TO AVOID FINES return on or before date due.

 

{fl LIBRARY
[Michigan State
L Unlverslty

 

 

PLACE IN RETURN BOX to remove this checkout from your record.

DATE DUE DATE DUE DATE DUE I

 

 

NE

 

 

 

‘ ‘ I . r .
hi4! "w t '7

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_—|l ll

 

 

 

MSU Is An Affirmative ActiorVEquel Opportunity Institution '
chS-DJ

EFFECTS OF GROUP COHESION, BELIEFS OF OWN DESERVING,
AND BELIEFS OF OTHER’S DESERVING ON
COOPERATION IN SOCIAL DILEMMAS
By

David Nathan Sattler

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Psychology

1990

ABSTRACT
EFFECTS OF GROUP COHESION, BELIEFS OF OWN DESERVING,
AND BELIEFS OF OTI-IER’S DESERVING ON
COOPERATION IN SOCIAL DILEMMAS
By

David Nathan Sattler

An experiment investigated hOw group cohesion, an individual’s belief that S/he
deserved to obtain a resource, and an individual’s belief that his/his partner deserved
to obtain a resource affects cooperative behavior in a two—person simulated social
dilemma. Based on Deutsch’s (1975) equity and equality disu‘ibution principles and
Clark and Mills’ (1979) theory of communal and exchange relationships, it was
hypothesized that different allocation norms or distribution principles would be
employed in cohesive and non-cohesive groups. Beliefs of own and other’s deserving
were manipulated by varying the amount of time an individual and his/her partner
participated in a task. Subjects harvested lumber from a Simulated community foreSt
and shared equal responsibility for fines levied for overzealous resource use. The
hypotheses were supported. Individuals in cohesive groups appeared to employ the
equality distribution rule and individuals in non-cohesive groups appeared to employ
the equity distribution rule. Implications of the findings are discussed, and future

research is proposed.

Copyright by
IMAVHI>bhATTLAhISAUFTLIHt

1990

To my family

iv

ACKNOWLEDGMENTS

It is a special pleasure to thank the many individuals who have contributed to this
dissertation and to my graduate education:

To Professors Norbert L. Kerr and Lawrence A. Messé, mentors and friends, it is
with deep respect and admiration that I thank you both for your guidance, support,
time, and constructive suggestions on this project and throughout my graduate
education. I feel very fortunate to have had the opportunity to work with both of you;

to Professors Galen Bodenhausen, Dan Ilgen, and Joel Aronoff, thank you for your
constructive suggestions and expertise;

to the wonderful team of undergraduate students who served as experimenters on
this project, thank you for your dedication and comradery. This project could nOt
have been completed without your help. Thank you Brian Gussin, Beth Hockman,
Matt Hunt, Debbie Jacobs, Leemor Katz, Jeff Skore, and Mary Villegas;

to Professor Gordon Wood, Department Chair, thank you for the opportunity to
teach introductory social psychology and introductory statistics, and for allowing me to
discover the joys of teaching;

to Vicki Alexander, Sharon Anderson, Seka Davis, Cheryl Mollitor, Suzy Pavick,
and Hank Wieferich--part of an exceptional support Staff at the psychology
department--thank you for your warmth, friendship, and assistance throughout my stay

at MSU;

to fellow social/personality graduate students Rick Hamish, Geoff Kramer, and Dr.
Linda Sullivan, "jolly jokers," Uncle Milton Ant Farm aficionados, and all around
great people, thank you for your friendship; and to former i/o graduate Student Dr. Jeff
Vancouver, see you on the racquetball court.

A diverse group of individuals in San Diego helped me to keep perspective and
have enjoyable breaks:

To Peter Ballantyne and the Staff at the Mission Bay Aquatic Center, see you on
the course; to Julie Phalen and Angie Hougan, thanks for your support;

to Dr. Alden "Doc" Holmes and Mr. Clyde L. Glasson, Scoutrnaster and
Hikemaster respectively of Boy Scouts of America Troop 974 in San Diego, two wise
men and incredibly good friends for the past 16 years, I look forward to many more
years hiking with you and the trOOp in the mountains and deserts of southern
California;

to my family--my parents, Dr. Jerome M. Sattler and Dr. Virginia R. Shabatay, I
love you both very much. You have been constant sources of support, encouragement,
and inspiration. Thank you for going through this experience with me; and to my
sister Heidi Sattler, LCSW, brOther-in-law Walter Phillips, LCSW, Debbie, Keith,

Bonnie, and Yehuda, thank you for your support.

vi

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES
CHAPTER I: INTRODUCTION
Social Dilemma Research Paradigms
Prisoner’s Dilemma and N-Person Dilemma
Social and Temporal Traps
Nontechnical Solutions to Social Dilemmas .

Focus of Present Research .

Deserving for a Resource, Resource Distribution, and Distributive Justice .

Distribution Principles and Norms
Experimental Research on Deservingness

Social Ties and Social Dilemmas
Experimental Research on Social Identity

Hypotheses

CHAPTER II: METHOD

Design

Subjects .

Experimental Task

Procedure

vii

xi

11

14

17

19

25

27

27

27

28

29

TABLE OF CONTENTS (cont’d)
CHAPTER III: RESULTS
Manipulation Checks
Group Cohesion
Own Belief of Deserving
Beliefs of Other’s Deserving
Check for Sample Differences and Sex Differences on Harvest Trials
Harvesting Behavior .
Impressions of Partner
Additional Findings .
CHAPTER IV: DISCUSSION

Conclusion

. FOOTNOTES

REFERENCES
APPENDIX A: EXPERIMENTAL MATERIALS
Departmental Research Consent Form .
Questionnaire Booklet . . .-
Initial Ranking Form
Final Ranking Form .
Winter Survival Questionnaire
Feedback Sheets
Questionnaire
APPENDIX B: EXPERIMENTAL INSTRUCTIONS

viii

35

35

35

37

39

41

42

59

59

61

71

73

74

86

86

87

94

95

96

97

99

103

TABLE OF CONTENTS (cont’d)
APPENDIX C: ANOVA RESULTS ON THE GROUP COHESION

MANIPULATION CHECK ITEMS . . . . . . . 118

ix

LIST OF TABLES

Factor Loadings on Questionnaire Designed to Measure Group Cohesion

ANOVA Means and Standard Deviations for Partner Deservingness x
Class Interaction on Partner Deserving Manipulation Check

ANOVA Means and Standard Deviations for Cohesion x Own
Deserving Interaction on Harvest Trials

ANOVA: Cohesion x Own Deserving x Partner Deservingness on
Harvest Trials: Between-Subject Factors.

ANOVA: Cohesion x Own Deserving x Partner Deserving x Trials:
Effects of Repeated Measures

Means and Standard Deviations for Partner Deserving x Trials Simple
Interaction for High Own Deserving Condition for Eight Trials .

Means and Standard Deviations for Partner Deserving x Trials Simple
Interaction for High Cohesion Condition for First Six Trials

Means and Standard Deviations for Partner Deserving x Trials Simple
Interaction for High Own Deserving Condition for First Six Trials

Factor Loadings for Partner Impression

40

45

45

53

55

55

1.

2.

LIST OF FIGURES

Cohesion x Partner Deserving x Trials Interaction

Own Deserving x Partner Deserving x Trials Interaction

xi

48

50

CHAPTER I

INTRODUCTION

Individuals in a myriad of situations frequently must choose between actions that
will either promote their self-interest and in turn harm the collective or promote the
collective-interest but at personal expense. Examples abound. A conflict between
individual and collective interests exists in labor organizations, where individuals can
gain the benefits of a union without paying dues, on sinking ships, where individual
rationality dictates that people rush to save themselves while risking the lives of
everyone, and in resource use (e.g., petroleum), where individuals can gain from using
as much of the resource as possible even though this actions leads to pollution and
elimination of the resource. Such situations, known in social psychology as social
dilemmas, exist when self-interested behavior appears rational from an individual’s
perspective, but if all act in their self-interest, then the result is a less desirable
outcome than if all had acted in the collective-interest (Dawes, 1980). Social
dilemmas are characterized by two properties: (a) each individual receives a higher
outcome for choosing a self-interested option than for selecting a
collectively-interested option, no matter which options the other individuals in the
group select, but (b) all individuals in the group receive higher outcomes if they select

the collectively-beneficial option than if they choose the self-interested option (Dawes,

1980).

The conflict between individual and collective interests was characterized by
Garrett Hardin (1968) in his classic article "The Tragedy of the Commons." Hardin
explored the situation in which a number of herdsmen, sharing use of a common
pasture, muSt decide how many cattle to graze on the commons. Each herdsman
reasoned that it is in his best interest to keep adding animals to his herd, for the
personal gain from adding animals outweighs his proportionate Share of the damage
inflicted on the commons, which is shared equally by all. Since the commons is
finite, the day must come when the total number of cattle will cause the pasture to
deteriorate and eventually destroy the resource on which the herdsmen depend. And
therein lies the tragedy: the logic of the commons dictates that the collective
consequences of self-interested behavior result in an undesirable outcome for everyone.

Social Dilemma Research Paradigms
Prisoner’s Dilemma ad N-mrson Dilemma

Social psychologists have invoked two main research paradigms to Study the
conflict between individual and collective interests: the prisoner’s dilemma (Luce &
Raiffa, 1957) and the N-person dilemma (or social dilemma) (e.g., Kramer & Brewer,
1984; Orbell & Dawes, 1981). In both paradigms, (a) an individual may make a
choice that benefits either him or herself or the collective, (b) an individual’s
outcomes are influenced by his or her own choice and the choices made by the other
group member(s) (i.e., both paradigms are considered to be a class of outcome

interdependent Situations (Kelley & Grzelak, 1972)), (c) it is assumed that individuals

3

are "rational" and attempt to maximize their own interests, and (d) the main dependent
variable is whether an individual acts in his or her own interest or in the
collective-interest. In addition, several psychological factors may lessen any incentives
to act in the collective-interest, especially when it is not clear whether Others will act
in their self-interest or in the collective-interest. Examples of such factors include
motivation losses (Kerr, 1983), social loafing (Latané, Williams, & Harkins, 1979),
diffusion of responsibility (Latané & Darley, 1970), and feelings of inefficacy (Kerr,
1989).

There are at least three differences between the prisoner’s dilemma and the
N-person dilemma. First, in the prisoner’s dilemma, negative consequences resulting
from self-interested behavior are focused on one individual but are Spread out among
at least two individuals in the N-person dilemma. Second, choices are n0t anonymous
in the prisoner’s dilemma, whereas choices may be anonymous in the N-person
dilemma. Finally, in the prisoner’s dilemma each individual has some degree of
influence over the other individual, but in the N-person dilemma it is difficult, if not
impossible, for one individual to influence or control either anOther’s behavior or fate
(Thibaut & Kelley, 1959; Dawes, 1980; Komorita, 1976).

m and 1m T_raE§

In addition to the conflict between individual and collective interests, many social
dilemmas also contain a conflict between short-term and long-term interests. Platt
(1973) refers to the conflict between individual and collective interests as a social trap,

and Messick and McClelland (1983) refer to the confliCt between short-term and

4

long-terrn interests as a temporal trap. Real world examples of Situations where the
pursuit of short-term self-interested goals can adversely affect long-term
collective-outcomes are commonplace (Schelling, 1971). As an example, Schelling
(1971) describes the decay of rail service when individuals began using their own
automobiles. As more individuals drove their own automobiles, traffic congestion
increased and the quality of the railroad decreased. In time, it appeared that the
efficient railroad of the past was preferred to the automobile.

Messick and McClelland (1983) investigated how social and temporal traps affect
collectively-interested behavior in a social dilemma by manipulating the number of
individuals Sharing a resource. Subjects experienced either a temporal trap (those in
the one-person condition; Since there was no group in this condition, a social trap did
not exist) or both a social trap and a temporal trap (those in the three- and six-person
conditions). Fifty-Six introductory psychology Students were told that their task was to
withdraw as many points as possible from a regenerating resource pool without
depleting the pool over a series of trials. The pool was replenished at the end of each
trial by one-third the number of pOints withdrawn on the proceeding trial. The results
showed that the subjects in the one-person condition maintained the pool longer than
those experiencing both the social and temporal traps. The findings suggest that social
and temporal traps have differential effects on collectively-interested behavior and
either may or may not occur concurrently in either a prisoner’s dilemma or a social

dilemma.

