studies for the current research is that
competitive contexts such as athletic competitions lead to increased reported feelings of
competitiveness.

It is reasonable to assume that working in a group at a task in an athletic
environment, as mentioned above, would lead to an increase in self-reported
competitiveness. In this situation, a person is presented with the opportunity to compare
his or her skills to others publicly for the potential benefit of the team. The added
component of identifiability makes it possible for him or her to compare his or her own
skills to those of others. Identifiability should also motivate the person to avoid
performing poorly and thus hurt the group’s chances for success.

While such an opportunity for public social comparison should lead to increases
in competitiveness, it is also logical that competitiveness should lead to better
performance in a group task. First, those who come into a group task and perceive it as
an opportunity to compete will be more willing and able to perform with and against
others. These perceptions should motivate the person to perform better than when alone.
The perceived demands of the situation should get one’s competitive juices flowing in
anticipation of the task at hand. Activating one’s competitiveness should lead to
performance enhancement at a group task. However, since people perceive situations
differently, we can cannot assume that they would see a group task as an opportunity to

compete. By manipulating one’s perceptions of the environment, then testing their

performance at group tasks, we can explore more cleanly the extent to which
competitiveness leads to performance enhancement.

Other studies from education and social psychology provide a better
understanding of the nature of the environment’s influence on competitiveness. For
example, Aronson and Osherow (1980) created an interdependent, cooperative classroom
environment in which students took turns at tasks within 5-6 person groups. The results
showed that this cooperative environment had several positive outcomes for minority
students but not white students when compared to a normal classroom. In addition,
feelings of competitiveness decreased. However, the classroom environment differs from
other environments in type of achievement, identifiability, and goals of individuals. So
we cannot be sure that results obtained in the classroom will generalize to the other
contexts. Secondly, this study characterizes cooperation and competition as antithetical
in the classroom. This may not be the case. In other areas they may be interdependent or
occur independently. For example, a leader of a group may be highly competitive with
others but in order to achieve his goals, he must cooperate with his teammates in order to
perform well. The two are examined separately in this paper.

Person-Situation Interaction Approaches to Competitiveness

Some researchers have attempted to integrate the trait and state approaches to
competitiveness. These researchers have measured individual differences in trait
competitiveness and then put participants in competitive situations to see if those who are
high in competitiveness differ from those low in competitiveness in terms of certain
behaviors and cognitions. Once again, some of these studies have been conducted in

social and personality psychology, while others are from sport psychology.

Within the field of social psychology, there has been some interactional research on
competitiveness. For example, Graziano, Hair, and Finch (1997) examined the role of
competitiveness and other personality traits on performance in group task situations.
Trait competitiveness was rated by both self-report and observer-report. These trait
ratings of competitiveness were negatively correlated to agreeableness and positively
correlated to task performance. The effect of agreeableness on task performance was
partially mediated by trait competitiveness. So, this research used multiple measures of
trait competitiveness to examine its role on task performance in competitive situations.
While this research focused more on the trait aspects of competitiveness, it did predict
performance in state competitive group tasks.

Other studies in marketing have shown that those high in trait competitiveness set
higher goals when they perceive the organizational climate as competitive. However,
those low in trait competitiveness set lower goals regardless of their perceptions about the
level of competitiveness of the organizational climate. These goals in turn are positively
related to performance (Brown, Cron, & Slocum, 1998). So the person-situation
interaction approach to competitiveness has implications in the world of business as well.

Other studies have used nonconscious, subliminal priming to attempt to
manipulate the salience of trait competitiveness. Subliminal priming involves pairing a
target concept with an unrelated, distracter concept. The target concept is presented
outside of conscious awareness but exerts effects on behavior or thoughts related to that
concept (Bargh & Chartrand, 2000). Priming procedures and background will be

explained in more depth later in this paper.

Priming the concept of competitiveness seems to interact with trait
competitiveness in particular situations. For example, Neuberg (1988) subliminally
primed participants with either competitive or neutral words. The researcher also
measured level of trait competitiveness. Results showed that those high in trait
competitiveness behaved more competitively in a prisoner’s dilemma game when primed
for competitiveness than when exposed to neutral primes. Also, among all those who
were primed for competitiveness, the participants highest in trait competitiveness
behaved more competitively than those low in trait competitiveness. A prisoner’s
dilemma game allows the participant to interact with other group members toward either
a common or individual goal. Those who worked toward the individual goal and
disregarded the group goal were scored as competitive. So we see that competitiveness is
once again conceptualized as the polar opposite of cooperation. Be that as it may, this
study shows that individual differences in competitiveness can be influenced by
unconscious primes in particular situations.

These studies show that competitive behavior was changed as a function of the
interaction between personality characteristics and the environment. Neuberg’s study
suggests that subliminal priming can change people’s behavior in a mixed-motive
situation in which competitiveness is one of a number of possible orientations. However,
subliminal priming interacts with participants’ trait competitiveness such that those who
are more competitive to begin with will behave more competitively. The present research
examined participants’ trait competitiveness, the effectiveness of supralirninal priming,
and their task performance in different group situations. As such it in part conceptually

replicated Neuberg’s work, and it also generalized his results to other tasks. Neuberg

10

found that both person factors and situational factors affected competitive behavior and
its relation to priming. The interaction between these two sets of factors also exerted an
influence on behavior. Therefore, I examined how participants’ trait competitiveness is
related to both the effectiveness of priming and their performance at different group
tasks.
Gender Differences

The results of much of the previous research show that there are gender
differences in competitiveness. Most show that men are more competitive than women
are in general. Most of these gender differences have been tested using self-report
measures of trait competitiveness. Several studies have shown that men score higher on
the competitiveness subscale of the SOQ (Gill, Kelley, Martin, & Caruso, 1991; Ruan,
1993)

Other self-report measures show mixed results. For example, a study using the
SP1 found that while males were more competitive that females in a novice athlete group,
the reverse was true of college athletes (Jones, et al., 2001). Another study found that
among 155 professional tennis players, females were significantly more competitive on a
trait measure (Houston, Carter, Smither, 1997). A study of Taiwanese swimmers after
participation in a tournament found that females were more competitive than males on
the SOQ (Lee, 1997). Of course after a competition we might expect that people will
report feeling competitive but it is not clear why women would feel more competitive
than men. This reversal may only be the case for these small, restricted samples of
professional tennis players and Taiwanese swimmers, but not for athletes in general. So

as a whole, trait competitiveness measures have shown that males tend to be more

11

 

competitive than females in general. Results that show females as more competitive had
small samples of high level athletes which are not typical of females in general while the
males samples were larger and more representative. Although the results have not been
unanimous, there seems to be reasonable support for this gender difference.

The potential causes for this gender difference are varied and interesting. They
could arise from differences in socialization, biology, a combination of both, cognitive
strategies, and/or a host of other reasons. Previous research has suggested that gender
differences in competitiveness are not due to biological sex, but rather to differences in
social roles (Olds & Shaver, 1980). This research has shown that the more masculine a
person is, as measured by the Personal Attributes Questionnaire (Spence & Helrnreich,
1979), the more competitive he or she is regardless of his or her biological sex. Other
research has shown that the gender difference may exist because females have a stronger
desire to maintain equitable outcomes based on socialized difl‘erences in submissiveness
and dominance between the genders (Wyer & Malinowski, 1972). So it seems that this
gender difl'erence may be learned rather than biologically based. While I acknowledge
that this gender difference likely exists, the causes will not be discussed at length here.
Future research needs to investigate this issue.

Evidence of this gender difference in competitiveness is also found in state
competitiveness and the interaction between trait and state competitiveness. For
example, Kline and Sell (1996) looked at trait competitiveness across group tasks of
varying degrees of competitiveness. Groups were composed of all highly competitive
individuals, all highly cooperative individuals, and mixed groups. These groups were

then put into competitive or cooperative situations. There was no gender difference in

12

performance at these tasks. So females performed just as well as males at competitive
tasks regardless of level of trait competitiveness. Unfortunately, once again
competitiveness and cooperation are separated and operationalized as opposites in this
experiment. As a result we do not see how they can jointly affect performance such as in
team sports where one must 000perate with one’s team and compete against other teams
and individuals in order to achieve success. There was support for the idea that women
will perform just as well as men in competitive tasks even if there is a gender difference
in trait competitiveness. Unfortunately, the authors do not mention whether there were
gender differences in trait competitiveness in this sample.

Previous research has shown that trait competitiveness is activated only in
particular situations and that this trait-situation link varies across genders. Results of a
previous study show that when participants work at a group task and they believe their
input is less than their partner’s input they will allocate the same amount of reward to
each member of the group. There is not a gender difference in this condition. However,
when participants believe their input to the group is greater than the input of their partner,
males tend to give more reward to themselves and females still give the same amount of
reward to their partner as to themselves (Carles & Carver, 1979). This change in strategy
can be interpreted as an increase in competitiveness in males when they believe they are
helping the group more than other members are. It could also be the case that females are
more likely to adhere to fairness norms. Either way this brings up an interesting point.
That is that competitiveness can be conceptualized and measured in an interpersonal
context not just in sports. So in this case competitiveness is not defined as desire to

compare one’s skills to those of an opponent but rather to defeat an opponent by

13

allocating more resources to oneself than to others. This desire to gain an advantage over
one’s opponent is part and parcel of the essence of competitiveness. By seeking an
advantage over an opponent or using social comparison to evaluate one’s performance, a
person still seeks to fulfill the same goal of performing one’s best in relation to others and
to one’s own standards of excellence. For measurement purposes competitiveness can be
measured as one’s motivation to succeed and one’s willingness to gain advantage relative
to others because this behavior serves the goal of performing one’s best.

Another study that examined gender differences in competitiveness across
situations was conducted with children. By using observer report, researchers assessed
children’s competitiveness when worldng at a task with same sex partners or opposite sex
partners. This study found that children in preschool only behaved competitively when
they worked with a member of the opposite sex. This was true of both boys and girls.
However, when children were a little older (7-9), girls only behaved competitively when
performing a task with the opposite sex while boys were competitive when working with
the same or opposite sex (Moely, Skarin, & Wei], 1979). The authors believe that males
develop a more general tendency to compete across situations, while females compete
only in particular situations. In certain situations women may be just as competitive as
men, but men appear to be more competitive in general across situations. So how this
gender difference in trait competitiveness manifests in behavior and under what
circumstances is still largely an open question. Past research has looked at these
differences from a variety of perspectives and measures. This research suggests that

while gender difl‘erences in competitiveness exist, they do not exist in all situations.

14

Tasks

I measured the effects of competitiveness on performance at a persistence task. I
collaborated on prior research on the Kohler motivation gain efl‘ect which has used
persistence tasks to explore performance (Hertel, Kerr, & Messé, 2000; Kerr, Messé,
Park, & Sambolec, in preparation). Using a persistence task as opposed to other types of
cognitive or physical tasks has several advantages. First of all, few participants have
experience with persistence tasks, such as holding up one’s arm as long as possible. So it
is unlikely that certain individuals will come into the laboratory with known advantages
over others at these tasks. Second they are relatively non-evaluative. Failure or success
at the task says little about a participant’s intrinsic qualities such as personality or
intelligence. Performance at cognitive or physical tasks often causes participants to make
attributions about their abilities and worth.

Previous research has shown that performance at persistence tasks can be
influenced by factors such as the presence of others (Triplett, 1898), relative ability
difference between self and co-worker (Messé, Hertel, Kerr, Lount, & Park, 2002), and
group efficiency (Kohler, 1926, 1927). These studies have focused on the factors that
facilitate performance enhancement in small groups. This concept of individual
performance improvement in a team setting is referred to as a group motivation gain.
That is, a motivation gain occurs when an individual’s performance in a group is greater
than his/her performance when working at the same task alone. So the individual
exceeds expectations that would be predicted by his or her solitary performance when
taking into account any losses in performance due to physical demands of the group task,

i.e., loss of time at a team rowing task when one member’s car hits another’s (Steiner,

15

 

1972). Previous research in motivation gains has placed participants in contexts in which
they worked with and/or competed against other individuals or teams (e.g., Hertel, Kerr,
& Messé, 1999; Williams & Karau, 1991). However, very little research in this area has
looked at individual differences that may moderate performance. It seems that
competitiveness is a likely candidate for a personality moderator in this type of research.

Competitiveness may be of particular importance within certain group
performance contexts. The first is at a conjunctive task. In such a task, each member of
a group performs the same function and the group’s success is determined by the weakest
member (Steiner, 1972). Some past work investigating conjunctive tasks manipulated
co-workers’ supposed ability so that participants were always the least capable members
of the group (Hertel, et al., 1999, Study 2; Messé, et al., 2002, Study 2). The participant’s
performance is also essential to the group’s performance because his or her score is the
one that all members of the groups will receive. Research has shown that when a person
is placed in such a situation, he or she often shows motivation gains (Kohler, 1926, 1927;
Hertel, et al., 1999; Messé et al., 2002). A highly competitive person or one placed in a
highly competitive context may be more motivated to beat the more capable member(s)
of the groups and refirse to quit first. In the process the competitive person maximizes
benefit for the group unintentionally. Therefore, competitiveness brought into the
situation by priming or initial predisposition may increase performance under conjunctive
conditions. However, because the assumed goal of conjunctive tasks is to help the tearn’s
performance, it could also be that the desire to beat one’s opponent may not be as salient
in this performance context and heightened competitiveness may not increase

performance.

16

In contrast, it is more likely the case that the task demands of a coactive group
situation might well be afl‘ected by competitiveness. In such a situation participants are
able to simply monitor one another’s performance. There is no group goal or outcome.
They are given no specific instructions on how they should behave toward this other
person but are told that their own score is simply the outcome of their own performance.
Under these circumstances a highly competitive person, as a function of either
personality or environment, will be more likely to see this interpersonal configuration as
an opportunity to compete, by gauging his or her own performance against that of other
group member(s). In this situation, it is the presence of another which allows the
opportunity to evaluate one’s own performance relative to this person and one’s own
standards that will motivate the participant to enhance his/her performance. Additionally
those participants who are high in trait competitiveness would interpret this situation as a
chance to beat an opponent thereby pushing their own performance. Once again,
competitiveness, either brought in to the lab as a predisposition or induced by priming
will lead to motivation gains under coactive task conditions.

There has been little research on the relationships between competitiveness, task
performance, and mood. However, it seems plausible that those whose personality
complements their situation should experience a better mood than those whose
personality and situation are incompatible. Since, this is a person-situation analysis of
behavior, mood was measured to examine this relationship. Positive and negative afi‘ect
are of primary interest so the PANAS (Watson, Clark, & Tellegen, 1988) was included as

a mood measure.