5

Nontechnical Solutions to Social Dilemmas

In an excellent review of the social dilemma literature, Messick and Brewer
(1983) identified two types of "strategies" to solve social dilemmas. These strategies,
or nontechnical solutions, attempt to persuade individuals to consider collective
outcomes in their decision-makings. Structural solutions attempt to eliminate or alter
the incentive structure for self-interested behavior by making either self-interested
behavior more costly than collectively-interested behavior or collectively-interested
behavior more attractive than self-interested behavior. For example, high electricity
and water rates can be considered Structural solutions since they quantitatively change
the incentives for use of the resource (Pope, Stepp, & Lytle, 1975). Hardin’s (1968)
"mutual coercion, mutually agreed upon" solution in which group members agree to
establish a strong central authority to constrain self-interested behavior also can be
considered a structural solution.

Individual solutions attempt to influence individuals to voluntarily act in the
collectively-interest. For example, providing information about the consequences of
self-interested behavior (Stern, 1976) and making appeals to individuals to conserve a
resource such as water (Orbell & Dawes, 1981; Maki, Hoffman, & Berk, 197 8; Sattler
& Kerr, in press) can be considered individual solutions since they attempt to
influence an individual’s decision-making. Ingroup or social identity may be an
individual solution variable, for "the sense of membership in a common group or
social category probably...influence[s] individuals’ willingness to exercise personal

restraint in the interests of the collective welfare" (Messick & Brewer, 1983, p. 27).

6

Another individual solution variable may be individual differences in orientation
to maximize an individual’s own outcomes or collecrive outcomes. Such differences,
known by a variety of names including social motives (McClintock, 1972),
motivational orientations (Kuhlman & Marshello, 1975), and value orientations
(Liebrand, 1983) have been found to influence behavior in both the prisoner’s dilemma
(e.g., Kuhlman & Marshello, 1975) and the N-person dilemma (e.g., Knight & Dubro,
1984; Liebrand, 1984; Messick & McClintock, 1968). The composition of the group
with regard to social motives has been found to influence the likelihood that an
individual will act in the collective interest. For example, Kelley and Stahelski (1970)
found that competitors (competitors attempt to maximize the difference between their
own and others’ outcomes) tend to act in their own interest regardless of their
partner’s behavior, whereas cooperators (cooperators attempt to maximize their own
and others’ outcomes) assimilate to their partner’s behavior. Thus, the composition of
the group with regard to social motives (i.e., the symmetry of social motives within a
group) appears to affect collectively interested behavior.

Focus of Present Research

Many theoretical approaches to social dilemmas assume that "individuals affected
by common pool problems are self-interested, rational" (Ostrom, 1977, p. 174), and
attempt to realize the largest outcomes possible for themselves. With some exceptions
(e.g., Messé, Kerr, & Sattler, in press; Wilke, de Boer, & Liebrand, 1986), few social
dilemma studies have manipulated self-interest. Typically, social dilemma‘studies

have investigated groups whose members have the same inputs or beliefs of

7

deservingness for the shared resource. The present study investigates how individuals
might act in a social dilemma when self-interest (e.g., beliefs of deserving) is
particularly salient for one or more group members (e.g., symmetric beliefs of
deserving: own and others’ beliefs are high; own and others’ beliefs are low; and
asymmetric beliefs of deserving: own belief high, Others’ low; own belief low, Others’
high). Do individuals M assign more weight to their own outcomes (i.e., act in
their self-interest), or are there conditions in which individuals will assign equal or
more weight to others’ outcomes in a social dilemma? Might intragroup social
relations moderate the degree to which individuals act (a) to further only their own
interest, (b) to further the interests of all group members, including themselves, or (c)
to further the interest of only their fellow group members? Thus, the main focus of
the present research are the interactive effects of intragroup social relations, an
individual’s beliefs of his or her own deserving for a resource, and an individual’s
perception that his or her fellow group member deserves a resource on cooperative
behavior in a social dilemma.

Situations in which some portion of individuals sharing use of a resource (ranging
from zero to 100%) believe that they deserve to obtain the resource are related to
situations involving resource scarcity. Scarcity exists when there is an insufficient
supply to meet demand for the resource. The perception that resource scarcity exists
is usually created when the supply available dwindles and is no longer able to meet
demand. A handful of social dilemma studies have manipulated scarcity by varying

the resource supply (e.g., Rutte, Wilke, & Messick, 1987). These Studies tend to

8

demonstrate that self-interested behavior increases as the resource supply decreases.
However, scarcity is also created when demand to acquire the resource exceeds the
available supply. If scarcity can be manipulated by keeping the resource supply
constant and varying demand (i.e., by creating beliefs that either all or a subset of
group members deserve to obtain the resource whose supply is fixed), then would
self-interested behavior increase as when demand remains constant but supply
decreases? A review of the literature did not reveal any social dilemma studies
manipulating scarcity by varying demand for the resource.
Deserving for a Resource, Resource Distribution,
and Distributive Justice

Social dilemmas and the domain of distributive justice are concerned with how
individuals distribute goods that affect an individual’s own and others’ psychological,
physiological, economic, and social well-being (cf. Deutsch, 1975). Distributive

justice refers to the fairness of resource allocation among individuals and can be

 

characterized by three properties: (a) the fairness of allocation decisions are
detemrined by comparing an individual’s own outcome to another’s outcome
(Festinger, 1954), (b) to make meaningful comparisons between an individual’s own
outcome and another’s outcome, the other individual must be perceived as similar on
relevant dimensions, and (c) if an individual’s own outcome is not equal to Others’
outcomes, then it is expected that the individual allocating the resource will justify the
allocation decision (Kamiol & Miller, 1981). Thus, one aspect of distributive justice

is concerned with whether an individual allocates resources in his or her own

self-interest or in the collective interest.

Several theorists have speculated about the conditions that influence resource
distribution. For example, Deutsch (1975, 1982) and Schwinger (1981, 1986)
suggested that the values or principles an individual employs to distribute resources is
influenced by (a) the type of interpersonal relationship existing among group members,
and (b) the environmental circumstances facing the group. In addition, Deutsch (1975)
indicated that "there are undoubtedly some minimal conditions of individual well-being
and human dignity which are necessary to sustain continued cooperative participation
in a group’s activities and vice versa. An individual can tolerate only a certain degree
of inconsistency, rejection, isolation, abuse, or terror from his [or her] group before he
[or she] no longer will be willing or competent to cooperate; his [or her] threshold of
tolerance for such practices will undoubtedly decrease if he [or She] sees that others
arenot treated similarly" (p. 141) (see Perelman, 1967, p. 51).

Deutsch (1975), Lerner (1975), and Leventhal (1976) proposed three principles
regarding the conditions-environmental circumstances and interpersonal relationS--that
affect the just or fair distribution of resources. Each principle addresses the degree to
which individuals may weigh their own and others’ inputs and outcomes in their
decision-making. The first principle states that in impersonal and competitive social
situations in which economic productivity is a primary goal, equity will be the
dominant principle used to distribute resources. The equity principle States that
individuals should receive outcomes in proportion to their respective contributions to

the group. For example, all other variables being equal, an individual who works at a

10

task longer than another individual should receive more wages. Individuals might also
consider outcomes to be equitable when higher status individuals--who supposedly
have higher inputS--obtain larger outcomes than lower Status individuals-who
supposedly have lower inputS--in a social dilemma (see Messé, Kerr, & Sattler, in
press).

The second principle states that in social situations that can be characterized by
"face-to.face contact" (Deutsch, 1975, p. 148) between group memberS--provided that
none of the group members are considered to be dependent upon another--in which the
fostering or maintenance of enjoyable social relations is the common goal, the
dominant principle to distribute resources will be equality. The equality principle
states all group members should Share equally the group’s outcome. For example,
consider the situation in which two individuals are neighbors and are working together
to paint a house for pay. The equality principle would indicate that payment for
painting should be distributed in equal amounts to both individuals, even if one
individual worked a few more hours than the other.

The third principle states that in social relations in which a strong degree of
friendship exists within the group in which the fostering of personal development and
personal welfare is the common goal, the dominant principle to distribute resources
will be need. The need principle states that all individuals Should receive outcomes in
proportion to their respective need for the resource. For example, consider the
situation in which a married couple becomes lost on a backpack trip and have a

limited supply of water. The need principle would indicate that the thirstiest person

11

should be allotted the most water. "Heightening the relative needs of group members,
heightening the sense of responsibility in relation to the needs of the others, etc. would
lead to the use of need rather than equalin or equity as the principle of distributive
justice" (Deutsch, 1975, p. 149).

For Deutsch, Lerner, and Leventhal, the way in which resources will be
distributed among group members is largely influenced by the social ties or relations
existing among the group members. If social ties among group members range from
weak to strong, then as social ties become stronger, individuals Should become less
concerned about their own outcomes and more concerned about the outcomes of their
fellow group members.

Distribution Principles gig _No_rm_s
The equity, equality, and need principles may be considered as defining different

types of Standards, rules, or norms for appropriate behavior in various conditions. A

 

norm is an expectation about how one ought to act that is enforced by the threat of
sanctions or the promise of reward (Schwartz & Howard, 1982) (see Kerr (in press)
for an excellent discussion on the role of norms in social dilemmas). Two types of
social normsunorms in which the consequences of adherence to and violation of
norms are administered by an actor’s fellow group members (Schwartz, 1977)-—are
relevant to these principles.1 First, benevolence norms prescribe actions that benefit
others directly and may "prescribe cooperation, sacrifice, and trust within a
group...[and are] implicit in such notions as ’teamwork,’ ’allegiance,’ ’loyalty,’

’patriotism,’ ’nationalism,’ ’solidarity,’ and ’community’ (Kerr, in press). Second,

12

gem interaction n_o_rrns_ "promote group member welfare indirecfly, by providing
fairly general rules of interaction and exchange" (Kerr, in press). Examples of general
interaction norms include the norm of equity, equality, and reciprocity. The norms of
equity and equality also may be considered to be types of allocation no_r_m_§ that
involve "a social rule which specifies criteria that define certain distributions of
rewards and resources as fair and just" (Leventhal, 1976, p. 94). Such allocation
norms have also been referred to as fththe ng_r*_nrs__ (cf. Sampson, 1969, 1975; Deutsch,
1975; Leventhal, 1976; Lemer, 1975, 1977). The equity and equality general
interaction norms are most relevant to the equity and equality principles.

The equity, equality, and need principles suggest that situational and social
factors will influence the allocation or justice norms utilized. As Leventhal (1976)
suggested, "Prevailing rules and practices may encourage the allocator to follow
certain norms and ignore others" (p. 95). Thus, different allocation or justice norms
may be more appropriate in some situations than in other situations (e.g., equity norm
in social relations with weak social ties and the equality norm in social relations with
strong social ties).

Clark and Mills’ (1979) theory of communal and exchange relationships--
stimulated by Goffman’s (1961, p. 275) differentiation between social and economic
relationships-also explores how interpersonal relations moderate interpersonal
behavior. According to Clark and Mills (1979), in an exchange relationship members
assume that an individual who receives a benefit from another individual accepts a

debt or obligation to repay the individual. "A benefit can be anything a person can

13

choose to give to another person that is of use to the person receiving it" (Clark &
Mills, 1979, p. 12). In a communal relationship, each individual is concerned with the
welfare of the other and observes the "norm of mutual responsiveness" (Pruitt, 1972).
"In a communal relationship, the receipt of a benefit does not create a Specific debt or
obligation to return a comparable benefit" (Clark & Mills, 1979, p. 13). In addition,
"when a communal relationship does not yet exist but is desired, the receipt of a
benefit should have the same effect as when a communal relationship is assumed to
exist" (Clark & Mills, 1979, p. 13). These researchers conduCted two experiments that
explored how exchange and communal relationships and the benefits received
influence interpersonal attraction. The results indicated that interpersonal attraction
decreased when an individual treated a communal relationship as an exchange
relationship (e.g., in a communal relationship it was inappropriate for a receiver of a
benefit to respond by giving a benefit). That is, counter-normative behavior resulted
in decreased attraction for the individual.

Other researchers have highlighted additional forces that may influence
individuals to act in their self-interest at the expense of fair or just resource
distribution. Mikula (1974) and Leventhal and Anderson (1970) speculated that if
resource allocation is conducted by only one individual in a group, then motives such
as self-interest may influence the allocation decisions. Others have speculated the
existence of a self-maximizing justice motive (e.g., Walster, Berscheid, & Walster,
1973). Such self-interest morives may influence an individual to attempt to maximize

his or her own outcomes, even if it means distributing resources inequitably.

l4

Tomblom (1977) speculated that people are motivated to avoid injustices to
themselves more than to others. Lenski (1966) suggested that when individuals are
forced "to choose between their own [interest], their group’s interest, and the interests
of others, they nearly always choose the former" (p. 30). Finally, Homans (1961)
Stated that individuals are likely to emphasize the worth of the characteristics and the
behavior in which they are most outstanding. A variety of forces may be present in
any given situation to influence individuals to act in their self-interest.

Experimental Research o_n Deservingness

Research in the domain of distributive justice has investigated how beliefs about
deservingness effects resource distribution. These studies typically manipulate beliefs
of deservingness by varying either the amount of time various group members work at
a task or the amount of money an individual initially receives (e.g., Pepitone, 1971;
Lane & Messé, 1971; Wilke et al., 1986). Individuals who work longer at a task or
initially receive more money purportedly believe they deserve more of the resource
than those who work less or initially receive less money. Typically, the dependent
variable is how an individual distributes a resource (e.g., money) between him or
herself and another individual.

Pepitone (1971) investigated how resources are distributed among group members
when one member of a dyad believes s/he deserves to obtain the resource (i.e.,
asymmetric beliefs of deservingness existed within the group). One member of a dyad
initially received a payment of money either as a reward for success at a task (i.e., the

merit condition) or due to random chance (i.e., the arbitrary condition). On each of 50

15

trials, subjects were presented with various distributions of money to self and anOther
individual. Subjects then selected the distributions that were most preferable to them.
Thus, within the merit and the arbitrary conditions, for each dyad one subject was
rewarded and the other subject was not rewarded. The results indicated that the
rewarded subjects distributed more resources to themselves than did the non-rewarded
subjects in the merit condition. In contrast, non-rewarded subjects distributed more
resources to themselves than did the rewarded subjects in the random change
condition. The results illustrate how different conditions influence the way in which
individuals distribute resources and attempt to achieve equitable outcomes.

Lane and Messé (1971) investigated how resources are distributed as a function
Of the amount of time individuals work at a task and the available monetary resources
to pay the individuals for their work. Seventy-nine male undergraduate students
worked on a task for either 90 minutes (high-input condition), 30 minutes (low-input
condition), or 30 minutes in which subjects were told that they were expeCted to
perform poorly (explicit low input condition). After working on the task for the
allotted time, subjects were given money to disuibute between themselves and one
co-worker. The money served as payment for working on the task. A pretest
determined the amount of pay introductory psychology students considered fair for the
type of work performed in the study (i.e., the standard of fair pay). The results
indicated that when the amount of money available for distribution was congruent with
the subject’s standard of fair pay, subjects with co-workers who worked an equal

amount of time at the task divided the money equally between themselves and the

16

co-worker. In contrast, when the amount of money available was either substantially
more or substantially less than the subject’s standard of fair pay, subjects allocated
more money to themselves. That is, when there was a sufficient amount of money to
pay borh workers, subjects distributed the money equally. But when there was either
an insufficient amount of money or too much money to pay borh workers, subjects
took more money for themselves. The results suggest that several factors influence
resource distribution, including internal standards (e.g., norms) of juStice, comparisons
with Others (i.e., time inputs), and circumstances external to the group (i.e., the
resource supply). Whereas the dictates of the principle of equity were clear when the
individuals had equal inputs and the resource supply was congruent with the
individuals’ internal standard of fair pay, when the resource supply was not congruent
with the internal Standards it is possible that subjects simply did not have a relevant
norm or principle to guide their distribution. If so, then under such conditions
self-interest appears to be employed. In addition, the findings suggest that when the
supply of a resource is insufficient to meet demand (e.g., if own beliefs of deserving
for a resource are high), individuals act in their self-interest.

Other researchers have explored how beliefs concerning group members’ need for
a resource, as opposed to an individual’s own need, effecrs resource distribution.
Karuza and Leventhal (1976) investigated the effecr of need of anorher person on
allocation decisions. Subjects allocated food to hypothetical children who had either
low need or high need for food. The results showed that subjects gave more food to

children with high need than to children with low need. Leventhal and Weiss (1975)

l7

explored how the need of a co-worker influenced the distribution of monetary
payments for work performed. Subjects were paired with a bogus co-worker and
worked on a task. The money the subjects earned was Shared by bOth individuals.
Upon completion of the task, subjects were presented with feedback indicating that
they and their co-worker had performed similarly on the task. Additional information
indicated that the bogus co-worker had either low need or high need for the money.
Subjects then distributed the money between themselves and their co-worker. The
results showed that subjects distributed more money to co-workers with high need than
to co-workers with low need for the money. Although the Studies investigating need
did not manipulate subjects own need for the resource, the findings suggest that under
certain conditions, subjects may distribute more resources to another individual, even
though it may not be equitable to do so.

Together, these studies raise a fascinating question regarding resource
distribution: (a) are the equity and equality principles relevant to behavior in social
' dilemmas, and if so, (b) which principle will be used to distribute resources in a social
dilemma when group members’ deservingness for the resource is either symmetric or
asymmetric and the social ties existing among group members are either strong or
weak?

Social Ties and Social Dilemmas

The majority of social dilemma researchers assume that individuals are "rational"

and as such attempt to realize the greatest outcome possible for themselves. Thus,

most researchers assume that self-interest occurs at the individual level, as Opposed to

18

the group level. For example, Coleman (1961) noted that "classic economic theory
always assumed that the individual will act in his [or her] interest; but it never
carefully examined the entity to which his [or her] refers...in many situations men [and
women] act as if the his [or hers] refers to some entity larger than themselves. That
is, they appear to act in terms, not of their own interest, but in the intereStS of a
collectivity" (p. 24). Are individuals in groups with strong social ties or bonds more
likely to act in the collective interest than in their own self-interest?

Brewer and Kramer (1986) suggest that individuals in groups with high social
identity may act in the collective interest because they may assign "greater weight...to
joint (collective) gains over individual gains alone...making it less likely that they will
make sharp distinctions between their own and others’ welfare" (p. 545) (see Brewer,
1979, 1981; Tajfel & Turner, 1986). ReseaIch tends to support this position. For
example, field studies conducted by Levine and Campbell (1972) and Brewer and
Campbell (1976), and Brewer’s (1979) laboratory studies demonstrated that individuals
who are part of a social group tend to perceive other group members in a generally
favorable or positive light. Brewer and Silver (1978) found that individuals perceived
other members of their group as more likely to possess traits such as trustworthiness,
honesty, and cooperativeness.

Campbell (1958) proposed a second possible explanation for the group enhancing
effects of social identity. He suggested that collectively interested behaviOr may be a

consequence of cognitions associated with unit formation, as opposed to evaluative

biases and expectations of behavior associated with ingroup favoritism. Research by

l9

Brewer (1979) lends support to this speculation. Brewer (1979) found that when
individuals are part of a "common social boundary," the social distance among group
members is reduced and they are less likely to make sharp diStinctionS between their
own welfare and others’ welfare. AS such, an individual may assign equal weight to
his or her own outcome and to his or her fellow group members’ outcomes. In
addition, being identified as part of a social category may function as a "rule for
defining the boundaries of low risk interpersonal trust [e.g., depersonalized trust] that
bypasses the need for personal knowledge and the costs of negotiating reciprocity with
individual others" (Brewer, 1981, p. 356). One consequence of depersonalized trust
may be that individuals are more likely to act in the collective interest because they
believe other group members will reciprocate with cooperative behavior (see Messick,
Wilke, Brewer, Kramer, & Zemke, & Lui, 1983).
Experimental Research 9p Mal gen—tity

Social dilemma researchers exploring the effects of social ties on cOOperation
have typically done so by manipulating social identity. The social identity
manipulations usually involve varying whether or not an individual identifies with a
social category, such as a college or club. Although only a handful of Studies have
been conducted, they tend to Show that individuals who identify with their group are
more likely to act in the collective interest. Kramer and Brewer (1984) performed a
series of three experiments to investigate the effects of social identity on collectively
interested behavior in a social dilemma. Social identity was manipulated with

commonplace social categories such as age, association with an institution, and

20

common fate of the group members. In all experiments, subjects participated in a
simulation of a commons dilemma in which all decisions were anonymous. Subjects
were told that their goal in the task was to withdraw as many points as possible from
a resource pool in such a way that the resource would last as long as possible. The
initial size of the resource pool was 300 points and the pool size could not exceed 300
points. The pool was regenerated at the end of each trial by multiplying the number
of points remaining in the pool by 1.1. Each subject could withdraw between 0-10
points on each trial. The results of these experiments tended to indicate that
individuals acted in the collective interest under conditions of high social identity more
than under conditions of low social identity, even when the size of the resource pool
decreased. Kramer and Brewer concluded that high social identity increases
cooperative behavior when the supply of a Shared resource is insufficient to meet
demand. Note that scarcity was created in this paradigm by manipulating the resource
supply, not demand for the resource. As was suggested earlier, it is not clear whether
the effects of resource scarcity on collectively interested behavior are the same when
the cause of scarcity is due to a decrease in supply or an increase in demand.

Brewer and Kramer (1986) investigated how social identity, group size, and
framing of a social dilemma affect collectively interested behavior. Subjects in either
eight or 32-person groups participated in either a public goods task in which they
made investments in order to provide a resource or a commons dilemma task in which
they withdrew points from a resource supply that was Shared by all members. Social

identity was manipulated by varying the common fate of the group members. The

21

task was similar to the one used by Kramer and Brewer (1984), although the initial
pool size varied as a function of group size. The results Showed that subjects
participating in the commons dilemma task were more COOperative than those
participating in the public goods task. Social identity influenced behavior when the
resource became depleted for subjects in the large group conditions. Specifically,
compared to subjects in the low social identity condition, subjeCIs participating in the
commons dilemma acted in the collective interest more when they were in the high
social identity condition. In contrast, subjects participating in the public goods task
and small group condition acted in the collective interest when they were high and low
in social identity.

Fewer researchers investigating how social ties affect cooperation in a social
dilemma have done so by manipulating group cohesion (for a review of research
investigating cohesion on behavior in the prisoner’s dilemma, see Pruitt and Kimmel,
1977). Group cohesion may be defined as "the resultant of all the forces acting on the
members to remain in the group" (Cartwright & Zander, 1960, p. 74). "Thus, the
members of a highly cohesive group, in contrast to one with a low level of
cohesiveness, are more concerned with their membership and are therefore more
strongly motivated to contribute to the group’s welfare, to advance its objectives, and
to participate in its activities. Cohesiveness contributes to a group’s potency and
vitality" (Cartwright, 1968, p. 91).

While similarities exist between social identity and group cohesion, there clearly

are differences. For example, social identity may be considered as mainly a cognitive

22

phenomenon. For social identity to exist, there need nOt been any interaction between
members of the "group." Indeed, "group members" need not see the Others or expect
future interpersonal interaction. The perception that an individual is part of a social
category is the sufficient condition for social identity. In contrast, group cohesion
contains more of an affective component than social identity and usually involves
social interaction among group members, in addition to numerous other components
that may stem from interaction (e.g., success/failure on a task, interpersonal attraction,
etc.). Such differences between social identity and group cohesion raise the possibility
that the way in which social ties are operationalized in a social dilemma might effect
cooperative behavior. For example, various norms might be more salient for members
of cohesive groups than for individuals who identify with a social category (e.g., Clark
& Mills, 1979).