17

Hypotheses

In Study 1, I tested the effectiveness of a priming manipulation, as well as
examined gender differences in competitiveness. Based on previous research, I expected
that scores on trait competitiveness, measured by a variety of measures including the
SOQ competitiveness subscale (Gill, Dzewaltowski, & Deeter, 1988), the
Competitiveness Index (CI; Smither & Houston, 1992), and a Measure of Social Values
(Van Lange, Otten, De Bruin, & Joireman, 1997) would be higher for those who are
primed for competitiveness in comparison to those who are not (N euberg, 1988). The
SOQ and the CI both purport to measure competitiveness as a motivation to perform well
in athletic contexts in order to beat one’s opponents and/or to live up to one’s own
standards of excellence. The Measure of Social Values operationalizes competitiveness
as a gaining advantage over others by allocating the greatest possible resources to
oneself. Each measures a different competitive behavior but both are motivated by the
same underlying concept so their ratings should be highly consistent with one another.

Also as shown by previous research, there are gender differences in trait
competitiveness (e. g., Gill, et al., 1991). However, previous research has found no
gender difference in efl‘ectiveness of priming competitiveness (N euberg, 1988). So, I
predicted that males would score higher than females on initial trait competitiveness. All
participants were measured on trait competitiveness before and after the priming
procedure to test the effects of the priming. Although I expected males to be higher on
the pretest measure of trait competitiveness, I had no apriori hypothesis on gender
differences on the posttest measure. However, there was no reason to believe there

should have been any post-test gender differences beyond that found on the pretest. It

18

was unclear as to whether the prime would be equally effective for both sexes. So results
explored possible gender differences in priming effects, but no specific apriori
hypotheses about gender as a moderator of priming were formulated.

In Study 2, I tested the effects of priming on task performance. Based on
previous research on the link between trait competitiveness and performance, I expected
that those who were higher on trait competitiveness would outperform those lower in trait
competitiveness (Higgs, 1972a). I also expected that this difference would occur in both
group and individual conditions. However, I also proposed that these individual
variations in trait competitiveness could be overshadowed by nonconscious priming. So,
I predicted that priming would increase performance at tasks, particularly in interpersonal
conditions, regardless of initial level of trait competitiveness. To test this hypothesis, half
of the participants were primed for competitiveness after they had been pretested for trait
competitiveness, the other halfwere not. I expected that trait competitiveness scores, as
measured by the SOQ competitiveness subscale, would predict performance in the
control group but not in the primed group. As noted above, I also hypothesized that
males, in general, would score higher in trait competitiveness than would females. Once
again, I proposed no hypotheses with regard to gender difl’erences on the efl’ectiveness of
priming. I did hypothesize gender difl‘erences across situations. In general males’
performance should have been enhanced by level of competitiveness across situations,
while females’ competitiveness should have only enhanced performance in competitive
situations. That is, males should have enhanced performance in all group conditions
when primed while females should have only exhibited performance enhancement in the

coactor condition when primed. Previous research has shown that males behave more

19

competitively in general while females behave competitively in competitive situations
(Jones, et al., 2001).

I predicted that task setting and demands would exert a significant influence on
performance. I expected to replicate the Kbhler effect (Kohler, 1926; 1927). That is, I
expected people working under conjunctive task demands would outperform those in the
individual control condition. I also expected that those who were primed for
competitiveness would exhibit enhanced performance in the coactive and conjunctive
task demands conditions. Those in the coactive conditions who were primed should have
particularly outperformed those in the individual condition because they had been primed
and were in a situation that could be interpreted as competitive. Those in the conjunctive
condition should have outperformed those in the individual condition. However, priming
might not have greatly enhanced performance for those in the conjunctive condition
because they were not instructed to defeat their partner but rather work with him. So I
expected that priming effects would overshadow task demands in the conjunctive
condition and thereby moderate performance but not as much as in the coactive
condition.

The Positive and Negative Affect Schedule (PAN AS, Watson, Clark, & Tellegen,
1988) was included in Study 2 as a measure of mood. I expected that those whose
priming and task demands conditions most closely matched their pretest level of
competitiveness would have exhibited the highest positive and the lowest negative affect.
So I expected those who were highly competitive, primed for competitiveness, and who
were allowed to compete, i.e., those in the coactive (and perhaps, conjunctive) condition,

would have left the study with a more positive mood than those who were low in

20

competitiveness. Participants low in trait competitiveness would have exhibited less

positive affect if primed and forced to compete (group conditions).

21

Chapter 2

STUDY 1

£291.14

Participants

One hundred seventy-four undergraduate psychology students from Michigan
State University participated in exchange for partial fulfillment of course requirements.
Participants were recruited into the study using the university’s subject pool website.
Forty-eight were male and one hundred twenty-six were female.
Measures

Participants completed the Measure of Social Values as a pre- and posttest
measure of competition and cooperation (Van Lange, Otten, De Bruin, & J oireman,
1997). This measure consists of nine items. Participants are presented with a
hypotheticm situation in which they are asked to divide points among themselves and
another person. Respondents have three forced choice options of how to divide these
points. One option divides the points evenly and is scored as a prosocial choice. The two
other options provide a discrepancy between the participant’s and the other’s points such
that the participant gets more points. The option that provides for a moderate
discrepancy is scored as an individualistic choice and the large discrepancy is scored as a
competitive choice. There was no option which allowed participants to receive less than
the other. So a participant receives three scores for this measure. If a participant makes

six of these nine choices in one particular category then he or she is classified as a

22

member of this category (prosocial, individualistic, or competitive). To see the fill]
measure, refer to Appendix A.

Participants also completed two measures of competitiveness that were
administered as posttests only. The first was the Competitiveness Index (CI; Smither &
Houston, 1992). It consists of 20 statements to which a participant can choose to either
agree or disagree. Items include statements such as “I am a competitive person” and “I
hate to lose.” For the full measure, refer to Appendix B.

Participants also filled out the Sports Orientation Questionnaire (SOQ; Gill,
Dzewaltowski, & Deeter, 1988). This 25-item measure allows participants to rate their
agreement on a 5-point scale fiom “strongly agree” to “strongly disagree.” The SOQ
includes three subscales for competitiveness, goal orientation, and win orientation.
Competitiveness items included: “I perform my best when I am competing against an
opponent.” Goal orientation items included: “I try my hardest to win when I have a
specific goal.” Win orientation items included: “I hate to lose.” For a firll version of the
SOQ refer to Appendix C.

Two different types of supralirninal priming manipulations were examined, a
word search task and a story. These were further divided into competitive, neutral, and
cooperative conditions to create a 2 X 3 between subjects factorial design. The word task
priming manipulation was based on techniques shown to be effective in past research
(Bargh & Chartrand, 2000). This manipulation was a list of 25 words, 15 of them are
descriptors of competition or cooperation (target words). The other 10 were neutral
distracter words (see Appendices D & E). In the neutral or no prime condition,

participants were given a list of 25 neutral words (see Appendix F). The target and

23

distracter words were selected by research assistants within the Small Groups Lab at
Michigan State University. Final selections for inclusions of these words were made by
the author and one of his advisors (Messé). Examples of competitive words are:
competition, victorious, rivalry, and achievement. Examples of cooperative words are:
connection, partnership, and support. Examples of neutral words are: glassware, edifice,
stowaway, and circumlocution. According to the directions, the participants, in 15
minutes, had to find as many words as they could that are three letters or longer hidden
within these given words. They were instructed not to unscramble words, but only to
find smaller words that existed within the given order of letters of the stimulus word.
That is all the three letter or longer words had to be contiguous letters already contained
in the stimulus words.

The story priming manipulation consisted of three different stories that were as
similar as possible to one another in context, while still representing different ideas. All
three described a running race. The competitive story told of a solitary runner who
endured years of hard work and maintained an extremely competitive outlook to beat his
competition and qualify for the Olympics. The cooperative story described a team of
athletes running a relay at the World Track and Field Championships who valued their
common bond above all and succeeding because of it. The neutral story described a day
in the life of a marathon race oflicial. It told of his duties to prepare for and clean up the
course. It made no mention of competition or cooperation. The stories were developed
by members of the small groups lab and then edited by the researcher and one of his

advisors (Messé). For full versions of these stories refer to Appendices (G-I).

24

 

Procedure

In order to participate in Study 1, participants had to complete the Social Values
Measure (SVM) as a pretest questionnaire. This measure was posted on the university’s
psychology subject pool website as a “global questionnaire.” These questionnaires allow
researchers to collect data on students without linking themselves to any particular study.
These questionnaires are listed on the website when students logon to it. They were told
that by filling out these questionnaires they could gain opportunities to participate in
studies that would allow them additional chances at course credit. Approximately 500
students filled out the SVM on the web.

The experimenter then contacted all of the students who had filled out the SVM
correctly by e-mail and invited them participate in the “Puzzling Words Study.”
Experimental sessions consisted of six to thirty students in a large lecture hall. When
participants arrived at a session, an experimenter welcomed them in the hallway outside
the classroom and asked them to sign their names on a log sheet. Another experimenter
brought participants into the classroom and sat them with at least one space between each
other. The experimenter then described the basic purpose and procedure of the study. He
told the participants that the study was intended to measure their performance on a
variety of tasks. He went on further to say that not all participants would have the same
tasks but they would all take approximately the same amount of time. They could leave
after 30 minutes if they had finished and were allowed as much as one hour. The lower
time limit allowed most people to finish and prevented them from distracting those who

had not. It also compelled them to spend a reasonable amount of time on the tasks.

25

The experimenter then passed out packets containing a consent form, the priming
manipulation, and the measures. The priming manipulation was titled a measure of
verbal and creative abilities. Participants first filled out the neutral version of the
opposite condition prime. That is, if someone was in the competitive word task prime
condition, he or she would first fill out the neutral story prime. If one was in the
cooperative story prime condition, he or she would first fill out the neutral word task
prime. This procedure is presented in Figure 2.1. So in the event a participant looked
over at the neighboring participant he or she would see an identical task. This was
implemented to reduce potential suspicion that some participants had different tasks.
Also there was no reason to believe that the neutral prime would affect scores on the
dependent variables so its inclusion before the target prime should not have influenced

the other prime types effects.

26

Word Prime

 

 

 

 

NeutrAPrinre (Control Competitive Prime
Neutral Story Neutral Story
Neutral Words Competitive Words
Stag Prime

Neutral Prime (Control) Competitive Prime
Neutral Words Neutral Words
Neutral Story Competitive Story

Figure 2.1. Presentation of Priming Stimuli in Study 1

Ram
Malleability of Social Values.
In order to assess the effectiveness of priming competitiveness a pilot test was
conducted measuring participants’ base rate of competitiveness on the Social Values

Measure (SVM; Van Lange, et al., 1997). After participants completed the priming

27

 

procedure, I measured their responses on this scale once more. In order to be classified
on this scale participants must make six of nine consistent choices. Results on the pretest
showed that of the 124 participants who filled out the Social Value Measure in the
sample, 16 (12.9%) were classified as competitive, 29 (23.4%) as individualistic and 59
(47.6%) as prosocial. This left 20 participants (16.1%) who could not be classified into
any of the three categories. Twenty-seven of the participants who filled out the SVM
were male. Four (14.8%) were prosocial, five (18.5%) individualistic, and twelve
(44.4%) prosocial. The rest (6) could not be classified. Ninety-seven females filled out
this measure. Of these 12 (12.4%) were competitive, 24 (24.7%) individualistic, and 47
(48.5%) prosocial. This left 14 females who could not be classified. For the firll
distributions of competitiveness scores on this measure by gender, refer to Figures 2.2

and 2.3.

 

Frequency

 

 

 

 

Figt_rre 2.2. Pretest Competitiveness Scores on Social Values Measures for Males in
Study 1

28

 

so
70
so
so
40
so
20
10
o
§p '\ ‘Io '5 Ft 93 ‘b ’\ ‘b $9
6“)“ of

00 9‘00

‘5.
e"? a?

l Females

Frequency

 

 

 

 

Figure 2.3 Pretest Competitiveness Scores on Social Values Measure for Females in
Study 1

Results showed that the mean pretest scores on the SVM for all those who
participated was 1.47 on competitive items, 2.77 on individualist items, and 4.76 on
prosocial items. The possible scores on each subscale range fi'om 0-9. Males mean
pretest scores on the competitive, individualistic, and prosocial items were M = 1.78, M =
2.74, and M = 4.48 respectively. Females mean scores on the same items were M = 1.38,
M = 2.78, and M = 4.84 respectively. These gender differences were nonsignificant (all
F’s < 1).
Posttest Measures of Competitiveness.

Participants also completed the SOQ (Gill, Dzewaltowski, & Deeter, 1988) and

CI (Smither & Houston, 1992) as posttest measures. The SOQ achieved acceptable level

29

of reliability, Cronbach’s alpha = .95. The competitiveness subscale of the SOQ also
achieved a high level of reliability, Cronbach’s alpha = .94. as did the CI, Cronbach’s
alpha = .84. The original factor structure of the SOQ was also replicated such that items

loaded on appropriate subscales. For a complete list of factor loadings refer to Table 2.1.

Table 2.1

Factor Loadings for the SOQ in Study 1

 

 

 

Factor
Item Competitiveness Win Goal Unpredicted
1. I am a determined competitor .579 .134 .251 .602
2. Winning is important .376 .479 .006 .563
3. I am a competitive person .512 .241 .230 .612
4. I set goals for myself when .330 .222 .535 .512
I compete
5. I try my hardest to win .292 .292 .454 .604
6. Scoring more points than my .417 .532 .000 .443
my opponent is very important
to me
7. I look forward to competing .773 .007 .180 .465
8. I am most competitive when .197 .009 .745 .192
I try to achieve personal goals
9. I enjoy competing against .772 .009 .222 .394
others
10. I hate to lose .177 .770 .004 .279
11. I thrive on competition .747 .240 .189 .303
12. I try my hardest to win when .254 .166 .684 .279
I have a specific goal
13. My goal is to be the best .558 .387 .288 .219
athlete possible
14. The only time I am satisfied .180 .754 .009 -.007
is when I win
15. I want to be successful in .594 .448 .174 .206
sports

30

16. Performing to the best of my .009 .167 .468 .605

ability is very important to me

17. I work hard to be successful .646 .350 .276 .168
in sports

18. Losing upsets me .109 .783 .124 .251

19. The best test of my ability is .766 .282 .158 -.003
competing against others

20. Reaching personal .164 .137 .780 .179

performance goals is very
goals is very important to me

21. I look forward to the .695 .138 .443 .209
opportunity to test my skills
in competition

22. I have the most fun when I win .322 .666 .284 .005

23. I perform my best when I am .758 .293 .176 .117
competing against an opponent

24. The best way to determine my .207 .009 .786 -.001
ability is to set a goal and try
to reach it

25. I want to be the best every .312 .548 .304 .290
time I compete

 

Note. The above table presents factor loadings based on a Principal Components Factor
Analysis with Varimax Rotation and Kaiser Normalization. The above is a Rotated
Component Matrix.

3 X 2 X 2 AN OVAs analyzing Prime (competitive, cooperative, or neutral),
Prime type (words or story), and Gender were performed on the posttest measures of
competitiveness: the CI, SOQ and SOQ competitiveness subscale. Results revealed a
nonsignificant Priming effect on the SOQ and it’s competitiveness subscale, F’s < 1.
Although this relationship was stronger when using the CI as the dependent variable, it
was still nonsignificant, F(2, 160) = 1.91, p < .16. The effect of prime type (word or
story) was marginally significant on the SOQ, F(1, 160) = 3.48, p < .07 such that those in
the word condition scored higher than those in the story condition. This effect was also
marginally significant for the competitiveness subscale, F(1, 160) = 3. 10, p < .09, but

nonsignificant for the CI, F < 1. The effect of Gender was significant on both the

31

competitiveness subscale, F(1, 160) = 4.46, p < .04 and the CI, F(1, 160) = 8.82, p < .005
such that males scored higher than females on these measures. This effect was
nonsignificant for the total SOQ, F(1, 160) = 1.52, p < .22.

Interactions terms were computed among the independent variables. Results
revealed no significant interactions. There was a marginally significant prime X prime
type interaction on the SOQ , F(2, 160) = 2.60, p < .08 as well as a marginally significant
prime X gender interaction, F (1, 160) = 3.42, p < .07. However these interactions did not
even reach marginally significance on the other measures of competitiveness. A
marginally significant prime X prime type X gender interaction was detected by both the
competitiveness subscale of the SOQ, F(2, 160) = 2.62, p < .08 and the CI, F(2,160) =
2.84, p < .07. This effect was not significant for the overall SOQ.

To examine the nature of the 3-way interaction contrasts were performed which
revealed that males primed for competitiveness by words scored marginally higher than
males primed by the neutral words on SOQ competitiveness subscale, t(160) = 1.64, p =
.10. This effect did not reach significance on the other measures of competitiveness (t’s
< 1.6). However when comparing the competitive prime versus the neutral prime for
males primed by the story method and for female primed by either method, we see that
competitiveness scores are opposite of the predicted direction. That is under these
conditions, competitiveness priming made participants score lower on measures of
competitiveness. For all means refer to Table 2.2. This suggests that competitiveness
priming may only have worked for males in the word task condition but not males on the
story condition or for females using either priming stimulus. These results also suggest

that the SOQ, its competitiveness subscale and the C1 are measuring similar but not

32

identical concepts. They also suggest that priming was only selectively successful in

creating differences on trait measures of competitiveness.

Table 2.2

Scores on Competitiveness Measures After Priming by Condition

 

 

 

 

 

 

Prime
Competitive Neutral Cooperative
Prime Type
Measure Word Task Story Word Task Story Word Task Story
Males
SOQ 104 90.33 88.7 92.2 108.67 81
SOQ:
Competitiveness 55.5 46.67 45.3 48.2 57.11 42.11
Subscale
CI 15.5 11.83 12.8 14.2 15.83 15
Females
SOQ 82.2 93.05 91.38 86.68 96.3 90.05
SOQ:
Competitiveness 39.56 46.47 46.48 42.26 48.75 44.85
Subscale
CI 10.56 12 13.1 10.61 12.71 12.9

 

33

Note. Scores on the SOQ are totaled fiom all 25 items on the measure. Items are on a 5-
point scale. Higher scores indicate a higher sports orientation. Scores on the SOQ:
Competitiveness Subscale are totaled from 13 items on the SOQ. Items are on a 5-point
scale. Higher scores indicate higher competitiveness. Scores on the C1 are totaled from
all 20 items on the measure. Scores are dichotomous. Items are scored as l-point for
agreement or O-points for disagree. Higher scores indicate higher competitiveness.

ANOVA’s on the SVM revealed no significant effects (all F’s < 1.2) for any of
the three types of choices (competitive, individualist, or prosocial). Based of these
results, I decided to use the SOQ and its competitiveness subscale as pre and posttest
measures of competitiveness in the main experiment (Study 2). Since the word task
seemed to be a better method for priming competitiveness than the story task, I decided to
use the word task as the priming manipulation in the Study 2. Also since only males
responded to the competitive prime, Study 2 only examined males.

Discussion

The results showed that scores on the SVM were not affected by the priming
procedures. Of the 174 participants in the study only 124 completed the SVM as a
pretest measure. So we can only use these participants to measure differences caused by
the priming procedures. Results show that only 15 of these participants changed their
classification on this measure (competitive, individualistic, or prosocial) after the prime
was administered. Only six of these were in the predicted direction. That is only six
participants were measured as more competitive after the competitive prime or more
prosocial after the cooperative prime. Four were in the opposite direction indicating
possible reactance or ineffectiveness of the prime. The remaining five consisted of

changes in classification after the neutral prime. Three participants were measured as

more prosocial and two were more competitive. So it appears that social values as

34

 

measured by this instrument are either very difiicult to change or are relatively unaffected
by cooperative and competitive priming procedures. Perhaps this is because the
instrument really measures allocation of resources which is not related to one’s
motivation to beat others at a task.

There was a lack of significant results on this measure for prime, prime type, and
gender, indicating that it is not a useful procedure for purposes of exploring the target
concepts. However, results from the posttest competitiveness measures were more
promising. Males primed for competitiveness by the word task scored somewhat higher
on every posttest measure of competitiveness when compared to the neutral prime
condition. However this effect was only marginally significant on one measure (SOQ
competitiveness subscale) and did not reach statistical significance on the other two
measures.

Unexpectedly, the cooperative word prime made males just as competitive or
more so than the competitive prime, although not significantly more than the neutral
condition (see Table 2.2). The reasons for these trends are unclear. The cooperative
word prime (Appendix E) contains words such as “teamwork,” “camaraderie,” and
“squadron” which may bring to mind image of competitive sports for males. Other
words contained within the instrument which seem much more related to cooperation
such as “interconnected”, “collective”, and “support” may actually bring to mind
cooperation but cause reactance in males who are inclined to be more competitive. This
reactance may lead males to think or behave more competitively when primed for

cooperation by these words.

35

For males primed by the story method, the results show a different pattern. The
highest posttest competitiveness scores were obtained from the neutral priming condition,
then competitive, and finally cooperative on the SOQ and its competitiveness subscale.
For the CI, participants scored highest when primed for cooperation, then neutral, and
finally competitive. So most of the SOQ results were at least in the right direction for the
cooperative prime, although not significant but in the wrong direction on the CI. These
conflicting results may be due to the content of the story primes (Appendices G, H, and
I). All three describe a running race. The competitive story describes a man who runs
his best due to his competitive spirit. The cooperative story describes a group of
teammates who do their best and enjoy themselves because of their cooperation. The
neutral story just describes the days events of a road race oflicial but does not mention
any competitive or cooperative content. To males, who are more competitive than
females, the mention of a sporting context may trigger competitive juices, generally,
regardless of the context in which it is phrased. This may be why scores in the
cooperative prime condition were higher than expected.

For females, the competitive word task prime did not work at all. Females scored
lowest on all posttest measures of competitiveness in the competitive word task prime
condition when compared to both the neutral and cooperative prime condition, although
this effect was not significant. Females primed for competitiveness by the story prime
did score higher than those in the neutral condition on all three measures but not
significantly so. So the competitive story did not make females significantly more
competitive and the competitive word task prime actually makes them less competitive.

It appears then, that competitiveness priming is ineffective for females, at least when

36

measured by these instruments. When we look at the content of the competitive word
task prime, we see words such as ‘Victorious,” “vanquishing,” and “superiority” which
may not activate accessible concepts to most females. These words are related to
defeating others and hierarchies which may not be as salient to women as they are to
men.

Males were overall more competitive than females on the SOQ competitiveness
subscale and the CI, but contrary to expectations females did score higher than males in
the neutral word task condition on many of the measures, although none of these
differences even approached significance. It appears that the competitive word task
prime made males somewhat more competitive. This priming procedure did not work for
females. The competitive story prime worked for neither. As a result I decided to
eliminate the story prime in the subsequent study. I also decided to focus specifically on
males in Study 2 because overall the priming was ineffective for females. Since the
cooperative prime tended to yield higher competitiveness scores than the neutral
condition, it too was eliminated from Study 2. Competitiveness seemed more likely a
moderator of performance in group tasks because a desire to outperform one’s partner
should increase motivation while cooperation should not. This would be especially true
in the coactor condition in which a participant is not told to work with another person but
simply that he can monitor the other’s performance. Here one has the opportunity to
compete against another and competition has been made salient by the prime so priming
should moderate performance it the coactive condition.

It is less clear what competitiveness priming should do in the conjunctive

condition because beating one’s partner will not necessarily help the group which should

37

be the goal in a conjunctive task. Cooperation may improve the group’s performance
under these circumstances but since priming cooperation did not make people more
cooperative on the SVM and they actually appeared more competitive on the other
measures, cooperative priming was dropped. Ifthe competitive prime is enough to make
males perceive a group situation as competitive, then the prime’s effects should override
those of task demands and priming will moderate performance in the conjunctive
condition. If however, the cooperative nature of conjunctive task demands override the
salience of competitiveness, then priming will not moderate performance in the
conjunctive condition. However, priming was expected to increase performance in the
co-actor condition, in which concerns with relative performance could not possibly be

tempered by cooperative focus (e.g., a focus on team success, group cohesion, and so

forth).

38

Chapter 3

STUDY 2

M
Participants

One hundred twenty male undergraduate psychology students from Michigan
State University participated in exchange for partial fulfillment of course requirements. I
recruited only males for this study because the Study 1 showed that priming
competitiveness was more efl‘ective for males than females. Participants were recruited
into the study using the university’s subject pool website. This sample originally
included 132 students. However, 12 were excluded for explicit verbal suspicion
expressed to the experimenter after the study that there was no partner,
computer/equipment failure, misunderstanding of directions/language which prevented
completion of the priming procedure, and sleeping during the task.

Six of the aforementioned participants failed to complete at least 50% of the word
priming task and were excluded from analyses. Of these, two were in the neutral prime
individual condition, one in the neutral prime conjunctive condition, as well as one each
from the competitive prime individual, conjunctive, and coactor conditions. Two
participants told the experimenter they did not believe there was another participant in
their group and were therefore excluded from analyses. These were both given the
competitive prime treatment, one each in the coactor and conjunctive conditions. Two
participants in the competitive prime coactor condition were issued the wrong hand

weight by the experimenter and their data was excluded. One participant in the neutral

39

prime condition fell asleep during the session. Finally one participant in the neutral
prime individual condition experienced computer failure.
Apparatus and Measures

In order to measure participants’ trait competitiveness, I used the competitiveness
subscale of the Sports Orientation Questionnaire (Gill, Dzewaltowski, & Deeter, 1988).
The SOQ is self-report questionnaire in which participants rate their agreement with
items on a 5-point scale from strongly agree to strongly disagree. The SOQ contains 25
items divided into 3 subscales. To see the fill] SOQ, refer to Appendix C. The
competitiveness subscale consists of all the odd-numbered items. This subscale contains
13 items. It has demonstrated acceptable levels of internal consistency, reliability, and
validity (Gill, Dzewaltowski, & Deeter, 1988; Gill & Dzewaltowski, 1988) both in the
US and in Taiwan (Kang, Gill, Acevado, & Deeter, 1990).

As in Study 1, the priming manipulation was based on techniques shown to be
effective in past research (Bargh & Chartrand, 2000). Again, this manipulation was a list
of 25 words, 15 of which connote competition (target words). The other 10 are neutral
masking words (see Appendix D). In the neutral or no prime condition, participants were
given a list of 25 neutral words (see Appendix F). Examples of competitive words were:
competition, victorious, rivalry, and achievement. Examples of neutral words were:
glassware, edifice, stowaway, and circumlocution. As in Study 1, the participants had to
find as many words as they can that are three contiguous letters or longer hidden within
these given words in 15 minutes. They were instructed not to unscramble words, but

only to find smaller words that exist within the given ordering of letters. They were

40

given the word glassware as an example and told from this word they could extract words
such as glass, war, lass, ass, etc.

The performance task used was a persistence task based on previous research in
motivation gains (Kohler, 1926, 1927; Hertel, Kerr, & Messé, 1999). Participants were
instructed to hold their arms above a tripwire for as long as possible. Participants held a
4 lb. hand-weight so that the trials would tend to last only for a short period. Previous
research has shown that participants using comparable weights can hold their arms above
the tripwire for about 2.5 minutes on average (Hertel, Kerr, & Messé, 1999). Also based
on previous research, participants used a 41b. weight (Hertel, Kerr, & Messé, 1999). The
tripwire was extended between two pulleys and above a keyboard such that once the
participant’s arm fell on the tripwire, a spring-loaded lever attached to the pulley system
depresses the spacebar on the keyboard. This device is heretofore referred to as the
Weston machine, named after its designer, Eric Weston. During this task, the
participants were seated in fiont of a computer that recorded the time for each trial of the
persistence task. As soon as the spacebar was depressed from the dropping of the
participant’s arm, the clock stopped for the trial.

I administered the Positive and Negative Affect Schedule (PAN AS; Watson,
Clark, & Tellegen, 1988) as a measure of participants’ mood. The PANAS has exhibited
acceptable levels of reliability (Cronbach’s Alpha’s: .84 - .90; Watson, Clark, &
Tellegen, 1988). As stated earlier, I included this measure for exploratory reasons. I
suspected that participants whose initial state of competitiveness more closely matched

the state they were placed in the experiment would experience more positive affect.

41

Procedure

In order to participate in the study, participants had to fill out the SOQ as a pretest
measure of trait competitiveness. All students in introductory psychology are required to
participate in a certain number of studies or to complete an alternative service in order to
fulfill requirements for course credit. To participate, students registered on the university
subject pool website. The SOQ was posted on the university subject pool website as a
“global questionnaire.” All students who sign-on to the website are notified that they can
fill out these global questionnaires to qualify for additional studies. These global
questionnaires are not explicitly associated with any particular study on the website. The
SOQ was titled a measure of “sports and competition.” I also included a gender question
and a question assessing level of athletic involvement on a 5-point scale from “no
participation in sports” to “participation in college athletics.” Students were required to
enter their e-mail addresses when filling out these measures. Eight hundred forty eight
students filled out the pretest measure. About 30% were male. The experimenter then
contacted all male students who completed the questionnaire and invited them to
participate in the study via e-mail. In this e-mail the experimenter notified the students
that they had been selected to participate in a study that would assess their performance at
a variety of tasks. They were also told that there was a chance they could earn money for
their participation. Finally, the e-mail instructed the students how to sign-up for sessions
of the experiment.