Studies investigating group cohesion in social dilemmas and in the prisoner’s
dilemma do not provide a definitive answer regarding the effect of cohesion on
COOperation. AS discussed below, many of these Studies suffer from methodological
problems that confound the cohesion manipulation with other powerful factors. Baird
(1982) explored how group cohesion and previous experiences in which individuals
Shared a resource affects COOperation in a commons dilemma. Cohesion was
manipulated by using Back’s (1951) technique in which subjecrs are provided with
bogus .test results that indicate group members are either very similar and congenial or
are not similar and scored very differently on most items. Prior sharing was

manipulated by having subjects work alone on a crossword puzzle. In the prior

23

sharing condition, a monetary reward for the number of correct words was determined
by averaging words across group members, whereas in the no prior Sharing condition
the reward was determined by an individual’s own performance. The commons task
was a modified version of Edney and Harper’s (1978) game in which three-person
groups withdrew quarters from a bowl of quarters that was replenished at the end of
each trial by a factor of two. The results indicated that the manipulation check for
cohesion was not significant. Thus, there was no main effect for cohesion on
withdrawals from the commons. The main effect for prior Sharing was significant.
Those in the prior sharing condition were more cooperative than those in the no prior
sharing condition. Since the cohesion manipulation was not successful, the role of
cohesion on behavior in a commons dilemma is unclear.

Oskamp and Perlman (1966) explored the effects of friendship on cooperation in
the prisoner’s dilemma. They predicted higher levels of cooperation in Strong
friendships than in weak friendships. A questionnaire that was designed to identify an
individual’s best friend, acquaintances, and individuals who were disliked was
administered to introductory psychology classes at two colleges. Based on responses
to the questionnaire, two person groups were formed. Unfortunately, the
experimenters took "care...to insure that [subjects] saw each other clearly before
entering the experimental room" (p. 223). This procedure may have confounded the
type of relationship with identifiability. Identifiability has been found to increase
cooperative behavior in a variety of contexts, including the prisoner’s dilemma

(Bixtentine, Levitt, & Wilson, 1966; Jerdee & Rosen, 1974; Fox & Guyer, 1978).

24

Subjects were then separated and no communication was allowed. Subjects
participated in a multi-trial prisoner’s dilemma. Veridical feedback was presented.
The results indicated that there was no main effect for friendship on COOperation.
Additional analyses revealed that friends selected from Pomona College, but not from
Claremont Mens College, were more cooperative than acquaintances, disliked
individuals, and a fourth condition in which subjects did not know the individual.
Although the Pomona students were more cooperative, it is not clear whether these
effects were due to the strong social ties or to identifiability.

Swingle and Gillis (1968) explored the effects of emotional relationships on
behavior in the prisoner’s dilemma. Using a procedure that was Similar to Oskamp
and Perlman’s, Swingle and Gillis had male junior high school students indicate three
individuals whom they liked the most and whom they liked the least. Unfortunately,
like Oskamp and Perlman, these researchers may have confounded type of relationship
with identifiability. Although confidentiality of choices was stressed and subjeCtS were
not allowed to communicate, subjects were told that their partner was informed of
their identity. Subjects participated in a 100 trial prisoner’s dilemma in which false
feedback indicated the partner was making either 95% cooperative or 95% competitive
choices for the first 50 trials and then switched to the other strategy for the last 50
trials. The results indicated that the partners who liked each other were generally
more cooperative than those who disliked each other. In addition, individuals who
liked their partner were more likely to match the behavior of their partner than those

who did not like their partner. These researchers concluded that under certain

25

conditions, cooperation in friendly dyadS may not be Stable. The evidence indicating
the group cohesion and positive affective ties influence cOOperative behavior in the
social dilemma and the prisoner’s dilemma is not conclusive.

Hypgtheses

The present study explores how an individual’s beliefs that S/he deserves to
Obtain a shared resource, an individual’s perception that a fellow group member
deserves to obtain a shared resource, and the social ties existing among the group
members affect cooperative behavior in a social dilemma. There are three predictions
stemming from previous research that do not directly bear on the main focus of the
present study. It was predicted that there would be a main effect for beliefs of own
deservingness, beliefs of partner deservingness, and group cohesion. FirSt, individuals
who believe they deserve more of the resource Should take more of the resource than
those who are low in own beliefs of deserving. Second, individuals who believe their
partner deserves more of the resource Should take less of the resource than those who
believe their partner is low in deserving for the resource. Third, individuals in
cohesive groups should take less of the reSOurce than those in less cohesive groups.

It was also expected that the composition of the group with regard to an
individual’s beliefs of own deservingness and an individuals beliefs of his or her
partner’s deservingness would interact to influence cooperation. This expectation is in
a vein similar to Kelley and Stahelski’s (1970) findings (presented earlier) regarding
group composition and social motives. However, Specific predictions are not advanced

regarding the nature of this interaction. It is not clear whether individuals will

26

withdraw resources in a Similar manner when symmetry exists between group
members-both group members have either high or low in beliefs of deserving.
However, when asymmetry exists, it is expected that individuals with high beliefs of
deservingness who believe their partner is low in deserving should take more of the
resource than individuals with low beliefs of deservingness who believe their partner is
high in deserving.

Finally, it was predicted that an individual’s beliefs Of own deserving, an
individual’s perception of his or her partner’s deserving for the resource, and the
social ties among group members would affect jointly interested behavior. The
following predictions are based on the equity and equality principles. In cohesive
groups, individuals should be less concerned about achieving equitable outcomes and
Should take similar amounts of the resource, regardless of their own or their partners
deserving. In contrast, in low cohesive groups individuals should act to achieve
equitable outcomes for all group members. Individuals who believe they deserve to
obtain the resource should take more of the resource when their partner is low in
deserving than when their partner is high in deserving. In contrast, individuals who
believe they are low in deserving Should take less of the resource when their partner is

high in deserving than low in deserving.

CHAPTER H

METHOD

Desig

The design was a four-way (2 x 2 x 2 x 8) mixed analysis of variance (ANOVA)
with repeated measures on the last factor. The variables were group cohesion (high,
low), beliefs of own deserving (high, low), beliefs of other’s deserving (high, low),
and trials. The dependent variable was the number of acres of lumber harvested by
each subject.

Megs

Subjects were 151 (48 male, 96 female, 7 unknown) psychology Students at
Michigan State University. Of these, 91 (31 male, 58 female, 2 unknown) were from
introductory psychology courses, and 60 (17 male, 38 female, 5 unknown) were from
an introductory statistics course that was taught by the author. Slightly different
procedures were used to recruit subjects from the two courses. Students in both
courses were told that the session would last for 1 1/2 hours. Students in the
introductory psychology course were told they would receive two 1/2 hour credits plus
have an opportunity to earn money for the additional 30 minutes. In contrast, students
in the introductory statistics course were told they would receive three credits for their
participation and that they would have an opportunity to earn money. However, the
statistics Students were not told that each credit represented 30 minutes. This is

relevant because the instructions in the Study were designed for the two credit plus

27

28

money scenario. Thus, subjects recruited from the introductory Statistics course for
three credits who were expecting three credits heard the pre-recorded instructions
indicate they would receive two credits. Of course, all introductory statistics Students
received three credits for their participation.

Exmrimental M

The experimental task was a multi-trial two-person take-some social dilemma
(Orbell & Dawes, 1981). A take-some game is an experimental game used to Study
the conflict between individual and collective interests. In a take-some game, group
members acting in their self-interest receive larger outcomes than those acting in the
collective interest, but all group members receive larger outcomes if all act in the
collective interest. In addition, all group members are penalized (e.g., fined) if one or
more individuals acts in his or her self-interest.

The take-some game was simulated with a paradigm in which group members
Share access to a 200 acre forest (cf. Sattler & Kerr, in press). On each of a series of
trials, group members decide how many acres of lumber to harvest (i.e., withdraw) for
private use (i.e., the acres an individual group member harvests are not Shared with the
other group member). Each member could withdrawal between 20 and 100 acres.
The goal of each subject was to make withdrawals in such a way to acquire the largest
personal outcome possible.

The conflict between individual and collective interest was achieved through a
fine system that penalized overzealous harvesring. The fine was based on the tOIal

number of acres harvested by both group members. The function of the fine was

29

linear and the fine increased as the total number of acres harvested increased. For
example, no fine would be imposed if both members took 20 acres, whereas a fine of
200 acres would be imposed if the total taken by the group was 200 (the maximum
withdrawal allowed). Group members equally shared responsibility for paying the
fine. For example, if the fine was 60 acres, then each member would be fined half
this amount.

Group member interdependence was stressed to illustrate that each group
member’s choices affected both his or her own outcome as well as the Other group
member’s outcome. Trials were independent such that a withdrawal on one trial did
not affect the size of the forest on subsequent trials (i.e., the number of acres in the
forest was constant across trials). Trial independence minimizes concern with Short-
term and long-term consequences of withdrawal decisions (i.e., a temporal trap (see
Messick & McClelland, 1983)). There were eight no feedback trials. Subjecrs did not
receive information regarding the partner’s harvests, his or her own fine, the partner’s
fine, or the group outcome. Subjects were told that there would be between five to 10
trials.

Procedure

Subjects were randomly assigned to condition. Subjects were never referred to
by name and personal identities were not revealed. Rather, subjects were referred to
only by color code. These procedures attempted to increase feelings of anonymity, for
identifiability has been found to influence an individual’s decisions in group

interdependence situations (see Bixtentine et al., 1966; Jerdee & Rosen, 1974; Fox &

30

Guyer, 1978). Each subject was asked to give his or her written consent to participate
in the study. A copy of the consent form is presented in Appendix A.

The manipulation of beliefs of own deserving occurred during the first part of the
session. Four subjects were run in each session, and for each session two of the
subjects were scheduled to arrive 30 minutes earlier than the other two subjects. On
arrival at' the lab, subjects were placed in individual rooms and told that the study
would be conducted in several parts. Subjects were reminded that the length of the
experiment was 1 1/2 hours but that they only received 1 hour of credit. Subjects
were also told that they would have the opportunity to earn money in one of the tasks.
The money they earned in the task would be the way in which they would be
compensated for the 1/2 hour for which they were not going to receive class credit.
The first two subjects to arrive were then told that the first part of the session involved
completing a battery of questionnaires. The questionnaires contained open-ended
questions dealing with issues facing Michigan State University Students (cf. Lane &
Messé, 1971)). Subjects had 30 minutes to work on the questionnaires. The
questionnaire is presented in Appendix A.

At the end of the 30 minute period, the questionnaires were collected and the
second pair of subjects scheduled to arrive late were escorted into individual rooms in
the lab. With all four subjects now present, the subjects were told that four people
were participating in the session and that the two people who just arrived would
participate in the session for only 1 hour. The instrucrions stressed that two people

already had been working for 1/2 hour on a task and did not receive credit for

31

participating so far, but would be compensated for the 1/2 hour by earning money in
one of the tasks. However, all subjects were told that everyone had the opportunity to
earn money in one of the tasks.