Once a participants had signed up, an experimenter sent him an e-mail reminder
the night before his session to increase the show-up rate. Participants signed up for

sessions that could accommodate one to four students. These sessions were randomly

42

assigned to conditions. When the participant(s) arrived, an experimenter welcomed the
participant(s) into the laboratory and asked them if they had any arm, shoulder, or back
injuries that would disqualify them from performing the persistence task. Any students
who had such injuries would be given credit and dismissed.

The experimenter seated participants around a large table where they completed
informed consent forms. As the participants did this, the experimenter stated that he or
she was going to call another laboratory where a similar study was being conducted. The
experimenter informed the participants that depending on what condition they were in,
the participants might be paired via a computer link with a participant from this other lab
during the study. This procedure was actually an experimental deception used to
accomplish two goals. First, for many of the experimental sessions only one student
signed up. This meant that without such deception, it would not be possible to complete
all of the dyad conditions. With other supposed participants in another lab, students were
led to believe that they would not be completing the experimental tasks alone. Second,
when participants in the dyad conditions complete the tasks on the computer, the program
tells them that the name of their partner is “Robert Jackson.” So if students knew all the
other participants in their session, they could see through this deception. By telling them
that another group of students in a different lab is participating in this study, they would
be more likely to believe that their experimental partner existed. The script for the bogus
fake phone call read by the experimenter was as follows: “Hi John. I have (number of
participants present) students. How many do you have? Oh you have (same number)

also. Okay, well we’re ready to start. All right goodbye.”

43

After the phone call, the experimenter described and demonstrated portions of the
experimental task for the participants. First the experimenter informed participants that
they would be performing the study in small booths adjacent to the lab. The remainder of
the study would run by computer and the experimenter would be in the main lab room.
The experimenter then demonstrated the persistence task to the participants stressing that
they should grasp the hand weight firmly, sit up straight, keep elbow locked, and place
their free hand on the control key on the computer keyboard during the task trials. The
experimenter reiterated that the computer would present participants with all necessary
information throughout the study.

At this time the experimenter took each participant to his respective booth
individually. To do this, the experimenter printed out the list of all participants signed up
for the session prior to this time. The experiment needed to enter each participant’s name
into the computer program before each session for two important reasons. The first was
in order to identify and retrieve data by condition and session. Second, after each task
trial, the computer gave participants feedback information on how well they and, if
applicable, their partner had done. The participant’s first name was used in this feedback
screen. Booths were labeled AD. The experimenter entered names in the order they
appeared on the computer print out. To match this predetermined order to the proper
participant, the experimenter collected consent forms from the participants, looked at
their signatures and sent them to the appropriate booth. The experimenter was careful not
to use the participants’ names so that when the feedback screen came up with the name
“Robert Jackson” in the dyad conditions, participants would not know who precisely his

partner was. Finally the experimenter told participants that if they had any questions

44

while in the booth, or if they had completed the study that they should slip a red sheet
provided to them under the door. At that point the experimenter would come to get them
from the booth. The experimenter asked if participants had any questions, answered
them, and then put each participant in the proper booth, instructed him to follow the
directions on the computer, and closed the door to the booth.

All booths were approximately equal in size (10ft. x 10ft.) and contained the same
materials and equipment arranged in the same manner. There was a small desk centrally
located. On this desk was a computer with a keyboard and Weston machine in front of it.
A small web-cam was mounted next the keyboard to encourage participants to complete
the task trials properly. In addition, experimenters placed a clipboard containing the
priming task and the red sheet that would be used to summon the experimenter on top of
the computer tower. Experimenters also provided a hand weight on the desk and a paper
bag where participants were instructed to place any watches of bracelets during the study
to ensure they could not time themselves or wear something that would interfere with the
hand weight.

Participants were randomly assigned to conditions within a 2 (competitive prime
vs. neutral prime) X 3 (individual, coactor, or conjunctive) design. The priming
condition differed according to which priming stimulus the participant was given. The
task context condition (individual, coactor, or conjunctive) differed by what instructions
and feedback the computer program gave to participants. Participants in all conditions
received the same instruction screens before the first task trial. These initial instructions
described how participants should perform the task trials. These were very similar to the

instructions the experimenter gave to participants before entering the booths.

45

After the computer program instructed participants on how to perform the
persistence task, it described how they could receive money for their participation. The
instructions stated that two people would be randomly selected at the end of the study to
win money. Further everyone had an equal chance to win money regardless of their
performance on the trials. However, the money awarded would be based on how long a
participant lasted such that for each second one held his arm above the tripwire, he would
gain five cents. So winning the prize was not contingent upon performance, however the
amount of that prize was. I included this external motivation so that participants would
be more likely to exert reasonable effort at the task.

Then, the computer program instructed the participants to grasp the hand weight
in their dominant hands and get ready for the first of several trials. There was then a
countdown on the screen to signal the start of the first trial. As soon as a participant
could no longer hold up his arm, it would drop and depress the tripwire on the Weston
machine pushing the spacebar and ending the trial. The computer program recorded the
trial length and a screen appeared telling the participant that the first trial was now over.
The program further instructed the participant to switch the hand weight to his other arm
and await the start of the next trial that would be indicated by the computer countdown.
The procedure for the second trial was identical to the first in all conditions except for the
switching of arms as indicated above.

At this point, participants completed the priming task. The computer program
instructed participants to open the envelope next to the computer and follow the
instruction on the form contained within. This form was the priming task. It was labeled

a “Word Finding Task” at the top of the page. Instructions stated that the task was

designed to measure the participants’ verbal and creative abilities. In order to complete
the task participants had to find words of three letters or longer hidden within the given
words. For complete copies of the priming tasks refer to Appendices D and F.

Half the participants were placed in the neutral prime condition, the other halfin
the competitive prime condition. As mentioned before the neutral prime contained 25
unrelated neutral words while the competitive prime contained 15 competitive words and
10 neutral words. The participants were given 15 minutes to complete this task. The
computer screen displayed the number of seconds remaining to complete the task. If
participants finished before this time, they could click on a button that would take them to
the next screen in the computer program. However, if this elapsed time was less than 10
minutes the computer would continue the countdown until at least 10 minutes had
elapsed before advancing to the next screen. prarticipants took longer than 15 minutes,
the computer program automatically advanced to the next screen instructing participants
to move on. This limited the time to complete the task between 10 and 15 minutes.

Subsequent instruction screens after the participants finished the word task
priming procedure differed depending on condition. For those in the individual
condition, the computer program provided instructions identical to those of the first and
second trial for the next two trials. However, in all conditions, participants received
comparable rest between trials to ensure that there was not differential fatigue between
conditions. So, participants in the individual condition completed four trials with
identical instructions except for which arm to use (alternated by trial) interrupted only by

the priming procedure after the second trial.

47

Instructions for the conjunctive condition told participants that for the next few
trials they would be following a somewhat different procedure than before. During these
trials they would be paired with another student at the task to create a two-man team.
Each participant would receive the same score. That score would be the trial score of the
person whose arm dropped first. They were told that during the trials two boxes would
appear on the screen. Each box would have one of the participant’s names in it. The box
would remain green as long as that person’s arm remained above the tripwire. As soon as
one person’s arm touched the tripwire, their corresponding box would turn red. This
would signal the end of the trial and the other person could lower his arm. The partner’s
name was always given as “Robert Jackson.” As described earlier the participants
believed this partner was in a booth either next to them or in the other lab. In actuality
the partner’s box would never turn red. The real participant would necessarily always
quit first. Finally the instructions informed participants that they could still win money as
previously mentioned but that it would be awarded to one of the two-man teams. This
was done to motivate the participants to try their best. The computer program followed
this procedure for two trials, one with participant’s dominant arm and one with the non-
donrinant arm.

In the coactor condition, computer instructions also mentioned to participants that
they would be following a somewhat difl‘erent procedure than before. On the next few
trials they would be able to monitor another participant in the study. As in the
conjunctive condition, this was always a fake participant named “Robert Jackson” who
would always outlast the real participant. Also like the conjunctive condition, both

participants’ names would be in boxes lit green if their arms were still above the tripwire

48

and red when their arms hit the tripwire. However, unlike the conjunctive condition, the
two participants’ performances were independent of one another. That is, the
performance of one did not affect the score the other received. So participants were
instructed that they were simply able to monitor the performance of another student, but
not that they were teammates. In the conjunctive condition, the participant’s box on the
computer screen would turn red as soon as his arm hit the tripwire and the trial would end
meaning “Robert’s” box would never turn red. In the co-active condition, “Robert’s” box
would stay lit green for 1.42 times as long as the participant. This ratio is based on
Kohler’s original studies (Kohler, 1926; 1927) and more recent work (Messé, et al.,
2002) which has found that this discrepancy between partner’s leads to a maximal
motivation gain under conjunctive task demand conditions. The computer followed this
procedure for the third and fourth trials. The third trial instructed participants to use their
dominant arms, and then prior to the fourth trial, their non-dominant arms. Finally the
program informed participants that they could still win money according to the same
procedure mentioned at the beginning of the program. That is, two participants fiom the
study would be randomly selected to win money. How much money they would earn
would be contingent on their performance.

After the fourth trial of the persistence task in all conditions, the computer
program informed participants that they had just completed the last trial. Now they
would be required to answer some questions on the computer. These questions consisted
of a manipulation check that asked participants if they worked with another person at the
task over computer, how hard they worked, the PAN AS, and the SOQ. After completing

these questions, participants were given the option of reading a short lecture about the

49

purpose of the study or being dismissed. If they choose to be dismissed, the program
instructed them to place their red sheet under the door and wait for the experimenter to
get them. Those who read the lecture were given the same instructions after reading it.
The lecture informed participants of some major findings in group research and also gave
them some information about the present study. Once participants placed their red sheets
under the door, the experimenter opened the door to the booth, asked them if they were
done and took them back into the main room. At this point the experimenter debriefed,
thanked, and dismissed the participants.

Mulls
Characteristics of Pretest and Experimental Samples.

As in the pilot study (Study 1), in order to recruit participants for the main study
(Study 2), the experimenter posted the SOQ on the university’s subject pool website as a
global questionnaire. In addition, there were questions about the students’ level of
athletic involvement on a 5-point scale from 1 (none) to 5 (college athlete) and gender.
Eight hundred forty eight students filled out this pretest measure. Two hundred eighty-
eight were male and 560 were female.

Mean scores on the SOQ were as follows. Males scored an average of 100.05 on
the overall SOQ, including means of 51.82 on the competitiveness subscale, 22.22 on the
win orientation subscale, and 26.01 on the goal orientation subscale. Females’ overall
mean SOQ score was 90.06 with means of 44.75, 19.79,and 25.52 for the
competitiveness, win orientation, and goal orientation subscales respectively. AN OVAs
revealed that males scored significantly higher on all these measures (all F’s > 40, p <

.0001) with the exception of the goal orientation subscale, F(1, 846) = 3.51, p < .07.

50

(refer to Table 3.1). However, only the overall SOQ and the competitiveness subscale

were examined in this study.

Table 3.1

Scores on SOQ and Subscales by Gender on Pretest Sample

 

 

 

Measure

F df

 

SOQ

Competitiveness
Subscale

Wm Orientation
Subscale

Goal Orientation
Subscale

Female
M S
90.06 16.59
44.75 10.88
19.79 4.81
25.52 3.71

71.89M 1, 846

86.94“ 1, 846

48.21“ 1, 846

351* 1,846

 

Note. All items are on 5-point scale. Higher scores indicate more endorsement of items.
SOQ contains all 25 items on measure. Competitiveness subscale contains items 1, 3, 5,
7, 9, 11, 13, 15, 17, 19, 21, 23, and 25; Win Orientation contains items 2, 6, 10, 14, 18,
and 22; Goal Orientation contains items 4, 8, 12, 16, and 20. PS with p values < .05 are
marked with **. Those that are marginally significant (p < .10) are marked with *.

A factor analysis was performed in order to assess reliability of the SOQ. The

overall SOQ achieved acceptable levels of internal consistency, Cronbach’s alpha = .94.

The factor analysis using a varimax rotation extracted four factors. The first factor

contained almost all of the items from the competitiveness subscale with the exceptions

of items 3,5, and 25. That is all these items had a factor loading of .4 or above on a
Principal Components factor analysis with a Varimax rotation and Kaiser Normalization.
Only one item not on the competitiveness subscale loaded on this factor (item 14) with a
.407 factor loading. Factors 3 and 4 loaded almost perfectly with the win orientation and
goal orientation subscales using a factor loading of .4 as the criterion. Factor 2 contained
items 1-7, 10, and 11 which are not predicted by Gill, Dzewaltowski, and Deeter to
combine as a subscale. However, since I was only using the competitiveness subscale
and the overall SOQ as dependent measures, these issues do not concern the current

results. For a detailed presentation of the factor loadings for this scale, please refer to

 

 

 

Table 3.2.
Table 3.2
SOQ Pretest Factor Loadings
Factor
Item: Competitiveness Unpredicted Win Goal
1. I am a determined competitor .405 .678 .158 .268
2. Winning is important .184 .561 .521 .009
3. I am a competitive person .266 .743 .242 .201
4. I set goals for myselfwhen .204 .412 .110 .608
I compete
5. I try my hardest to win .219 .567 .207 .390
6. Scoring more points than my .198 .537 .515 .131
my opponent is very important
to me
7. I look forward to competing .614 .562 .007 .180
8. I am most competitive when -.005 .148 -.003 .703
I try to achieve personal goals
9. I enjoy competing against .555 .616 .103 .142
others

52

10. I hate to lose -.003 .235 .789 .119

11. I thrive on competition .508 .484 .320 .152

12. I try my hardest to win when .144 .258 .151 .636
I have a specific goal

13. My goal is to be the best .745 .149 .239 .212
athlete possible

14. The only time I am satisfied .407 .008 .668 -.002
is when I win

15. Iwant to be successfirlin .734 .217 .175 .104
sports

16. Performing to the best of my .309 .175 .006 .592
ability is very important to me

17. I work hard to be successfirl .785 .190 .131 .213
in sports

18. Losing upsets me .162 .114 .752 .101

19. The best test of my ability is .523 .194 .394 .102
competing against others

20. Reaching personal .203 .009 .005 .723
performance goals is very
goals is very important to me

21. I look forward to the .639 .355 .123 .323
opportunity to test my skills
in competition

22. I have the most firn when I win .190 .109 .672 -.004

23. I perform my best when I am .528 .303 .300 .165
competing against an opponent

24. The best way to determine my .160 -.004 -.008 .747
ability is to set a goal and try
to reach it

25. I want to be the best every time .388 .112 .464 .348
I compete

 

The final sample of participants in the study included 80 male students who had
filled out the pretest measures. Unfortunately not everyone in the study filled out the
pretest measure. It was necessary to include other students because the signup rate for
the study was relatively low. Means for those who filled out the pretest measure and
participated in the study were as follows. The overall mean SOQ score was 99.68

including means of 51.34, 22.10, and 25.74 on the competitiveness, win orientation, and

53

goal orientation subscales respectively. Scores on the athletic involvement item revealed
that two men did not participate in any sports, 16 sometimes played recreational sports,
14 played intramural or club sports, 39 had played for a high school team, and 9 played
for a university team.