Group cohesion was then manipulated. Subjects were told that the next part of
the session involved working on a winter survival problem. While the subjects were
in their individual rooms, tape recorded instructions described a winter survival
problem that required participants to imagine that everyone in the session was in a
light plane that crash-landed in the woods of northern Minnesota in mid-January
(Johnson & Johnson, 1975). The plane crash scenario indicated that 12 items were
salvaged from the plane (e.g., gun, extra clothing, newspaper, compass, candy bars).
Each of the 12 items were shown on a poster in each subject’s room. Subjects were
then told that they should think about how they would rank order the items in terms of
their importance for survival for two minutes, and to indicate their ranking on a form
titled "Initial Ranking Form." The Initial Ranking Form is presented in Appendix A.
After the Initial Ranking Form was completed, all subjects were asked to assemble
their chairs in a circle. Subjects in the low cohesion condition once again were asked
to think about the best way to rank the 12 items in terms of importance for survival
for the next five minutes and were instructed not to talk with any of the other
participants. In contrast, those in the high cohesion condition were instructed to
discuss with the Other group members the best way to rank the items in order of
importance for survival for the next five minutes. After the five minute period,

subjects returned to their rooms. All subjects then individually re-ranked the items

32
and indicated their ranking on a form titled "Final Ranking Form." The Final Ranking

Form is presented in Appendix A. Subjects were told that their final ranking solution
would be compared to the solutions made by experts and participants who worked on
the winter survival problem a previous term. In addition, subjects in the high cohesion
condition were told that their making solution would be averaged with the solutions of
the other group members to form a group solution. After a two minute period
allegedly to score the solutions, subjects in the high cohesion condition were told that
their group solution was better than 82% of the groups run the prior term, and those in
the low cohesion condition were told that their own individual solution was better than
82% Of those from the prior term.

A questionnaire titled "Winter Survival Questionnaire" that was designed to
assess the effectiveness of the group cohesion manipulation was then administered.
The instructions stressed that all responses on the questionnaire would be anonymous
and would not be shared with anyone. Items were answered on 7-point scales. The
questions included evaluation ratings of the other participants and how much they
liked being part of the group (cf. Gaertner, Mann, Murrell, & Dovidio, 1989). The
Winter Survival Questionnaire is presented in Appendix A.

The tape recorded instructions then indicated that the next part of session was
ready to begin. For this part, two person groups would be formed. Subjects were told
that for efficiency, there would be two separate two-person groups and that the two
groups would not be competing against one another. Subjects were randomly assigned

to a group, and subjects would not know which room their fellow group member

33
occupied.

Subjects then were made aware of their partner’s deserving. Subjecrs were given
a Sheet of paper indicating whether or not their partner worked on the questionnaire
task at the beginning of the session. This sheet is presented in Appendix A. It was
stressed that those who were present for the first 1/2 hour of the session did not
receive credit for that time. Further, although everyone would have the opportunity to
earn money in one of the tasks, those individuals who worked on the questionnaire
task during the first 1/2 hour would be compensated for their time through money they
may earn during the next task.

Subjects were then told that the instructions to the next task would be presented.
The harvesting task was then described. To familiarize subjects with the task, four
screens on the video monitors in each room presented sample harvests and outcomes.
Each example screen presented example harvests, example fines, and example
outcomes for bOth group members. The examples summarized the outcomes that
could result when group members made self-interested and collectively-interested
choices. For example, one example screen indicated that if both group members made
large harvests, then the fine would be severe and both individuals would receive very
small outcomes. Another screen indicated that if bOth members made small harvests,
then the fine would be minimal and both would receive moderate outcomes. The third
and fourth example screens indicated that if one group member made a large harvest
and the other member made a small harvest, then the fine would be moderate and the

individual making the large harvest would receive a moderate outcome but the

34

individual making the small harvest would receive very small outcomes.

Subjects were told that could earn money in the task in addition to two 1/2 hour
experimental credits they would receive for their participation in the session. The
amount of money each subject could earn was determined by the total number of acres
each individual harvested from the forest and the fine imposed on the group’s harvest.
Subjects would receive one penny for each acre of lumber s/he accumulated. The
instructions again stressed that although everyone had the opportunity to earn money
in the harvesting task, the money earned would serve as compensation for those who
completed the questionnaires at the beginning of the session.

The eight harvest trials then began. No feedback was presented concerning the
final number of acres either subject received or the amount of the fine imposed. At
end of the eight trial sequence, subjects completed a questionnaire. The questionnaire
included items that checked (a) the manipulation of the own beliefs of deserving and
beliefs of partner deserving factors, (b) subject suspiciousness about any Of the
experimental procedures, and (c) asked subjects to provide impressions of their partner
on a variety of evaluation items. The questionnaire is presented in Appendix A. After
completing the questionnaire, subjects were debriefed, received any money they
earned, thanked for participating, and excused. The experimental instructions are

presented in Appendix B.

CHAPTER III

RESULTS

Manipulation CM

@112 Cohesion. The responses on the six-item questionnaire that was designed
to measure the effectiveness of the group cohesion manipulation were evaluated by a
principle components factor analysis with varimax rotation. The results Showed that
one principle factor was present. Table 1 Shows the factor loading for each of the six
items. The loadings for the first four items ranged between .84 and .90. Item Six also
had a high loading of .63, and item five had the lowest loading of -.36.

The five items with the highest factor loadings were then used to create a factor
to evaluate the effectiveness of the cohesion manipulation. Each item was weighted
by multiplying the subject rating by the item’s factor loading. The fifth item on the
questionnaire was not included because it had a loading less than .40. The cohesion
manipulation was evaluated in relationship to the subjects’ class (introductory
psychology, statistics) and sex (male, female). Ideally, these two variables Should be
included in one analysis (ANOVA design). However, because of the small cell Size
for some of the variables, separate ANOVAS were performed. The first ANOVA
examined the subjects’ class. The design was a 2 x 2 x 2 x 2 between-subjects
ANOVA. The independent variables were group cohesion (high, low), beliefs of own
deserving (high, low), beliefs of partner deserving (high, low), and class (introductory

psychology, statistics). The dependent variable was the five item factor. There was

35

36
only a significant main effect for cohesion, 2(1, 135) = 22.57, p < .0001. The second

ANOVA examined the subjectS’ sex. The design was a 2 x 2 x 2 x 2
between-subjects ANOVA. The independent variables were group cohesion (high,
low), beliefs of own deserving (high, low), beliefs of partner deserving (high,
Table 1
Factor Loadings on Questionnaire Designed

to Measure Group Cohesion

 

Item Loading

 

"To what extent would you like to meet with these

same people again?" .84
"How much did you enjoy being in a group with

these people?" .88
"How much do you like the people in the group?" .90
"How much would you like to participate in another

experiment with the same people in the group?" .86
"How much would you like to be in a different

group in today’s session?" -.36
"How much do you feel as though yOu are a member

of a group?" .63

 

low) and sex (male, female). The dependent variable was the five item factor. There

37
was only a significant main effect for cohesion, E0, 135) = 31.87, p < .0001.

Another between-subjects ANOVA was used to evaluate the cohesion
manipulation without regard to subjects’ class or sex. The design was a 2 x 2 x 2
ANOVA. The independent variables were group cohesion (high, low), beliefs of own
deserving (high, low), and beliefs of partner deserving (high, low). There was only a
Significant main effect for group cohesion, 50,149) = 37.71, p < .0001. Subjects in
the high cohesion condition M = 4.02) were more favorable toward their group than
those in the low cohesion condition M = 3.20).

Ratings on the five individual questionnaire items were also evaluated in
relationship to cohesion, beliefs of own deserving, and beliefs of other’s deserving.
Five separate 2 x 2 x 2 between-subjects ANOVAS were performed. For each of the
five ANOVAS there was only a significant main effect for cohesion, such that subjects
in the high cohesion condition gave more favorable ratings than subjects in the low
cohesion condition. The ANOVA means and results for each item are summarized in
Appendix C. The ANOVA results indicate that the group cohesion manipulation was
successful.

m o_f _Ow_n Deserving. The belief of own deserving manipulation was
evaluated with two questions that were presented on the post-experimental
questionnaire. The first question was, "How much time have ypp been in the lab
today?" Since own deserving was manipulated in part by varying the amount of time
the subject was in the session, it was important to assess whether subjects were

cognizant of the amount of time they spent in the lab. Unfortunately, this question

38

was added to the questionnaire only after two-thirds of the subjects had participated in
the study. Forty-nine subjects responded to this question. Ideally, this question Should
be evaluated in relationship to subjects’ class. However, because of the small cell size
for some of the variables, such an analysis could not be performed. The question was
evaluated in relationship to cohesion, belief of own deserving, and belief of other’s
deserving. The design was a 2 x 2 x 2 between-subjects ANOVA. There was a
significant main effect for own belief of deserving, E(1, 41) = 183.70, p < .0001.
Subjects in the high own deserving condition who participated in the session for 90
minutes indicated they were in the lab longer (M = 88 min) than did those
participating for 60 minutes in the low own deserving condition (M = 54 min).

The second question was, "How much money do you think would be fair for ypp
to receive for participating in the study?" This item was evaluated in relationship to
the subjects’ class. The design was a 2 x 2 x 2 x 2 between-subjects ANOVA. The
independent variables were group cohesion (high, low), belief of own deserving (high,
low), beliefs of partner deserving (high, low), and class (introductory psychology,
statistics). The class main effect and the belief of own deserving x class interaction
were not significant. Another ANOVA was used to evaluate this question without
regard to class. The design was a 2 x 2 x 2 between-subjects ANOVA. The
independent variables were group cohesion (high, low), beliefs of own deserving (high,
low), and beliefs of partner deserving (high, low). There was a significant main effect
for belief of own deserving, 5(1, 143) = 10.76, p < .001. Subjects in the high own

deserving condition indicated a higher amount of money would be fair compensation

39
(M = $4.46) than those in the low own deserving conditions (M = $3.03). The

ANOVA results indicate that the belief of own deserving manipulation was successful.

B_eli_e_f_s o_f m Deserving. The manipulation for beliefs of other’s deserving
was evaluated with two questions that were presented on the post-experimental
questionnaire. Each question was evaluated in relationship to cohesion, beliefs of own
deserving, and beliefs of other’s deserving. The first question was, "How much time
has y_gt_r_r m been in the lab today?" Since other’s deserving was manipulated in
part by varying the amount of time the subjects were told their partner was in the
session, it was important to assess whether subjects were cognizant of the amount of
time their partner spent in the lab. Unfortunately, this question was added to the
questionnaire only after two-thirds of the subjects had participated in the study.
Forty-nine subjects responded to this question. Ideally, this question should be
evaluated in relationship to subjects’ class. However, because of the small cell Size
for some of the variables, such an analysis could not be performed. The question was
evaluated in relationship to cohesion, belief of own deserving, and belief of Other’s
deserving. There was a significant main effect for belief of other’s deserving, E(l, 41)
= 141.37, p < .0001. Subjects in the high partner deserving condition indicated that
their partner had been in the lab longer (M = 87 min) than those in the low partner
deserving condition (M = 54 min).

The second question was, "How much money do you think would be fair for
ygu_r ppm to receive for participating in the study?" This item was evaluated in

relationship to the subjects’ class. The design was a 2 x 2 x 2 x 2 between-subjects

40
ANOVA. The independent variables were group cohesion (high, low), belief of own

deserving (high, low), beliefs of partner deserving (high, low), and class (introductory
psychology, statistics). The belief of Other’s deserving x class interaction was
significant, E0, 135) --= 7.41, p < .01. The means underlying this interaction are
presented in Table 2. Post hoc contrasts (Winer, 1971) showed that subjects from the
introductory psychology class believed high deserving partners deserved more money
than low deserving partners, E0, 135) = 7.79, p < .01, but there was no significant
difference for partner deserving, 5(1, 135) = 2.22, p > .10.
Table 2
ANOVA Means (M) and Standard Deviations (SD) for Partner Deserving

x Class Interaction on Partner Deserving Manipulation Check

 

 

Class
Partner Introductory Introductory
Deserving Psychology Statistics
N M SD N M SD
High 45 472.53¢ 264.87 18 294.44¢ 175.64
Low 46 312.22¢ 259.43 42 409.02¢ 318.15

 

This interaction suggests that the belief of partner deserving manipulation was not
as effective for students from the introductory statistics class as for Students from the
introductory psychology class. This finding may be due to the slightly different

procedures that were used to recruit subjects from the two courses. Although Students

 

41

in both courses knew that the session would last one and one half hours, students in
the introductory psychology course were told they would receive two half hour credits
plus the opportunity to earn money for the additional 30 minutes. In contrast, Students
in the introductory Statistics course were told they would receive three credits for their
participation in the session and have the opportunity to earn money, but were not told
that each credit represented 30 minutes. Since the pro-recorded instructions presented
during the session were designed for the introductory psychology Students and
indicated that all subjects would receive two credits for their participation, it is
possible that for the introductory statistics students expecting three credits, these
instructions may have in some way altered their perception of partner deserving. For
example, the introductory statistics Students may have been more concerned about
receiving proper credit and this distraction may have caused these subjects to focus
more on their own situation. Thus, there is not conclusive evidence that the
manipulation of other’s deserving was effective.