Task Performance and Priming Effects.

As previously mentioned, participants performed a persistence task by holding up
their arms as long as possible for four trials alternating dominant and non-dominant arms.
To test if participants fatigued and at what rate with each arm across the four trials a 2
(Trial Block) X 2 (Trial within Block) repeated-measures AN OVA was conducted on
those in the control condition (individual). The first factor simply contrasts the first two
trials with the last two [(Trial 1 + Trial 2) / 2 vs. (Trial 3 + Trial 4) / 2] while the second
factor tests participants’ dominant versus non-dominant arm. A significant Trial Block
main effect emerged (F(1, 39) = 34.79, p < .0001) indicating that those in the control
condition persisted longer in the first trial with each arm (Block 1 mean = 103.65, s =
45.77) than on their second trial with each arm (Block 2 mean = 79.89, s = 38.24). There
was also a significant arm main effect (F(1, 39) = 23.84, p < .0001) indicating that these
participants persisted longer with their dominant arms (98.263, s = 41.53, n = 40) than
with their nondominant arms (85.28s, s = 40.61). Finally, a significant Block X Arm
interaction emerged (F(1, 39) = 4.51, p < .05) indicating that participants fatigued at
different rates with each arm. On average, control participants decreased performance by
28.535 from the first to the second trial with their dominant arms and 19.0s with their
nondonrinant arms. Examined multiplicatively, participants last 1.485 times longer on

the first compared to the second trial with their dominant arms and 1.3956 times longer

54

with their nondominant arms. So any participants in the conjunctive and coactive
conditions whose performance declines less than these figures from Block 1 to Block 2
would manifest a motivation gain.

In order to analyze the performance of the participants in the experimental
conditions (conjunctive and coactive), scores on the experimental trial (3 & 4) were
corrected for fatigue. This was done by multiplying scores for the experimental trials by
the rate of fatigue shown in the control condition by each arm (1.485 for dominant and
1.3856 for nondonrinant). These fatigue corrections replicate the methods of previous
research in motivation gains (Hertel, et al., 1999).1 I then subtracted the scores from the
first trial of each arm from these fatigue corrected scores to eliminate variability from
these control trials. The result was fatigue corrected difference scores. 80 any score
above zero on this metric represents a motivation gain since the result of the equation
eliminates fatigue and arm differences.

Fatigue corrected scores were analyzed separately for Trials 3 and 4. This was
done because feedback that participants received after each trial differs qualitatively.
Before Trial 3 the only difference between the conjunctive and coactive conditions was
the instructions that participants received was a task demands manipulation. After Trial
3, those in both the conjunctive and coactive conditions were both given feedback that
they had stopped before the other participant had. In the conjunctive condition, the trial
ended as soon as the participant’s arm hit the tripwire. Since this was a conjunctive team
task, the other participant did not continue and there was no information how much better

the other participant was, but simply that the other participant lasted longer. In the

 

‘ A number of different fatigue corrections could be used for the motivation scores. This was chosen
amongst alternatives because others yielded similar or attenuated results but were more complicated

55

coactive condition, the other participant outlasted the real participant as well, but since
their performances were independent, the other participants always persisted at the same
ratio beyond the real participant (1.42 times longer). So information about ability
discrepancy, task demands, and performance feedback are confounded on the fourth
trial:

In order to test for motivation gains and priming efl’ects in Trial 3, a 3 (Group
Condition: individual control, conjunctive, and coactive) X 2 (Prime) AN OVA was
performed on the fatigue corrected Trial 3 -— Trial 1 difference scores. Results revealed a
significant effect of Group, F (2, 114) = 16.95, p < .0001. Planned contrasts reveal that
those in the conjunctive condition (M = 72.7s, s = 65.6) significantly outlasted those in
the individual condition (M = 12.2s, s = 34.6, F(1, 114) = 23.40, p < .0001) indicating a
significant motivation gain for those in the conjunctive condition. Those in the coactive
condition (M = 74.7s, s = 63.5) also manifested a significant motivation gain over those
in the individual condition (F(1, 114) = 27.31, p < .0001). The conjunctive and coactive
conditions did not significantly differ from one another (F (1, 114) = .17, ns). So it
appears that in the third trial, conjunctive and coactive task demands are equally effective
in producing motivation gains. In past research, coactive conditions have led to increases
in performance over individual conditions but significantly lower than conjunctive
conditions (Kerr, Messé, Park, Seok, & Lount, in preparation). To test the Kohler effect,
that is that conjunctive task demands lead to the greatest motivation gains, a planned
contrast was performed comparing only unprimed conjunctive and coactive conditions.

This comparison most closely mimicked conditions in previous research. The contrast

 

2 Those who would like to see an examination of fatigue corrected experimental trial scores on the
conjunctive and coactive conditions compared by priming condition and arm can refer to Appendix J.

56

revealed that task demand condition still did not differ (F( 1, 114) .28, ns). Means for all
conditions on Trial 3 are presented in Figure 3.1.

The overall efl‘ect of priming was not significant (F(], 114) = 2.38, ns). However
this analysis included individual controls whom it was hypothesized, would not be
affected by the prime. Therefore, firrther contrasts were performed to test the
effectiveness of the priming manipulation. Contrasts reveal that those who were primed
lasted significantly longer on Trial 3 under coactive conditions (97.9s versus 57.5, F(1,
114) = 5.13,p < .03) but not under conjunctive conditions (78.2s versus 66.8, F(1, 114) =
.4, ns). This priming effect makes sense in light of the differential task demands of the
situations. Those who are in the presence of another who is not a teammate should be
more motivated to outlast the other participant when primed for competitiveness rather
than those who are supposed to be working collectively with that person. Unsurprisingly,
priming had no efi’ect in the individual condition when there was no one present with

which to compete (F < .5).

 

1 20
1 00
8O
60
4O
20
0

 

0
E
':
0.

Competitive
Neutral
Competitive
Neutral
Competitive
Neutral

Group Individual Conjunctive Coactive

 

 

 

Figure 3.1 Motivation Gain for Trial 3 by Condition

57

As stated previously, the fourth trial is qualitatively different from the third
because performance feedback and information about partner’s ability are confounded.
So Trial 4 data were examined separately. To examine the effects of task demands and
priming a 3 (Group) X 2 (Prime) ANOVA was performed on fatigue corrected Trial 4
scores. Analyses revealed a significant main effect of Group (F(2, 114) = 6.82, p < .01).
Further analyses revealed that those in the conjunctive condition (M = 41 .93, s = 64.9)
outlasted individual controls (M= 10.2s, s = 33.4, F(1, 114) = 8.44,p < .01). Coactors
(M= 45.7s, s = 40.2) also outlasted individual controls (F(], 114) = 11.75,p < .01). As
in Trial 3, those in the conjunctive and coactive conditions did not differ from one
another (F < .3). For all means by condition in Trial 4 refer to Figure 3.2. To test the
Kohler effect in Trial 4, unprimed scores were compared in the conjunctive and coactive

conditions revealing no significant difference (F < .2)

 

 

o 0 To 0 3 o E
2% t ‘2: =5; :3. i. ‘2:
Q 2 0. 2 Q 2
E E E
0 0 0
Group Individual Conjunctive Coactive

 

 

 

Figgre 3.2 Motivation Gain for Trial 4 by Condition

58

The overall priming effect in Trial 4 did not reach statistical significance (F (1,
114) = 1.92, ns). Once again though since individual controls were included in this
analyses, further planed contrasts were warranted. However, in Trial 4 the prime did not
reach statistical significance for either the conjunctive (F (1, 114) = 1.75, ns) or the
coactive conditions (F (1, 114) = 1.6, ns), although means were in the predicted direction.
This may mean that the prime wears off by Trial 4 or that the confound between
performance feedback and partner’s discrepancy decreases motivation. In Trial 4,
participants already know they have been beaten. This may decrease their motivation on
the fourth trial. Before Trial 3 they have no such information and have just been primed
for competitiveness. So since participants had been primed just before the start of the
third trial and have no information about the ability of their partner, they will have higher
motivation in the third trial.

In order to examine the effects of participants’ pretest trait competitiveness scores
a number of analyses were performed. First to measure the effect of pretest
competitiveness without any priming or task demands manipulation, bivariate
correlations were calculated between pretest competitiveness scores (as measured by both
the SOQ and its competitiveness subscale) and raw scores on Trials 1 and 2 which as you
recall were simply control trials. Results revealed low to moderate correlations ranging
from .13 to .27. All correlations are presented in Table 3.3. The only significant
correlation was between the SOQ pretest scores and Trial 2 performance (r(120) = .27, p
< .03). So it appears that competitiveness did exert a small effect on performance at a

persistence task.

59

Table 3.3

Correlations Between Pretest Competitiveness and Control Trials

 

Pretest SOQ Pretest SOQ competitive Trial 1 Trial 2

 

Pretest SOQ -

Pretest SOQ .92*** -

competitiveness

Trial 1 .21* .13 -

Trial 2 .27" .16 .86*** -

 

Note. This sample contained the 80 participants who completed the pretest questionnaire
and participated in the main study. *p < .10, **p < .05, ***p < .0001

To test whether pretest competitiveness predicted affected performance after the
priming and task demands manipulations a 2 (group condition) X 2 (prime) AN COVA
was conducted using pretest SOQ and pretest SOQ competitiveness subscale scores as
covariates. The effect of the SOQ was non significant on the fatigue corrected difference
scores examined either by each experimental trial (Trial 3 and 4 individually) or as
overall scores (average of trials 3 and 4, (F5 < .5). Similar results were obtained when
using the SOQ competitiveness subscale as a covariate (F’s < 1).

To test whether pretest competitiveness predicted performance in each group
condition, regressions were conducted using fatigue corrected difference scores examined
by each arm and as overall scores. Prime was entered as a dummy variable as were
pretest SOQ scores. For those in the conjunctive condition, SOQ scores did not exert a
significant effect. Bivariate correlations were all low, ranging from .05 to .19, ns (n =
31). None of the coefficients reached statistical significance including the interaction

between priming and SOQ scores. This was true for all fatigue corrected scores. The

60

same regression was performed in the coactive condition revealing similar results. All
bivariate correlations were low, ranging from .05 to .29 with the exception of the Prime X
SOQ interaction in the third trial (r(25) = .31, p < .08). However, none of the coefficients
reached statistical significance indicating that trait competitiveness did not significantly
moderate task performance by itself or in conjunction with the prime.

Posttest Measures of Trait Competitiveness

In order to test the hypothesis that scores of posttest competitiveness would be
changed by priming manipulation and task demands, a 2 (prime) X 3 (dyad condition)
ANOVA was conducted on posttest scores of the SOQ and its competitiveness subscale.
These analyses included all participants, even those who did not fill out the pretest SOQ.
Since participants were randomly assigned to conditions, any pretest differences should
be evenly distributed across conditions. Results reveal a marginally significant effect for
prime on both the overall SOQ and the competitiveness subscale respectively, F (l , 114)
= 3.88,p < .06; F(1, 114) = 3.13,p < .08. The effect ofdyad condition was not
significant for either measure, F’s < 2. The prime by group interaction was also
nonsignificant, F’ s < 1.

Since priming was somewhat effective in boosting motivation gain scores in the
coactive conditions, it may also be the case that competitiveness will increase in these
instances as well. For exploratory reasons, I performed contrasts which revealed a
significant effect of priming in the dyad conditions on both the SOQ and the
competitiveness subscale respectively, t(114) = 2.24, p < .03 and t(114) = 2.02, p < .05.
The priming effect did not appear in the individual condition, t < 1. So it appears that

priming can effect scores on a trait measure of competitiveness especially when one

61

competes (i.e., a group condition). The effect of task setting (group: coactive and
conjunctive vs. individual) was marginally significant for both measures, 1014) = 1.97, p
< .06; t(114) = 1.97, p < .06. Counter to the predictions, those in the competitive
condition scored higher (M = 102.3, s = 11.5) than the conjunctive (M = 96.8, s = 20.8) or
the coactive conditions (M = 94.6, s = 18.0) on the SOQ. A similar pattern emerged on
the competitiveness subscale in which individuals (M = 53.1, s = 7.1) scored higher than
those in conjunctive (M = 49.7, s = 12.2) and coactive (M = 48.4, s = 11.2) conditions.
Those in the conjunctive condition did not difl‘er from those in the individual condition
on either measure, ((114) = 1.53, ns; t(114) = 1.51, ns.
Mood Effects

Analyses on PANAS scores revealed that those above the median on the SOQ
competitiveness subscale pretest measure scored significantly higher on positive affect
than those below the median, F(1,76) = 6.56, p < .02. However there was no significant
difference between these two groups on negative affect, F (1 ,7 6) = 1.03, ns. There were
also no significant interactions between pretest competitiveness, as measured by a median
split, and priming or task demands conditions. There were no significant effects for
priming or task demands conditions (F’s < 1.3).

Discussion

Priming and Task Demands in Trial 3

Only Trial 3 data were be examined here because participant feedback in this trial
is qualitatively difl’erent from that in Trial 4. In Trial 3, participants have just been
primed and, in the group conditions, perform in the presence of another for the first time

without knowing the ability of this other person. In Trial 4, those in the group conditions

62

receive differing levels of information about the other person. In the coactive condition,
participants know the other is not only better than they are but by what amount. In the
conjunctive condition, participants know the other is better but not by what amount. So
the Trial 3 comparison between the two task demand conditions is a more viable one
where the Trial 4 data are confounded by feedback about one’s partner.

Results showed that priming effectively increased performance for those in the
first experimental trial (Trial 3) for the coactive condition but not the individual or
conjunctive condition. This makes sense in light of the task demand situations. When
competitiveness is made salient by priming, those who work alone still have no one with
whom to compete. Ifwe take competition to mean doing one’s best in relation to others,
it is clear why priming would not moderate performance in the individual condition. If
one desires to compete (or competitiveness is made salient to him), competition is not
possible and so priming should not moderate this performance. We can use the lock and
key analogy to understand this situation. In this case, the key is the concept made salient
by priming. The lock represents the demands of the situation. In order to overcome the
challenges of the situation and succeed, one must use the key to open the lock. In the
individual condition, priming provides a key for the wrong lock. That is competitiveness
is not applicable in this situation.