A preliminary analysis was conducted to determine whether the class from which
Students where drawn interacted with group cohesion, belief of own deserving, and
belief of other’s deserving on the harvesting trials. A 2 x 2 x 2 x 2 x 8 mixed
ANOVA with repeated measures on the last factor were performed on the eight
harvest trials. There were no Significant main or interaction effects involving class in
either the between-subjects or repeated measures portions of the analysis. Since there

was no evidence that the class from which subjects were drawn qualified any of the

42

effects on the harvesting trials, class was not included in subsequent analyses
involving the harvesting dependent variable.

The next analysis was conducted to determine whether subjects’ sex interacted
with group cohesion, belief of own deserving, and belief of other’s deserving on the
harvesting trials. A 2 x 2 x 2 x 2 x 8 mixed ANOVA with repeated measures on the
last factor were performed on the eight harvest trials. There were no significant main
or interaction effects involving sex in either the between-subjects or repeated measures
portions of the analysis. Since there was no evidence that sex qualified any of the
effects on the harvesting trials, sex was not included in subsequent analyses involving
the harvesting dependent variable.

Harvesting m

The hypotheses concerning subjects’ harvesting behavior were evaluated in a 2 x
2 x 2 x 8 mixed ANOVA with repeated measures on the last factor.2 The independent
variables were group cohesion (high, low), beliefs of own deserving (high, low),
beliefs of partner deserving (high, low), and trials. Based on the equity principle, it
was expected that there would be a main effect for beliefs of own deserving and for
beliefs of other’s deserving. However, the main effect for beliefs of own deserving
was not significant, E(1, 143) = 1.06, p = ns. Individuals who believed they deserved
to obtain the resource did not take more than those low in beliefs of own deserving.
In addition, the main effect for beliefs of other’s deserving was not significant, E(l,
143) = .40, p = nS. Individuals who believed their partner deserved more of the

resource did not take less of the resource than whose who believed their partner was

43

low in deserving.

It was also expected that there would be a main effect for group cohesion.
However, the main effect for cohesion was not significant, E(1, 143) = .95, p = ns.
Subjects in cohesive groups did not take less of the resource than those in low
cohesive groups. An additional analysis was performed to further explore the
non-Significant group cohesion main effect. A one-way ANOVA was performed on
the group cohesion variable for subjects in the low beliefs of own deserving and low
beliefs of partner’s deserving conditions, where the direct effect of cohesion,
uncomplicated by relative deserving considerations could be examined. The group
cohesion main effect was not Significant. Table 4 presents the summary table for the
between-subjects portion of the ANOVA.

The next hypotheses focused on how group cohesion might moderate equity
concerns. It was hypothesized that individuals in low cohesive groups would be more
concerned about equitable outcomes than those in high cohesive groups. Thus,
individuals in low cohesive group should be more influenced by their own deserving
than those in high cohesive groups. This hypothesis was supported. The Cohesion x
Own Deserving interaction was significant, _E(l, 143) = 5.63, p < .05. The means
underlying this interaction are Shown in Table 3. Simple effects tests (Winer, 1971)
indicated that in the low cohesion groups, high deserving subjects made larger harvests
than low deserving subjects, _E(1, 143) = 5.59, p < .05, but in the high cohesion groups
subjects made similar harvests regardless of their own deserving, 5(1, 143) = 0.97, p =

ns. Individuals in low cohesive groups appeared to be more concerned with equitable

44

outcomes than those in the cohesive groups.
Table 3

ANOVA Means and Standard Deviations for Cohesion

x Own Deserving Interaction on Harvest Trials

 

Own Deserving

 

Cohesion High Low

N M. SD N M SD
High 40 52.98 16.61 36 56.74 16.42
Low 43 61.63 17.31 32 52.48 14.89

 

The corresponding hypothesis predicted that individuals in low cohesive group
Should be more influenced by their partner’s deserving than those in high cohesive
groups. However, the Cohesion x Partner Deserving interaction was not Significant.
AS discussed earlier, the partner deserving manipulation check suggested that the
manipulation may have been more effective for Students selected from the introductory
psychology courses than from the Statistics course. Thus, an analysis was performed
only with students from the introductory psychology course to determine whether this
interaction was Significant for this group. A 2 x 2 x 2 x 8 mixed ANOVA with
repeated measures on the last factor was performed. The independent variables were
group cohesion (high, low), beliefs of own deserving (high, low), beliefs of Other’s

deserving (high, low), and trials. The Cohesion x Partner Deserving interaction was

45
Table 4

ANOVA: Cohesion x Own Deserving x Partner

Deserving on Harvest Trials: Between-Subject Factors

Source of Mean Sig.
Variation df Square F of F
Cohesion (C) 1 2089.14 .95 .331
Own Deserving (OD) 1 . 2327.231 .06 .305
Partner Deserving (PD) 1 874.05 .40 .529
C x OD 1 12396.7 5.63 .019
C x PD 1 1230.88 .56 .456
OD x PD 1 328.14 .15 .700
C x OD x PD 1 49.09 .02 .881
Error 143 2199.97
Table 5
ANOVA: Cohesion x Own Deserving x Partner
Deserving x Trials: Effects of Repeated‘ Measures
Source of Mean Sig.
Variation df Square F of F
Trials (T) 7 941.12 3.29 .002
Cohesion (C) x T 7 203.19 .71 .663
Own Deserving (OD) x T 7 105.11 .37 .921
Partner Deserving (PD) x T 7 381.89 1.34 .230
C x OD x T 7 478.46 1.67 .112
C x PD x T 7 544.09 1.90 .066
OD x PD x T 7 585.20 2.05 .047
C x OD x PD x T 7 182.84 .64 .724
Within Cells 1001 286.06

not significant. Together, these analyses suggest that the manipulation for partner
deserving may have been too weak (as suggested by the manipulation check) and/or

individuals may be more concerned with their own deserving than other’s deserving.

46

Although the Cohesion x Partner Deserving interacrion was not significant, there
were two unexpected 3-way interactions involving the partner deserving and trials
factors in the analysis performed on all subjects. These interactions indicate that
under some conditions individuals may be sensitive to Other’s deserving. The analysis
revealed a marginally significant Cohesion x Partner Deserving x Trials interaction,
F(7, 1001) = 1.90, p = .066, and a Significant Own Deserving x Partner Deserving x
Trials interaction, F(7, 1001) = 2.05, p < .05. Table 5 presents the summary table for
the repeated measures portion of the ANOVA.

The next analyses were performed to explore the Cohesion x Partner Deserving x
Trials interaction and the Own Deserving x Partner Deserving x Trials interaCtion that
indicated under certain conditions individuals may be sensitive to others’ deserving
when making their harvesting decisions. It was first suspected that both interactions
might be manifest in an early versus late trial pattern. Such a pattern might Show
subjects initially making smaller or larger withdrawals on later than earlier trials (e.g.,
making larger withdrawals because the trial sequence and the opportunity to harvest
would be over soon vs. making smaller withdrawals because the desired harvest has
been achieved). An analysis was performed on a block consisting of the first four
trials and a block consisting of the last four trials. The design was a 2 x 2 x 2 x 2
mixed ANOVA with repeated measures on the last factor. The independent variables
were group cohesion (high, low), own deserving (high, low), partner deserving (high,
low), and trials. There were no significant interactions involving the trials factor. It

was concluded that the two 3-way interactions were not due to an early versus late

47

trial effect.

It was also possible that the interactions were localized in an early versus middle
versus late trial pattern. Such a pattern might Show subjects making different harvests
during the middle trials than the early or late trials. An analysis was performed on a
block consisting of the first three, second three, and final two trials. The design was a
2 x 2 x 2 x 3 mixed ANOVA with repeated measures on the last factor. The
independent variables were group cohesion (high, low), own deserving (high, low),
partner deserving (high, low), and trials. There were no Significant interactions
involving the trials factor. It was concluded that the two 3-way interactions were nor
due to an early versus middle versus late trial effect.

The Cohesion x Partner Deserving x Trials interaction and the Own Deserving x
Partner Deserving x Trials interaction were plotted across trials to assist in identifying
the nature of the interactions. Visual inspection of the means underlying the Cohesion
x Partner Deserving x Trials interaction presented in Figure 1 suggested a sawtooth
alternation pattern on the first six trials for high cohesion subjects. High cohesion
subjects with high deserving partners appeared to make smaller harvests on Odd
numbered trials than on even numbered trials. In contrast, high cohesion subjects with
low deserving partners appeared to make smaller harvests on even numbered trials
than on odd numbered trials. Low cohesion subjects did not appear to make

alternating harvests. Visual inspection of the means underlying the Own

48

:2: 288 Bean.
:3 .2080 Ocean.
02.”. em:

 

 

5.5500 5.850 :9... .5.
20.6922. 23.... x 9.3580 55.3..
. 8%.".

 

6% aces...

49

3.... .238 seen.
33 .2800 Lenten.
ozm em:

 

cowficoo 5.8560 26... .3
5.699.... 2%... x 8.2680 Lenten
8.258 L 2%.".

-00
-mo
AK

 

e~.w emote:

50

 

 

:9: 5.580 35.8.. I I
26.. .2030 Conrad II

52650 9.2860 :30 no... Lo.
ozmom: 22599:. 22: x 3.3030 Lenten.
N 8sz

-00
-mo

 

-05

8m v.65:

51

:9... .2860 .0530
25.. .2630 55.9...
0zm0m...

 

 

cozficoo 9.3.680 :30 26.. .2
5.6292. 22.... x 9.3.300 .ecten.
8.2.8. N ocsmfi

 

35 «meter

52

Deserving x Partner Deserving x Trials interaction presented in Figure 2 also
suggested a sawtooth alternation pattern, but here it is for high own deserving subjects
on the first five trials. High own deserving subjects with high deserving partners
appeared to make smaller harvests on odd numbered trials than on even numbered
trials. In contrast, high deserving subjects with low deserving partners appeared to
make smaller harvests on even numbered trials than on odd numbered trials. Low
deserving subjects did nOt appear to make alternating harvests.

The next analyses were performed to determine if the two 3-way interactions
were manifest in an odd versus even trial alternation pattern. The first analysis was
performed on a block consisting of all odd numbered trials (trials 1, 3, 5, 7) and a
block consisting of all even numbered trials (trials 2, 4, 6, 8). The design was a 2 x 2
x 2 x 2 mixed ANOVA. The independent variables were cohesion (high, low), own
deserving (high, low), partner deserving (high, low), and trials. The Own Deserving x
Partner Deserving x Trials interaction was significant, 5(1, 143) = 6.71, p < .01.
Simple interaction effects were calculated for high own deserving and low own
deserving conditions. There were no Significant effects for the low own deserving
condition. In the high own deserving condition, the Partner Deserving x Trials
interaction was Significant, F(l, 81) = 10.44, p < .01. The means underlying this
Simple interaction are presented in Table 6. Further Simple effects tests revealed that
when own deserving was high, subjects with partners high in deserving made smaller
harvests on odd trials than on even trials, 5(1, 81) = 5.11, p < .05., whereas subjects

with partners low in deserving made larger harvests on odd trials than on even trials,

53

13(1, 81) = 4.98, p < .05. The Cohesion x Partner Deserving x Trials interaction was
not significant.
Table 6
Means and Standard Deviations for Partner Deserving x Trials

Simple Interaction for High Own Deserving Condition for Eight Trials

 

 

Trial
Partner Deserving Odd Even
N M SD N M SD
High 34 56.41 18.59 34 61.04 20.48
Low 49 58.49 16.89 49 54.69 18.37

 

Although the Cohesion x Partner Deserving x Trials interaction was not
significant in the Odd versus even trial analysis for all eight trials, visual inspection of
the interaction in Figure 1 suggested that both the interacrion effects were localized in
the first six trials. The next analysis was performed to determine if the Cohesion x
Partner Deserving x Trials interaction was localized in the first Six trials. A 2 x 2 x 2
x 2 mixed ANOVA was performed on a block consisting of odd trials 1, 3, and 5 and
a block consisting of even trials 2, 4, and 6. The independent variables were cohesion
(high, low), own deserving (high, low), partner deserving (high, low), and trials. The
Cohesion x Partner Deserving x Trials interaction was Significant, 5(1, 143) = 5.44, p
< .05, and the Own Deserving x Partner Deserving x Trials interaction was significant,

2(1, 143) = 8.17, p < .01. Simple interaction effects were first calculated for the

54

Cohesion x Partner Deserving x Trials interaction for the high cohesion and low
cohesion conditions. There were no significant effects for the low cohesion
conditions. In the high cohesion conditions, the Partner Deserving x Trials simple
interaction was significant, F(l, 74) = 7.95, p < .01. The means underlying this
interaction are presented in Table 7. Further simple effects tests revealed that when
cohesion was high, subjects with high deserving partners made smaller harvests on odd
trials than on even trials, F(l, 74) = 4.66, p < .05, whereas subjects with low
deserving partners tended to make larger harvests on odd trials than on even trials,
F(l, 74) = 3.31, p < .10. This finding suggests that high cohesion subjects were
influenced by partner deserving on the first six trials.

Next, simple interaction effects were calculated for the Own Deserving x Partner
Deserving x Trials interaction for the high and low own deserving conditions. There
were no significant effects for the low own deserving conditions. In the high own
deserving conditions, there was a significant Partner Deserving x Trials simple
interaction, F(1, 81) = 11.46, p < .001 The means underlying this interaction are
presented in Table 8. Further simple effects tests revealed that when own deserving
was high, subjects with high deserving partners made smaller harvests on odd trials
than on even trials, 5(1, 81) = 8.24, p < .01, whereas subjects with low deserving
partners tended to make larger harvests on odd trials than on even trials, 13(1, 81) =
3.33, p < .10. This finding suggests that high own deserving subjects were influenced

by partner deserving.

55
Table 7

Means and Standard Deviations for Partner Deserving x Trials

Simple Interaction for High Cohesion Condition for First Six Trials

 

 

Trial
Partner Deserving Odd Even
N M SD N M SD
High 36 50.20 17.00 36 55.74 20.20
Low 40 56.14 18.46 40 51.71 17.11

 

Table 8

Means and Standard Deviations for Partner Deserving x Trials

Simple Interaction for High Own Deserving Condition for First Six Trials

 

 

Trial
Partner Deserving Odd Even
N M SD N M SD
High 34 54.49 18.30 34 61.45 20.75
Low 49 58.22 17.82 49 54.46 19.90

 

Another analysis was performed to determine whether the Own Deserving x

Partner Deserving x Trial interaction was localized to the first six trials and non-

significant on trials 7 and 8. A 2 x 2 x 2 x 2 mixed ANOVA was performed on a

56

trials 7 and 8. The independent variables were cohesion (high, low), own deserving
(high, low), partner deserving (high, low), and trials. There were no Significant
interactions involving trials. It was concluded that the interaction was localized to the
first Six trials.

Even though the group means suggested an alternation pattern, it is possible that
individual subjects did not engage in the odd versus even trial alternation patterns
found in the Cohesion x Partner Deserving x Trials interaction and the Own Deserving
x Partner Deserving x Trials interaction. Instead, the mean alternations may have been
a congruence of some other pattem(s) among individual subjects. Several analyses
explored whether individual subjects alternated their harvests as implied by the group
mean data. These analyses were performed on the first Six trials.

The alternation patterns present in bOth of the 3-way interactions involving the
trials factor appeared to be mirror images of each other. In one pattern, the sawtooth
alternation appeared to begin with smaller harvests on the first trial than on the second
trial. This pattern will be referred to as alternation pattern A. In the other pattern, the
sawtooth alternation appeared to begin with larger harvests on the first trial than on
the second trial. This pattern will be referred to as alternation pattern B. The firsr set
of analyses were performed to determine if individual subjects engaged in the making
alternating harvests. These analyses calculated the number of times an individual
made harvests that were compatible with each leg of alternation pattern A and B. For
example, a harvest would have been considered compatible with the first leg of

alternation pattern A if the subject (a) made a smaller harvest on the firm trial than on

 

57

the second trial, and (b) the change was large enough to assume that they were not
simply small chance fluctuations (i.e., the harvest change between trials was equal to
or greater than 5 acres of lumber). The first analysis was performed on pattern A.
The design was a 2 x 2 x 2 ANOVA. The independent variables were cohesion (high,
low), own deserving (high, low), and partner deserving (high, low). The dependent
variable was the number of alternations on the first six trials that were consistent with
pattern A. The Cohesion x Partner Deserving interaction was significant, _E(1, 143) =
3.78, p < .05. Simple effects tests indicated that high cohesion SUbjCCIS made more
harvests that were consistent with alternation pattern A when their partner was high in
deserving than low in deserving, F(l, 74) = 5.17, p < .05. In addition, the Own
Deserving x Partner Deserving interaction was significant, F(1, 143) = 11.82, p < .
001. Simple effects tests indicated that high own deserving subjeCtS made more
harvests consistent with alternation pattern A when their partner was high in deserving
than low in deserving, 5(1, 81) = 6.43, p < .05.

The second analysis was performed on pattern B. The design was a 2 x 2 x 2
ANOVA. The independent variables were cohesion (high, low), own deserving (high,
low), and partner deserving (high, low). The dependent variable was the number of
alternations on the first Six trials that were consistent with pattern B. The Cohesion x
Partner Deserving interaction was significant, 5(1, 143) = 4.13, p < .05. Simple
effects tests indicated that high cohesion subjects made more harvests that were
consistent with alternation pattern B when their partner was low in deserving than high

in deserving, _Ij(1, 74) = 6.32, p < .05. In addition, the Own Deserving x Partner

58

Deserving interaction was marginally significant, 5(1, 143) = 3.29, p < . 10. Simple
effects tests indicated that high deserving subjects made more harveSts consistent with
alternation pattern B when their partner was low in deserving than high in deserving,
F(1, 81) = 5.97, p < .05. These analyses suggest that subjects were engaging in the
alternation patterns.

Although the two previous analyses suggest that individuals were engaging in the
alternation patterns, they do not make any allowance for the possibility that subjects
may have exhibited a linear trend in harvests over trials. Such a linear trend
superimposed on simple alternation might mask the latter, at least in the previous
method of analysis. To explore this possibility, for each subject a regression line
between trial numbers 1-6 and harveSt was computed and the residual between the
actual and predicted harvest was obtained. A harvest was considered compatible with
the alternation pattern if (a) the residual was equal to or larger than the criterion of
eight acres of lumber, and (b) was in the direction that was compatible with each leg
of the alternation pattern. The Cohesion x Partner Deserving interaction was
marginally significant, 13(1, 143) = 3.65, p = .058. Simple effects tests indicated that
high cohesion subjcCtS with high deserving partners tended to engage in the alternation
pattern, 5(1, 74) = 2.82, p < .10. In addition, the Own Deserving x Partner Deserving
interaction was significant, 13(1, 143) = 6.99, p = .01. Simple effects tests indicated
that high own deserving subjects made engaged in the alternation pattern, 5(1, 81) =

4.60, p < .05.

59

Impressions o_f Ml

Subjects provided their impressions of their partner after the trial sequence on
seven items. The responses to these items were evaluated by a principle components
factor analysis with varimax rotation. The results showed that one principle factor was
present. Table 9 shows the factor loading for each of the seven items. The loadings
ranged between .74 and .81. All seven item were then used to evaluate partner
impressions. Each item in the factor was weighted by multiplying the item score by
its factor loading.

Partner impressions were evaluated in relationship to subjects’ class. The design
was a 2 x 2 x 2 x 2 between-subjects ANOVA. The dependent variable was the seven
item factor score. There were no Significant main or interaction effects involving
class. Another between-subjects ANOVA was used to evaluate partner impressions
without regard to subjects’ class. A 2 x 2 x 2 ANOVA was conducted on the seven
item factor. There was a Significant Cohesion x Partner Deserving interaction, 5(1,
143) = 3.98, p < .05. Simple effects analyses showed that when cohesion was high,
subjects rated high deserving partners more favorably than low deserving partners, 5(1,
143) = 4.30, p < .05, but when cohesion was low, subjects did not differentiate
between high deserving and low deserving partners, 5(1, 143) = .33, p = ns.
Mortal. Finding

Subjects provided their impressions of the experiment after the trial sequence.
For the question asking how much an individual wanted to avoid the fine, there was a

marginally Significant Cohesion x Own Deserving interaction, 5(1, 143) = 3.65, p =

 

Items

gene}
COHCl
respc
hone:
Erica
trusti

C00p<

those
10w (
infilt-
Desc
Wit:
31 CO
6.54,

561181

60

.06. Simple effects tests indicated that when an individual was high in deserving,
Table 9

Factor Loadings for Partner Impression

 

 

Items Loading
generous/selfish .77
concerned for others/concemed for self .77
responsible/very irresponsible .73
honest/dishonest .8 l
friendly/unfriendly .76
trusting/not trusting .81
cooperative/competitive .7 4

 

those in the high cohesion groups tended to want to avoid the fine more than those in
low cohesive groups, _E(1, 81) = 3.48, p < .10. For the question asking how much an
individual’s partner wanted to win money, there was a significant Cohesion x Partner
Deserving interaction, F(1, 143) = 5.81, p < .05. Simple effects teSIS indicated that
when partner deserving was high, compared to those in low cohesive groups, subjeCts
in cohesive groups believed their partner wanted to win the money more, 5(1, 63) =
6.54, p < .05. Finally, compared to low deserving subjects, subjects high in deserving

believed their partner found the harvesting task easier, 5(1, 143) = 5.77, p < .05 .

CHAPTER IV

DISCUSSION

The harvest data support the hypothesis that interpersonal relations moderate the
influence of an individual’s beliefs of own deserving and beliefs of other’s deserving
for a resource on cooperative behavior in a social dilemma. Different disuibution
rules or allocation norms appeared to be employed in the cohesive and non-cohesive
groups. Deutsch’s (1975) equity distribution rule and Clark and Mills’ (1979)
classification of exchange relationships prescribe that in Situations in which social ties
among group members are weak or non-existent, individual group members Should
receive outcomes in proportion to their respective "contributions" to the group. Note
that the term "contribution" in the present study refers to an individual’s belief of his
or her own deserving and partner’s deserving for the resource as determined by the
amount of time an individual worked at a task. The equity rule implies that
individuals should consider the contributions of all group members when making their
harveSt decisions, for doing so allows an individual to determine how to behave
equitably (i.e., what outcomes each group member Should receive). Subjects in low
cohesive groups appeared to behave equitably, but only with regard to their own
contributions-~they were not influenced by the contributions of other group members.
Low cohesion subjects made larger harvests when they believed they deserved to
obtain the resource than when they believed they were low in deserving. Low

cohesion subjects did not make smaller harvests over all trials when they believed

61

62

their partner was high in deserving than when they believed their partner was low in
deserving. Subjects in the low cohesive groups were aware of their partners’
deserving (especially subjects from the introductory psychology courses), but this
awareness did not influence their harvest decisions.