In the coactive condition, competition becomes an appropriate behavior because
another person is present (a potential competitor). When one is primed for
competitiveness, beating this other person who is not framed as a teammate or ally
becomes salient. So in this situation priming provides the appropriate key for the look.

In the coactive condition participants were told that they could simply monitor another’s

63

performance. However, if participants were specifically instructed to beat this other
person or this coactor were framed as a participant, competitiveness priming would likely
moderate performance to an even greater degree.

In the conjunctive condition, priming did increase performance but not
significantly so. Again this makes sense in light of task demands. In a conjunctive
situation, the team score is equal to that of the member who quits first. So performing
one’s best in relation to others is important but not as important as giving one’s all for the
group which has more to do with cooperation. So in this case priming competitiveness
provides a key that doesn’t quite open the lock. Cooperative or group loyalty priming
might well have moderated task performance in this condition. This issue harkens back
to the idea that competitiveness and cooperation are not antithetical. Some in the
conjunctive condition may have enhanced performance through priming because they
saw beating their teammate as a way to increase the team’s score. We can also think of a
situation such as a team sports game where teammates must cooperate with one another
to compete against other teams.

Finally, examination of these data helps us to refine our definition of competition.
I previously defined competitiveness as one’s desire to perform at one’s best in relation to
others or one’s own standards. The data show that priming participants for
competitiveness motivated them to perform their best in relation to others but not against
their own standards or previous performance. So competition here is interpersonal rather

than intrapersonal.

64

Priming and Task Demands in Trial 4

The comparison between the two group task demands conditions for the Trial 4
data is more complicated than for Trial 3. As mentioned previously, those in the coactive
condition have more information about their partner’s ability than those in the
conjunctive condition. This should help us to understand the nonsignificant priming
effect in Trial 4.

As shown in Figure 3 .2, those in the primed for competitiveness outlasted controls
in the same task demand conditions in Trial 4 by 20.1 and 19.4 seconds for conjunctive
and coactive conditions respectively. This gain is nonsignificant. However this same
comparison in Trial 3 yielded performance increases of 11.5 and 40.4 seconds for the
conjunctive and coactive primed versus nonprimed groups, respectively. This
comparison was significant for the coactive but not the conjunctive condition. We can
speculate that the effect of the prime wears off or decreases by the second experimental
trial (Trial 4). It is also possible that upon receiving feedback that they have quit the task
first, participants are less motivated to try quite so hard again. However, for those in the
conjunctive condition, performance increases from Trial 3 to 4 although not to significant
levels over the unprimed condition. Those in the conjunctive condition received
information that they quit the task first but their partner’s ability was unknown. Those
primed for competitiveness may then believe that increased effort may allow them to beat
this partner. Those in the conjunctive condition also received feedback that they had quit
the task first but also know their partner lasted 1.42 times longer than they had.

Priming competitiveness before the third trial may lead to a performance or ego

orientation on this trial. Those performing under this orientation see competition as an

65

opportunity to showcase their abilities. However, when they lose and find out their
ability is inferior to their partner’s, they lose motivation. Future studies should explore
this possibility by manipulating failure with success feedback after trials and also
manipulating ego versus mastery orientation. Mastery orientation has been shown to
increase performance at a task when one is unskilled at the task (Elliot & Dweck, 1988).
So while the effect of priming was only significant for coactors in the third trial, it led to
increases in performance in every group condition and trial.
Trait Competitiveness

Results show that in general, trait competitiveness as measured by the SOQ at
pretest, had little effect on performance. The only significant effect of pretest SOQ
scores was on Trial 2 scores. The modest positive correlation (r = .27) showed that more
competitive people performed better on the second trial before any priming or task
demand manipulation had taken place. This effect was only marginally significant in
Trial 1 (r = .21). The relationship to experimental trials was lower. Correlations between
the SOQ measures and raw Trial 3 and 4 scores ranged from .09 to .18 and all were non
significant. While correlations between fatigue corrected trial scores, which eliminated
performance difi‘erences that might have been due to trait competitiveness, were near
zero (r’s < .02, ns). So trait competitiveness was not related to performance on
experimental trials whether corrected for fatigue or not.

On the experimental trials, pretest SOQ scores affected nothing significantly.
Those primed in Trial 3 showed a marginally significant trend toward better performance.
However, trait competitiveness did not predict performance in prime or task demand

conditions. This pattern goes against the hypotheses. The results were limited by a small

66

number of participants (n = 80) who filled out the pretest SOQ. However, the lack of
significant results on the performance measures lends credence to the idea that the power
of the situation (task demands condition) and state competitiveness (induced by priming)
are more predictive of performance at some group tasks than trait competitiveness.
Malleability of Trait Competitiveness

Results on posttest measure of trait competitiveness (SOQ & SOQ
competitiveness subscale) were similar to those obtained on the performance measures.
That is, those who worked in either of the two dyad conditions were marginally more
competitive on these measures than those who worked alone. Although priming did not
make all those primed more competitive, it did make those in the two dyad condition
appear significantly more competitive when compared to those in their conditions who
were not primed. All these analyses were done on posttest SOQ scores. When looking at
SOQ difference scores from pretest to posttest, there were no significant efl‘ects.
However since only 80 participants yielded this data, power is decreased when compared
to data from the whole sample (n = 120).

It appears that competitiveness as measured by the SOQ was very difficult to
change. Correlations fi'om pretest to posttest scores on the SOQ and the competitiveness
subscale were .72 and .76 respectively. This makes it all the more impressive that in
particular situations such when working in dyads or when primed for competitiveness in
dyads, one can actually become more competitive when measured by these self-report
instruments. So while competitiveness measured by the SOQ appears to be very stable

and trait like, it is not unchangeable.

67

Mood Measures

Scores on the PAN AS revealed only one significant effect. Those who were
above the median in pretest competitiveness exhibited more positive affect than those
below the median. However, counter to the hypotheses, those who were more
competitive, when primed and placed in dyad conditions, did not exhibit any more
positive affect than other groups. In fact there were no significant effects on the PAN AS
for prime or task demands conditions. This is counterintuitive. The hypotheses assumed
that those who are chronically more competitive and get to compete and/or are primed for
competition should be in a better mood. Since, participants filled out the PAN AS at the
end of the study, losing on the experimental trials may have affected their mood.
However, it is unclear why mood was unaffected for the various conditions.

One possible explanation is that the wording of the PANAS makes it unlikely that
those who just completed a physical persistence task would highly endorse the items.
Examples of some adjectives that participants have to endorse to score high on positive
affect are “interested”, “excited”, and “proud”. Even though participants might feel
more competitive after being primed or have performed better, performance at this type
of task is unlikely to elicit such emotion. This task was chosen specifically because it is
uninteresting to most people and performance at it should have little efl‘ect on self-esteem
because it is not a task most people have performed before or consider important.

The effect that competitive people have higher positive affect is most surprising.
It is possible that most competitive people see the task as a competitive one and then
enjoy themselves more as a result. This explanation appears unlikely though because

PAN AS scores did were not moderated by prime or condition. Since some conditions are

68

more competitive than others, we would expect competitiveness to interact with condition
or prime but it does not. It could be that competitive people are generally in a better

mood, but there is no previous research to support that idea.

69

Chapter 4

GENERAL DISCUSSION

Measures

Based on the results from Study 1 and Study 2, the SOQ appeared to be the best
way to measure competitiveness among these samples, while the SVM seemed to be the
worst measure of competitiveness. As mentioned previously, this may be because the
SVM measures allocation of resources while the SOQ measures motivation to perform
well in relation to others and one’s own goals. One criterion for establishing the validity
of these competitiveness measures is to test their relation to competitive behavior.
Athletes, for example are chronically exposed to competitive situations and one could
argue that they are more competitive because of the effects of these situation or that their
competitiveness leads them into these situations. There was a measure of athletic
involvement included in both studies. This measure was then correlated with all of the
competitiveness measures. Athletic involvement was most highly correlated with the
SOQ competitiveness subscale (r = .53,p < .0001, Study 1; r = 64,p < .0001 , Study 2
pretest; r = .61 , p < .0001, Study 2 posttest). The overall SOQ was also highly
correlated to athletic involvement with correlations ranging from .49 to .59 (p’s < .0001).
The other subscales of the SOQ (win orientation and goal orientation) were also
significantly correlated to athletic involvement but not as highly as the previously
mentioned scales with r’s ranging from .26 to .41,(p’s < .01).

The CI, which was used only in Study 1, also exhibited a significant relationship

to athletic involvement, r = .32, p < .0001. Correlations between athletic involvement

70

and SVM scales were low. Pretest SVM competitiveness scores were not significantly
related to athletic involvement (r = .08, ns) while posttest scores on this measure yielded
a low but significant correlation (r = .17, p < .05). SVM individualism scores were not
significantly related to athletic involvement and ranged from .06 to .17. SVM prosocial
exhibited a small but significant negative relationship to athletic involvement scores with
correlations ranging from -.18 to -.21. This makes sense within the structure of the SVM
since participants cannot simultaneously make prosocial and competitive choices.

So it seems that the overall SOQ and its competitiveness subscale are the best
measures of competitiveness among these samples. The original factor structure was
very closely replicated with the exception of an additional nonpredicted factor. The
overall SOQ is highly correlated to the competitiveness subscale (r = .97, p < .0001). So
the competitiveness subscale can be thought of as a shorter version of the overall scale
although both seem to be very effective at measuring competitiveness.

Implications for Motivation Gains and the Kohler Effect

In Study 2, participants in the coactive condition did not perform significantly
differently fi'om those in the conjunctive condition. This is true even when examining
those who were not primed which most closely replicates past studies. This result is not
consistent with some previous research (Kerr, Messé, Park, Seok, & Lount, in
preparation). However in Kerr et al. (in preparation), only females participants were
examined. In a partial replication of Hertel, Kerr, and Messé, Experiment 2 (2000),
Hertel (personal communication, November, 2002) found that male participants showed
equal motivation gains in conjunctive and coactive task demands conditions. These male

participants were recruited from a gym and as a result may have been especially

71

competitive. The experimenters in this study were attractive females whom Hertel
hypothesizes may have triggered participants’ impression management concerns. This
overshadowed motivation for group success which should be salient in conjunctive
conditions. As a result, this impression management increased motivation for both
conjunctive and coactive conditions.

It has been hypothesized that the high instrumentality that a weaker member feels
in contributing to the group’s success motivates him or her to perform especially well in
conjunctive conditions. However, since this concern for the group’s success and
indispensability of the weaker person’s input is absent in coactive or additive conditions,
there is little instrumentality to motivate people in this condition (Kerr, 1990; Hertel,
Kerr, & Messé, 2000). A competing explanation for motivation gains in certain tasks is
that social comparison and goal setting lead performers to do as well as stronger members
when performances can be compared (Stroebe, Diehl, Abakoumkin, & Amscheid, 1990).
This model predicts motivation gains for conjunctive, coactive, and additive conditions
but without differences between them.

It would appear that based on the results in the current studies, that the social
comparison/ goal setting model has been supported. However, as mentioned previously,
the gender differences between these studies may be the real explanation for the disparate
results. Since males are more competitive than females, it likely that they engage in
social comparison during these tasks to a larger extent. Their goal is more likely to be to
beat their partner with little concern for the success of the group in the conjunctive
condition and therefore conjunctive and coactive conditions do not differ for males.

However, for the less competitive females, group success and instrumentality of the

72

weakest member leads to motivation gains in the conjunctive condition (over and above
the coactive condition) where instrumentality to group success is salient but competition
is not. More research is necessary to tease apart these competing motivations for each
gender.
Limitations

The priming effects and motivation gains in general are limited to particular
circumstances. For example a significant effect was limited to only the third trial in the
coactive condition in Study 2. Also many researchers have focused on the negative
motivational aspects of groups such as social loafing (Latané, Williams, & Harkins,
1979). So while motivation gains that were moderated by priming were demonstrated,
these results would likely not generalize to some other situations. Below are presented
several factors that would limit or bring into question the generalizeability of the results.

First of all, the physical persistence task used shared several aspects of everyday
tasks such as physical labor and exercise. However, it is unclear as to whether
motivation gains would occur using a similar paradigm but substituting cognitive tasks
for physical ones. Recent research however has replicated motivation gains effects on a
computer sales task (Hertel, Deter, & Konradt, 2001). However, it is unclear as to
whether priming effects would be replicated by such tasks. It is unlikely that priming
competitiveness would increase performance at complex, unfamiliar tasks. These types
of tasks require skill and practice so performance at them is not monotonically related to
effort as it is at a simple physical persistence task.

Another limitation of these effects is gender. While past research in motivation

gains has exhibited robust effect for females (Hertel, Kerr, & Messé, 2000),

73

competitiveness priming was ineffective for females in the present Study 1. This may be
to the content of priming procedures. For example, both of the competitive and
cooperative story prime in Study I referred to sports situations with males athletes which
could be more likely to make competition salient to men but may be irrelevant to women.
Perhaps a competitive story prime referring to females athletes or competition between
groups rather than between individuals would make competition more salient females.
Competitive words referring to exceptional performance in traditionally females domains
might also make competition more salient. Females have been characterized as more
connected to groups than males and as a result might perform better under conjunctive
conditions than males although this was not the case in previous research (Hertel et al.,
2000, Exp. 1). Perhaps, however, cooperation is more chronically salient to females and
priming this concept effectively would enhance performance at a conjunctive task.
Future research in this area should resolve these gender differences.

As with most psychological research, these studies were conducted on mostly
white, college undergraduates in introductory psychology classes. This sample was also
fi'om a large Midwestern university with an athletically oriented student body.
Attendance at athletic events is often at capacity and the performance of school sports
teams is often a hot topic on and around campus. Many students also participate in
varsity or intramural sports. The heavy sports interest at this school may make
competition particularly salient to this sample, especially males and as a result make them
easily primeable to competitiveness. This may not be the case for less sports oriented
communities. The age range may make these participants especially suggestible. The

first few years of college involve identity exploration and experimentation (Erikson,

74

1980). This may well make certain personality traits like competitiveness more
malleable during this time and make people more open to suggestions and priming. Later
in the life span when personality is more set, this priming may be less likely to work.
Males at this point in the lifespan are often engaged in aggressive and competitive
behavior. They are competing for mates, grades, and jobs. This may be caused by
societal as well as physiological factors (increased testosterone). When these demands
decrease, competitiveness should decrease making it less salient. So the priming and
motivation gain effects may be particularly strong for this sample. Future research
should examine other age ranges.

Culture may also play a role. More individualistic societies like our own approve
of the pursuit of personal glory and individual gain more than more collectivistic cultures.
In these cultures, priming interpersonal competition may cause reactance and actually
lead to decreases in performance. Perhaps priming intergroup competition or
international competition and national pride rather than interpersonal competition might
lead to motivation gains in a collective context such as a conjunctive task.

Directions for Future Research

In an efi‘ort to generalize these results, firture research should focus on different
situations which could be moderated by competitive and cooperative priming. For
example, athletic contexts are clearly associated with competition and cooperation.
Perhaps exposure to competitive slogans or words just before game time could boost
performance. This should be especially true for individual sports. For team sports,
cooperative priming could help teammates work with one another. Competitiveness

priming could potentially increase team performance if intergroup rather than

75

interpersonal competition were primed. Other contexts that should be explored are work
and legislative groups in which individuals or teams need to increase motivation to
increase productivity or efficiency.

As already mentioned, one area of interest would be the role of cooperative
priming in performance enhancement. Increased cooperation might make group goals
more salient and enhance performance at conjunctive or additive tasks especially for
females. Also competition can be interpersonal or intergroup. As mentioned earlier
competition and cooperation do not need to be thought of as antithetical concepts. In the
context of intergroup competition, one must cooperate within the group while competing
against other groups in order to attain maximal performance. Priming intergroup
competition or intragroup cooperation should both be effective means of enhancing
performance in these situations. An effective cooperative prime would need to be
developed first.

The present studies examined two person groups as performance settings.
However, it is plausible that this paradigm could be extended to include larger groups.
The procedure is much simpler with only two people but conjunctive or coactive task
could be simulated with more than two people. However, additional issues would arise,
such as ability discrepancies between all partners and feedback information. This
information would be much more complicated but if participants understood they were
the least capable member of the group and their performance was most vital, we could
likely replicate conjunctive task conditions. Kohler (1926, 1927) was able to demonstrate
these effects with three person groups, so perhaps they could be replicated with even

larger groups. So while studying two person groups is simpler and more cost eflicient,

76

most of the groups to which people belong or interact with consist of more than two
people. Expanding this research to larger groups would make it more generalizeable to
the types of groups people interact with every day. There is no reason to believe these
results would not generalize to somewhat larger groups.

In many previous motivation gains studies including the ones examined here,
participants are given feedback that they have performed worse at the task than their
partner or coworker. This is done so that participants will push performance to the
maximum. This is necessary in conjunctive tasks which by definition are dependent on
the weakest member. However, as mentioned earlier some people might be especially
likely to push performance on subsequent trials upon learning that they were the most
capable member of a group. Namely those with a performance or ego orientation would
be motivated to perform well at a task when they know they are better in order to display
their skill. Future studies could manipulate failure or success feedback from previous

trials and see how it relates to motivational orientation and performance.

77

Chapter 5

CONCLUSIONS AND IMPLICATIONS

The results of these studies show that not only do those working at coactive and
conjunctive tasks increase performance over those working alone, but priming in the
coactive condition can even further enhance this performance. So factors that enhance
performance in these groups tasks include group task demands and state competitiveness.
Also these effects were independent of one’s trait competitiveness. So people can
temporarily become more competitive in the laboratory and when another is present will
increase performance regardless of how competitive they were to begin with.

These results are promising for anyone wishing to get an edge at a task which
involves physical persistence such as sports or many types of manual work. It also could
help those who wish to enhance a group’s or team’s performance such as a business
manager or team sport coach. One does not have to be born competitive to gain this edge
but rather must attain a competitive state of mind. It seems there are rituals and
preparations which have adhered to this principle for some time. For example, many
athletes decorate their bedrooms and exercise areas with motivational slogans. Members
of the University of Notre Dame Football team must touch a sign before the start of every
home game hanging just above the entrance to the field which reads “Play Like a
Champion Today.” These types of competitive slogans serve to prime athletes for the
type of interpersonal competition necessary to perform well against others.

However, attempts at priming are not necessarily limited to athletics.

Boardrooms and offices are often decorated with pictures framed by slogans stressing the

78

value of achievement or effort. Managers and supervisors likely believe this type of
stimuli gets their workers in the right mindset to work productively. Results from this
study show that these people may have been right all along. We often do better when
working in the presence of others but may be motivated to do so outside our conscious

awareness.

79

APPENDIX A

80

APPENDIX A
Measure of Social Values

In this task we ask you to imagine that you have been randomly paired with another
person, whom we will refer to simply as the “Other”. This other person is someone you
do not know and that you will not knowingly meet in the future. Both you and the
“Other” person will be making choices by circling either the letter A, B, or C. Your own
choices will produce points for both yourself and the “Other” person. Likewise, the
“Other’s” choice will produce points for him/her and for you. Every point has value: the
more points you receive, the better for you, and the more points the “Other” receives, the
better for him/her. Here is an example of how this task works:

A B C
You get 500 500 550
Other gets 100 500 300

In this example, if you chose A you would receive 500 points and the other would receive
100 points; if you chose B, you would receive 500 points and the other 500; and if you
chose C, you would receive 550 points and the other 300. So, you see that your choice
influences both the number of points you receive and the number of points the other
receives.

Before you begin making choices, please keep in mind that there are no right or wrong
answers: choose the option that you prefer most, for whatever reason. Also, remember
that the points have value. That is the more of them you accumulate, the better for you.

Likewise, from the “other’s” point of View, the more points he/she accumulates, the
better for him/her.

For each of the nine choice situations, circle A, B, or C, depending on which column you
prefer most:

A B C A B C
(1) You get 480 540 480 (5) You get 560 500 490
Other gets 80 280 480 Other gets 300 500 90
A B C A B C
(2) You get 560 500 500 (6) You get 500 500 570
Other gets 300 500 100 Other gets 500 100 300

81

(3) You get

Other gets

(4) You get

Other gets

(9) You get

Other gets

A

520

520

500

100

480

100

(Measure of Social Values continued)

B

520

120

460

300

490

490

C
580

320

490

490

540

300

82

(7) You get

Other gets

(8) You get

Other gets

A

510

510

A

550

300

B

560

300

500

100

C
510

110

500

500

APPENDIX B

83

APPENDIX B

Competitiveness Index

For the following twenty questions, rate your attitude.

1. I like competition.

Agree_ Disagree_

2. 1 find competitive situations unpleasant.
Agree_ Disagree_

3. I don’t like competing against other people.
Agree_ Disagree_

4. I enjoy competing against an opponent.
Agree_ Disagree_

5. I try to avoid competing with others.
Agree_ Disagree_

6. 1 get satisfaction from competing with others.
Agree_ Disagree_

7. I dread competing against other people.
Agree_ Disagree_

8. I am a competitive individual.
Agree_ Disagree_

9. Competition destroys fiiendships.
Agree_ Disagree_

10. I will do almost anything to avoid an argument.
Agree_ Disagree_

11. I try to avoid arguments.
Agree_ Disagree_

12. I often remain quiet rather than risk hurting another person’s feelings.
Agree_ Disagree_

13. In general, I will go along with the group rather than create a conflict.
Agree_ Disagree_

14. I don’t enjoy challenging others even when I think they are wrong.
Agree_ Disagree_

15. I would like to be on a debating team.
Agree_ Disagree_

16. Games that have no clear-cut winner are boring.
Agree_ Disagree_

17. It’s usually not important to me to be the best.
Agree_ Disagree_

18. I often try to outperform others.
Agree_ Disagree_

19. When I play a game I like to keep score.
Agree_ Disagree_

20. I don’t like games that are winner-take-all.
Agree_ Disagree_

84

APPENDIX C

85

APPENDIX C
Sports Orientation Questionnaire

The following statements describe reactions to sport situations. We want to know how
you usually feel about sports and competition. Read each statement and circle the letter
that indicates how much you agree or disagree with each statement on the scale: A, B, C,
D, or E. There are no right or wrong answers; simply answer as you honestly feel. Do
not spend too much time on any one statement. Remember, choose the letter which
describes how you usually feel about sports and competition.

Strongly Slightly Neither Agree Slightly Strongly
Agree Agree Nor Disagree Disagree Disagree

1. I am a determined competitor. A B C D E
2. Winning is important. A B C D E
3. I am a competitive person. A B C D E
4. I set goals for myself when

I compete. A B C D E
5. I try my hardest to win. A B C D E
6. Scoring more points than my

opponent is very important to me. A B C D E
7. I look forward to competing. A C D E
8. I am most competitive when I try

to achieve personal goals. A B C D E
9. I enjoy competing against others. A B C D E
10. I hate to lose. A B C D E
11. I thrive on competition. A B C D E
12. I try my hardest to win when I

have a specific goal. A B C D E
13. My goal is to be the best athlete

possible. A B C D E

Turn to the next page please.

86

(Sports Orientation Questionnaire continued)

Strongly Slightly Neither Agree Slightly Strongly
Agree Agree Nor Disagree Disagree Disagree

14. The only time I am satisfied is

when I win. A B C D E
15. I want to be successful in sports. A B C D E
16. Performing to the best of my

ability is very important to me. A B C D E
17. I work hard to be successful

in sports. A B C D E
18. Losing upsets me. A B C D E
19. The best test of my ability is

competing against others. A B C D E
20. Reaching personal performance

goals is very important to me. A B C D E
21. I look forward to the opportunity

to test my skills in competition. A B C D E
22. I have the most fun when I win. A B C D E

23. I perform my best when I am
competing against an opponent. A B C D E

24. The best way to determine my ability
is to set a goal and try to reach it. A B C D E

25. I want to be the best every time I
compete. A B C D E

87

APPENDIX D

88

APPENDIX D

Competitive Work Task Prime

Word Puzzle Task

For each of the words listed below, find any smaller words of 3 or more letters
hidden within them. Then write the words you have discovered to the right of the
words below. So for example if you were given the word “glassware” you could list
words such as “glass”, “are”, “lass”, “ass”, and “war”. However, you may not
change the ordering of the letters. Do your best, you may be asked questions about
this task later. Please take your time to complete the task thoughtfully. However,
try to spend no more than 15 minutes on this task. When you are done with this
task, redirect you attention to the computer screen. Good Luck! You may begin.

1. Stadium

2. Dormitory

3. Tournament

4. Rivalry

5. Thoroughly

6. Achievement

7. Bracelet

8. Competition

9. Vanquishing

10. Predicate

Turn to the next page please.

89

ll.

12.

l3.

14.

15.

16.

l7.

18.

19.

20.

21.

22.

23.

24.

25.

(Competitive Word Task Prime continued)

Opposition

Victorious

Indeterminate

Superiority

Medal

Ranking

Challenging

Contemplation

Indefinite

Winning

Championship

Edifice

Conquering

Trophy

Comparison

90

APPENDIX E

91

APPENDIX E

Cooperative Word Task Prime

Word Puzzle Task

This is a task designed to measure your verbal and creative abilities. For each of the
words listed below, find any smaller words of 3 or more letters hidden within them.
Then write the words you have discovered to the right of the words. So for example
if you were given the word “glassware” you could list words such as “glass”, “are”,
“lass”, “ass”, and “war”. However, you may not change the ordering of the letters.
List as many words as you can. Please take your time to complete the task
thoughtfully. However, try not to spend more than ten minutes on this task. When
you are done with this task, turn the page. Good Luck! You may begin.

1. Collaboration

2. Dormitory

3. Interconnected

4. Teamwork

5. Thoroughly

6. Together

7. Camaraderie

8. Bracelet

9. Communal

10. Unification

Turn to the next page please.

92

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

(Cooperative Word Task Prime continued)

Collective

Predicate

Cooperation

Partnership

Indeterminate

Contemplation

Tandem

Cohesive

Indefinite

Networked

Squadron

Edifice

Reciprocity

Tearnmates

Support

93

APPENDIX F

94

APPENDIX F

Neutral Word Task Prime

Word Puzzle Task

For each of the words listed below, find any smaller words of 3 or more letters
hidden within them. Then write the words you have discovered to the right of the
words below. So for example if you were given the word “glassware” you could list
words such as “glass”, “are”, “lass”, “ass”, and “war”. However, you may not
change the ordering of the letters. Do your best, you may be asked questions about
this task later. Please take your time to complete the task thoughtfully. However,
try to spend no more than 15 minutes on this task. When you are done with this
task, redirect you attention to the computer screen. Good Luck! You may begin.

1. Thoroughly

2. Dormitory

3. Transfer

4. Mountain

5. Edifice

6. Cabinet

7. Circumlocution

8. Alleviate

9. Pavement

10. Contemplation

Turn to the next page please.

95

ll.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

(Neutral Word Task Prime continued)

Orange

Indeterminate

Tornado

Predicate

Formica

Indefinite

Beanbag

Interstellar

Meantime

Distraction

Bracelet

Stowaway

Intricacy

Equipment

Menagerie

96

APPENDIX G

97

APPENDIX G

Competitive Story Prime

Word Finding Task

This is a task designed to measure your verbal and creative abilities. Read the
passage below. As you do so, circle any grammatical or spelling errors. Please take
your time to complete the task thoughtfully, however try to spend no more than ten
minutes on this task. Good luck! You may begin.

Michael worked hard to be the best compared to those around him. Everything he
achieved in life was due to his hard work and competitive spirit. Since he was a teenager
Michael enjoyed running. He became very good in it in college. He was always able to
push a little hardest than anyone he ran against in practice, He would give all he had to
finish a step ahead of any other runner he ran against and he knew this was the key to his
success. At the age of 27 Michael finally got his bigbreak. He qualified for the
Marathon World Championships. He started off well on the rural course, running with
the lead pack. Slightly past the halfway point of the race Michael make his decision. He
took the lead at a pace he wasn’t sure he could hold the rest of the way. However, he
knew this was the only way he could win the race because some of the other runners in
the leed pack possessed superioir speed at the end of the race. Only four runners chosen
to go with him. Now this pack of five had broke away form the rest of the field. With
three miles to go Michael had fallen into forth place just two steps behind the leader. He
felt terrible, as is each step would be his last. However, he noticed that the men in front
of him were also laboring and looked very fatiging. He thought back to his arduous
training where his pushed his body and mind to its limits to finish ahead of the others he
trained with. He knew he had trained hardest of all these men. At that moment Michael
found an extra reserve of energy. He surged to the front of the pack running like a man
possessed. The others could not respond. He held a five step lead until the finnish line
were he collapsed. Michael had gave everything he had and was now the world
champion. He had beaten all the best runers in the world. Exhausted, Micheal pulled
himself to his feet to congratulate all of his competitors finishing the race. The media
quickly broadcasted this story. The public embraced Michael as a hero for being a fine
role model as a competitive athlete.

98

APPENDIX H

99

APPENDIX H

Cooperative Story Prime

Word Finding Task

This is a task designed to measure your verbal and creative abilities. Read the
passage below. As you do so, circle any grammatical or spelling errors. Please take
your time to complete the task thoughtfully, however try to spend no more than ten
minutes on this task. Good luck! You may begin.

Michael, Carl, Paul, and Ben worked hard to be the best. They were all members of the
same track club and had trained together since they were children. Everything they
achieved in life was due to there hard work. Last year they finally got their big break.
The team qualified for the Track and Field World Championships. They started off well
at the championships, winning their first two qualifing rounds. Although they had some
sloppy batton passes, the members did not panic because they knew each other’s
strengths and weakness so well that they did easily overcome these errors. The day of the
final had arrived. That morning the team members got together. They did not meet to
practice baton passes or sprints as they had so many times in the pass. Rather they met to
pledge their support to one another. No matter what happens in the race they would
always be the team that stuck together, the team that always had. The dream they had
worked so hard together for was finaly here and they would enjoy it together as a team.
Race time had finally arrived. Michael was the first runner. He got ofl‘ to a fairly good
start, As the other runners cheered him on he seemed to accelerate catching other team
and moving into third place. He had an excellant baton exchange with Carl the second
runner. Carl fed off Michael’s success to hand off the baton to Paul neck and neck for
first place. Now the three who run their legs chanted in unity to Ben the anchor runners.
The louder they chanted the more Ben increased his lead on the field. They had won.
The four ran to embrace one another. Because of the combined efforts of the team the
had done it. No one team member ran his best time but together they worked as a unit to
accomplish there goal and acheived it as a team. The media quickly took notice of their
story. At the awards ceremony, the crowd cheered and cheered. Although the team
members received medals and fame for their acomplishment they all agreed that being
with there team members made it all worthwhile. Being a part of a cohesive, successfull
group was its own reward.

100

APPENDIX I

101

APPENDIX I

Neutral Story Prime

Word Finding Task

This is a task designed to measure your verbal and creative abilities. Read the
passage below. As you do so, circle any grammatical or spelling errors. Please take
your time to complete the task thoughtfully, however try to spend no more than ten
minutes on this task. Good luck! You may begin.

Michael woke up early on Sunday, he had to go to work. He is a timer and ofl‘ical for a
marathon races. This particular Sunday was a very big race. Over 3,000 people were
entered so Michael knew that he would be having his hands firll. The first difficult task
was getting to the race itself. Many of the city’s streets were blocked since they were in
the race course. In order to get to the start line, Michael had to drive through the east side
of town, past the river, and through 3 aditional tolls. This alternate route was about 15
miles longer than the usual root. When he finally got there Michael had to make sure to
park further away from the start then he would’ve liked because he had to make sure he
could get his car out. Upon arrival at the start line, Michael greeted the usual crew of
race oflicials, Cindy Maher, his best fiiend Bill Coats, and the veteran official Fred
Houston who had been an official for 32 years. They all set out on their taks of setting up
cones to block trafiic, greeting volunteers and participants, and setting up the clocks. As
race time drew near, Fred and his crew herded the runners in to the start area which was
not an easy task When all the runners were finally herded in, the officials turned the
crowds attention to Enrico Palatzo, the Italian opera singer who woul sing the national
anthem. Then Michael made sure all the timers were ready and the gun was fired. Once
the start area was cleaned up after about a half hour, Michael and his crew would set off
to the finnish line to record times and finishing place of all the runners. This would be
the longest task of all with this many runeers involved. The last runner usually di not
finish until about 4 hours after the first one so Michael new he had a long day ahead of
him. However, he enjoy getting out in the breezy spring morning. The finish line work
went very smoothly amd all times and place were recorded accurately, a rarity in
Michael’s exerience. When the last runner finally did come, the remaining crew
applaude and cleaned up the finish area. It was almost 5PM by now and Micheal was
ready to go home satisfied after a job well clone.

102

APPENDIX J

103

APPENDIX J
Supplemental Analyses
(Repeated Measures AN OVA for Experimental Trials Examined by Priming Condition
and Arm)
A 2 (Group condition: Conjunctive vs. Coactive) X 2 (Priming condition) X 2
(Arm: dominant vs. nondominant) ANOVA with repeated measures on the latter factor
was performed on the fatigue corrected difference scores for the experimental trials.
Analyses revealed no significant main efl'ect of group (F(], 76) = .315, ns) indicating that
motivation gain did not differ by conjunctive and coactive groups. The Priming main
effect reached statistical significance (F(1, 76) = 5.391, p < .03) indicating that those
primed for competitiveness (M = 70.45, s = 47.31, n= 38) outlasted those not primed (M
= 48.19, = 39.5, n = 42). The Arm main effect was also significant (F(1, 76) = 12.02,p
< .001) indicating that as with controls, those in the experimental conditions fatigued at
different rates with each arm. However, the arm factor did not interact significantly with
Group or Priming conditions (F’s < 1). The Group X Prime main effect was also
nonsignificant (F(], 76) = .515, ns).
To further examine these effects by group condition, 2 (Prime) X 2 (Arm)

AN OVA’s with repeated measures on the latter factor were performed on the fatigue
corrected difference scores. In the conjunctive condition, the effect of Priming was non
significant (F(], 38) = 1.31, ns). The effect ofArm was marginally significant (F(], 38)
= 4.02, p < .06). The Prime X Arm interaction was non significant (F(], 38) = .08, ns).
The mean motivation gain scores plotted by Group and Priming condition are presented

in Figures 3.1 and 3.2. In the coactive condition, the Priming effect was significant

104

(F(1, 38) = 4.54,p < .05). The Arm efl‘ect was also significant (F(], 38) = 12.74,p <
.001). However the Prime X Arm interaction was non significant (F(1, 38) = 1.51, ns).
These results indicate that while the conjunctive and coactive groups did not differ
overall in magnitude of motivation gains, the effectiveness of priming was limited to the

coactive condition but did not interact with the arm factor.

105

REFERENCES

106

REFERENCES

Aronson, E., & Osherow, N. (1980). Cooperation, prosocial behavior, and academic
performance: Experiments in the desegregated classroom. Applied Social
Psychology Annual, 1, 163-196.

Bargh, J. A., & Chartrand, T. L. (2000). The mind in the middle. In H. T. Reis & C. M.
Judd (Eds), Research methods in social and personality psychology (pp. 253-
285). Cambridge, UK: Cambridge University Press.

Brown, S. P., Cron, W. L., & Slocum, J. W. (1998). Effects of trait competitiveness on

salesperson goal setting and performance. Journal of Marketing, 62, 88—98.

Caries, E. M., & Carver, C. S. (1979). Effects of person salience versus role salience on
reward allocation in a dyad. Journal of Personality and Social Psychology, 3 7,
2071-2080.

Elliott, E. S., & Dweck, C. S. (1988). Goals: An approach to motivation and
Achievement. Journal of Personality and Social Psychology, 54, 5-12.

Erikson, E. H. (1980). Identity and the Life Cycle. New York: W. W. Norton & Co. Inc.

Frederick, C. M. (2000). Competitiveness: Relations with GPA, locus of control, sex, and
athletic status. Perceptual and Motor Skills, 90, 413—414.

Gill, D. L., & Dzewaltowski, D. A. (1988). Competitive orientations among
intercollegiate athletes: Is winning the only thing? Sport Psychologist, 2, 212-

221.

107

Gill, D. L., Dzewaltowski, D. A., & Deeter, T. E. (1988). The relationship of
competitiveness and achievement orientation to participation in sport and
nonsport activities. Journal of Sport and Exercise Psychology, 10, 139-150.

Gill, D. L., Kelley, B. C., Martin, J. J., & Caruso, C. M. (1991). A comparison of
competitive-orientation measures. Journal of Sport and Exercise Psychology, 13,
266-280.

Graziano, W. G., Hair, E. C., Finch, J. F. (1997). Competitiveness mediates the link
between personality and group performance. Journal of Personality and Social
Psychology, 73, 1394-1408.

Hellandsig, E. (1998). Motivational predictors of high performance and discontinuation
in different types of sports among talented teenage athletes. International Journal
of Sport Psychology, 29, 27-44.

Hertel, G. (2002). Personal communication with the author

Hertel, G., Deter, C., & Konradt, U. (2001). Motivation gains in computer-mediated work
groups. Manuscript under editorial review.

Hertel, G., Kerr, N. L., & Messé, L. A (1999). Revisiting the Koehler efi‘ect: Does
diversity enhance motivation and performance in groups? Psychologische
Beitrage, 41, 320-337.

Hertel, G., Kerr, N. L., & Messé, L. A. (2000). Motivation gains in performance groups:
Paradigrnatic and theoretical developments of the Koehler effect. Journal of
Personality and Social Psychology, 79, 580-601.

Higgs, S. L. (1972a). The effect of competition upon endurance of college women.

International Journal of Sport Psychology, 3, 128- 1 40.

108

 

Higgs, S. L. (1972b). Personality traits and motor ability associated with observed
competitiveness in women physical education majors. Perceptual and Motor
Skills, 34, 219-222.

Houston, J. M., Carter, D., & Smither, R. D. (1997). Competitiveness in elite professional
athletes. Perceptual and Motor Skills, 84, 1447-1454.

Jackson, D. N., Ahmed, S. A., & Heapy, N. A (1976). Is achievement a unitary concept?

Journal of Research in Personality, 10, 1-21.

Jones, J. W., Neuman, G., Altrnann, R., & Dreschler, B. (2001). Development of the
sports performance inventory: A psychological measure of athletic potential.
Journal of Business and Psychology, 15, 491-503.

Kang, L., Gill, D. L., Acevado, E. 0., & Deeter, T. E. (1990). Competitiveness
orientations among athletes and nonathletes in Taiwan. International Journal of
Sport Psychology, 21, 146-157.

Kerr, N. L. (1990). Reflections on group productivity. Discussant in the Symposium
“Group Performance” (W. Stroebe, Chair), Annual Convention of the Society of
Experimental Social Psychology, Buffalo, NY, Oct. 11-13.

Kerr, N. L., Messé, L. A., Park, E. S., & Sambolec, E. J . (2002). Manuscript in
preparation.

Kerr, N. L., Messé, L. A., Park, E. S., Seok, D., & Lount R. (2002).

Motivation gains under diflerent task demands. Manuscript in preparation.

Kline, T. J. B., & Sell, Y. P. (1996). Cooperativeness versus competitiveness: Initial
findings regarding effects on the performance of individual and group problem

solving. Psychological Reports, 79, 3 55-3 65.

109

Kohler, O. (1926). Kraftleistungen bei Einzel- und Gruppenabeit [Physical performance

in individual and group situations]. Industrialle Psychotechnik, 3, 274-282.

Kohler, O. (1927). Uber den Gruppenwirkungsgrad der menschlichen Korperarbeit und
die Bedingung optimaler Kollektivkraftreaktion [On group efficiency of physical
labor and the conditions of optimal collective performance]. Industrialle
Psychotechnik, 4, 209-226.

Latané, B., Williams, K., & Harkins, S. (1979). Many hands make light the work: The
causes and consequences of social loafing. Journal of Personality and Social
Psychology, 37, 822-832.

Lee, F. A. (1997). Preference for competition in recreation swimming activities: a gender
and developmental analysis. Unpublished master’s thesis, Springfield College,
Springfield, Massachusetts.

Martens, R. (1975). Social Psychology and Physical Activity. New York: Harper and
Row.

Messé, L. A., Hertel, G., Kerr, N. L., Lount, R., & Park, E. S. (2002). Knowledge of
partner’s ability as a moderator of group motivation gains: Exploration of the
Kohler discrepancy effect. Journal of Personality and Social Psychology.

Moely, B. E., Skarin, K., & Weil, S. (1979). Sex differences in competition-cooperation
behavior of children at two age levels. Sex Roles, 5, 329-342.

Neuberg, S. L. (1988). Behavioral implications of information presented outside
conscious awareness: The effect of subliminal presentation of trait information on

behavior in the prisoner’s dilemma game. Social Cognition, 6, 207-230.

110

Olds, D. E., & Shaver, P (1980). Masculinity, femininity, academic performance, and
health: Further evidence concerning the androgyny controversy. Journal of
Personality, 48, 323- 341.

Ruan, N. (1993). Sport achievement orientation of male and female intercollegiate
athletes, intramural athletes, and nonathletes. Unpublished master’s thesis,
Springfield College, Springfield, Massachusetts.

Smither, R. D., & Houston, J. M. (1992). The nature of competitiveness: The
development and validation of the competitiveness index. Educational and
Psychological Measurement, 52, 407-418.

Spence, J. T., & Hehnreich, R L. (1979). Comparison of masculine and feminine
personality attributes and sex-role attitudes across age groups. Developmental
Psychology, 15, 582-583.

Steiner, I. D. (1972). Group Processes and Productivity. New York: Academic Press.

Stoebe, W., Diehl, M., Abakoumkin, G., & Amscheid, R (1990). The Kohler efl'ect:
Motivation gains in group performance. Paper presented at the annual meeting of
the Society pf Experimental Social Psychology, Buffalo, NY, October, 1990.

Stroope, S. L. (1993). Social loafing and crew: Ihe effects of athlete identifiability on
group performance. Unpublished master’s thesis, University of North Carolina at
Greensboro.

Triplett, N. (1898). The dynamogenic factors in pacemaking and competition. American

Journal of Psychology, 9, 507-533.

111

Van Lange, P. A. M., De Bruin, E. M. N., Often, W., & Joireman, J. A. (1997).
Development of prosocial, individualistic, and competitive orientations: Theory
and preliminary evidence. Journal of Personality and Social Psychology, 73,
733-746.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief
measures of positive and negative affect: The PAN AS scales. Journal of
Personality and Social Psychology, 5 4, 1063- l 070.

Williams, K. D., Harkins, S., & Latané, B. (1981). Identifiability as a deterrent to social
loafing: Two cheering experiments. Journal of Personality and Social
Psychology, 40, 303-311.

Williams, K. D., Nida, S. A., Baca, L. D., & Latane, B. (1989). Social loafing and
swimming: Effects of identifiability on individual and relay performance of
intercollegiate swimmers. Basic and Applied Social Psychology, 10, 7 3-8 1.

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.

Wyer, R. S., & Malinowski, C. (1972). Efi‘ects of sex and achievement level upon
individualism and competitiveness in social interaction. Journal of Experimental

Social Psychology, 8, 303 -3 14.

112