One explanation for this finding is that in a social dilemma, the equity rule may
be valid only for an individual group member’s decisions in relation to his or her own
contributions and outcomes, not for an individual group member’s decisions in relation
to the contributions and outcomes of all group members. There are two alternative
methodological explanations for this finding. First, the equity rule indicates that
individuals should receive equitable outcomes for their respective contributions to the
group. It is possible that the type or nature of the contribution is critical to an
individual’s concern with others’ outcomes. One definition of contribution for the
equity rule is the amount of effort an individual contributes toward achieving a
common group goal. For example, an equitable outcome for individuals working
together to achieve a goal, for pay, would be one where the individuals who
performed the most received the largest outcomes. The present Study manipulated
individual contributions by varying the length of time individuals worked on a task
that was unrelated to the social dilemma task. Although individuals who worked
longer at this task may have been more deserving of money than those who did not
work as long, it is possible that subjects did nOt consider it to be contribution.
Therefore, although low cohesion subjects were aware of others’ deserving for the

resource, they may not have believed it was relevant to the social dilemma task.

63

Second, the manipulation of deserving/contribution may have been too weak. This
possibility is suggested by lack of an overall effect for partner deserving. Subjects
recruited from the introductory psychology courses were more aware of their partner’s
deserving than were subjects recruited from the introductory Statistics course.

Clark and Mills’ (1979) classification of communal relationships and Deutsch’s
equality distribution rule prescribe that in situations in which social ties among group
members are strong and enjoyable social relations exist, all individual group members
should share equally in the group’s outcome regardless of their respective
contributions to the group. And indeed, subjects in high cohesive groups appeared to
be less concerned about equitable outcomes and appeared to consider overall their own
and their partner’s contribution when making their harvest decisions. However, the
Strategies high cohesion subjects used to make cooperative withdrawals were
sometimes somewhat complex. First, high cohesion subjects who believed they
deserved to obtain the resource harvested amounts Similar to those who believed they
were low in deserving. The strategy used was direct and simple-on every trial these
individuals consistently made moderately-sized harvests. Although the present design
did nor allow for a direct test of the equality distribution rule-~since subjects did not
provide their expectations regarding their partner’s harvesting behavior--these results
provide indications that equality may have been a factor in subjects’ harvesting
behavior.

Second, high cohesion subjects considered the contributions of their partner but

engaged in a fairly complex strategy. Subjects engaged in a sawtooth alternation

64

strategy in which they alternated between making smaller and larger harvests on each
of the first six trials. Subjects tended to make (non-significantly) smaller harvests on
the first trial when they believed their partner deserved to obtain the resource than
when they believed their partner was low in deserving. This difference in harvest size
on the first trial was related to nature of the alternation pattern. Subjects who believed
their partner deserved to obtain the resource made smaller harvests on odd numbered
trials than on even numbered trials. In contrast, subjects who believed their partner
was low in deserving made larger harvests on odd numbered trials than on even
numbered trials. Thus, individuals who withdrew a little more on the first trial
compensated for doing so on the following trial by withdrawing a little less, and so
on. Likewise, individuals who withdrew a little less on the first trial compensated for
doing so on the following trial by withdrawing a little more, and so on.

The complex alternation patterns suggest that the Strategy individuals in cohesive
groups may use when others’ deserving is salient is not straightforward. Although
high cohesion subjects were influenced by partner deserving, they did not consistently
make moderately-Sized harvests as in the Cohesion x Own Deserving interaction. One
explanation for this alternation pattern is that high cohesion subjects may have
experienced a conflict between not exploiting a partner who they liked versus the risk
of being exploited by their partner if they consistently made small harvests (i.e., the
sucker effect (see Kerr, 1983)). Alternating harvest size may have been a resolution to
this problem. An alternative explanation is that own deserving may be more salient

and have more direct and powerful influences than other’s deserving. A third

65

explanation is the same methodological issue raised earlier for contribution (i.e., the
contribution manipulation may have been too weak). The manipulation may have
been strong enough to influence behavior initially but not strong enough to sustain its
influence over time.

The present results, theory advanced by Deutsch (1975, 1982) and Schwinger
(1986), and the findings of Clark and Mills (1979) highlight the important role of
Situational and social influences on distribution norms (viz., that interpersonal relations
moderate the type of allocation norm that may be utilized) (see Greenberg & Cohen,
1982). In addition, Walster, Walster, and Berscheid (1978) have speculated that
individuals in intimate relationships, "through identification with and empathy for their
partners, come to define themselves as a gait; as 9pc couple. They see themselves nor
merely as individuals interacting with others, but also as part of a partnership,
interacting with other individuals, partnerships, and groups. This characteristic may
have a dramatic impact on intimates’ perceptions of what is and is not equitable" (p.
152-153). This speculation is consistent with theory and the present research by
suggesting that in relationships that are characterized by strong social ties, individuals
may seek to distribute resources based on equality or need (see Schwinger (1986) for a
discussion of resource distribution based on need), rather than distributing benefits or
resources in an equitable manner.

The present results and theory discussed above also provide a clearer picture of
the conditions in which the norm of equity is salient and may be acted upon in a

social dilemma. First, when the social relations of individuals sharing a resource are

66

characterized by weak social ties among group members "in which economic
productivity is a primary goal [e.g., low cohesive groups or an exchange relationship],
equity rather than equality or need will be the dominant principle of distributive justice
[i.e., the dominant allocation/justice norm]" (Deutsch, 1975, p. 143). Second, when
the social relations of individuals Sharing a resource are characterized by Strong social
ties and "fostering or maintenance of enjoyable social relations" (Deutsch, 1975, p.
146) is emphasized (e.g., cohesive groups or a communal relationship), concerns for
equity will be minimal. As Leventhal (1976) asserts, equality occurs when
"maintenance of harmony and solidarity among receivers is important" (p. 218) (cf.
Lerner, Miller, & Holmes, 1976).

These findings have important implications for collectively interested behavior in
social dilemmas and for solutions to social dilemmas. First, as suggested by Kerr (in
press), norms "can and do have important effects on behavior in social dilemmas."
Second, when the appropriate social conditions exist, certain norms may increase the
likelihood that individuals will seek to promote the collective interest. Although
equality per se may not insure collectively interested behavior, it may inhibit
individuals from competing with one another. Third, discussion among group
members may further generate norms that alter the collective welfare. In the present
study, discussion about topics unrelated to the social dilemma was a factor leading to
social ties among group members, which in turn appeared to promote group welfare
norms under certain conditions.’ Further, these social ties appear to endure and carry

over to new Situations (i.e., in the present research, they developed during the winter

 

67

survival situation and carried over to the social dilemma Situation).

It is interesting to consider the possible ramifications if group members do not
adhere to the dominant norms in a group facing a social dilemma. As Clark and Mills
(1979) demonstrated, counter-normative behavior may result in decreased attraction for
the individual and/or decreased acceptance of an individual by Other group members.
Bonner (1959, p. 69) suggested that deviants from group norms are not accepted by
the group. In a social dilemma, what might be the consequences of behavior that is
counter to the dominant justice norm? For example, if the norm of equality was
dominant in a cohesive group and one group member acted competitively, would the
other group members continue to adhere to or reject the norm. Possibilities include
the following: the group members might (a) dislike the competitive individual, allow
him or her to remain in the group, but continue to act in an equitable manner, (b)
rejeCt the individual from the group, or (c) reject the equality norm and act in a
competitive manner. This issue could be explored in future research with a design
similar to the present one, with the addition of a feedback variable. Feedback could
indicate whether a fellow group member was acting in line with the norm In essence,
this issue is Similar to one raised by Kerr (1983) is his Studies on the free-rider,
sucker, and social loafing effects in social dilemmas.

It should be noted that several theorists have argued that norms may not be an
effective solution for social dilemmas. Orbell and Dawes (1981) speculate that when
acting in one’s self-interest is particularly attractive in a social dilemma, internalized

norms such as conscience "might be a thin reed for a society to lean on" (p. 50).

Hard
in wf
"nev
belie
beha

will

co'n
ind]
affe

PFC:

 

68
Hardin (1977), the proponent of the "mutual coercion mutually agreed upon" solution

in which individuals are coerced into collectively interested behavior, recommends
"never ask a person to act against his [or her] own self-interest" (p. 2). Hardin
believes that in a social dilemma, voluntary cooperation is ineffective, self-interested
behavior is inevitable, and if a few individuals act in their self-interest, then others
will follow.

The findings also raise a very important issue regarding the influence of
interpersonal relationships on COOperative behavior in social dilemmas. In the present
study the main effect for group cohesion was not significant. Contrary to Brewer and
Kramer’s social identity findings and group cohesion research and theory in domains
other than social dilemmas, in the present Study subjects in cohesive groups were not
more cooperative than those in low cohesive groups. One explanation for this finding
may lie in the distinction between social identity and group cohesion. Social identity
involves cognitions associated with unit formation, categorization, and shared fate
(e.g., Campbell, 1958). In contrast, a substantial component of group cohesion is the
affect that individual group members feel for one anOther. Brewer and Kramer (1984,
1986) did not find an increase in positive feelings for Other group members in high
than in low social identity conditions, whereas research indicated that individuals in
cohesive groups feel more positive affect for their fellow group members than do
individuals in low cohesive groups. The distinction between the cognitive and
affective components of social identity and group cohesion, and the findings of the

present research and Brewer and Kramer’s (1984, 1986) research, suggests that it may

69

be the cognitions associated with unit formation and Shared fate, not positive affect,
that influences cooperation in social dilemmas.

Why might cognitions concerning shared fate and unit formation, rather than
affect for group members, increase cooperation? A clue to the answer is provided by
French and Raven’s (1959) concept of referent power and Kelman’s (1958) concept of
identification. Referent power involves an individual’s desire to compare and evaluate
him or herself to an individual with whom S/he identifies. An individual may gain a
sense of intrinsic personal satisfaction from identification with another individual.
Identification refers to a form of influence that stems from a relationship between two
individuals. In such a relationship, an individual may not agree completely with the
others’ beliefs, but may Still "adOpt the induced behavior because it is associated with
the desired relationship" (Kelman, 1958, p. 33). Although an individual may admire
or feel positive affect for another individual, it does not necessarily follow that the
individual’s actions will be influenced by the other individual. In contrast, an
individual who identifies with another individual and who wants a relationship (i.e., a
unit relationship) with the individual may see his or her behavior as closely tied to the
other individual (i.e., shared fate may be very salient). Wit (1989) has Shown how
factors that break up a unit relationship decrease cooperation in a social dilemma.
Subjects who believed that their fellow group members either had made their decisions
three weeks earlier or would make their decisions three weeks in the future were less
cooperative than subjects who believed they were making their decisions concurrently

with others. Thus, the variable driving COOperation may be the cognitions associated

70
with shared fate and the unit rather than affect (cf. Campbell, 1958). This suggests a

critical test between shared fate and positive affect in a social dilemma. We (Kerr,
Sattler, and Kaufman, 1990) are currently conducting a study to explore this issue.

The present results also indicate that scarcity may be created by increasing
demand for a resource when the supply remains constant. It is interesting to note that
strong social ties among group members did not moderate self-interested behavior
when beliefs of own and other’s deserving were high, for an interaction between
beliefs of own deserving, beliefs of other’s deserving, and group cohesion was not
observed. Might individuals in groups with even stronger social ties than those in the
present study exercise restraint under conditions of demand induced scarcity? What
factors might moderate the effects of demand induced scarcity on cooperative
behavior? These important questions remain unanswered and it is strongly suggested
that future research attempt to answer these questions.

It is also interesting to speculate about the role of Deutsch’s (1975) need
principle in a social dilemma. The present author found no research exploring how
need for a resource affects cooperative behavior in a social dilemma. Real world
situations exist in which individuals sharing scarce resource need to obtain the
resource. An example of such a situation is that of Shipwreck survivors who are
stranded on a lifeboat that contains limited rations. Some survivors may have a
greater physical need to obtain food and water than others. It is conceptually possible
that need may be distinguished from deserving. An individual may physically need to

obtain food but not believe s/he deserves to obtain it. Likewise, an individual may

bell:

dist

C011

Clll'.

Sut

1111'

dis

C01

71

believe S/he deserves to obtain food but not physically need to obtain it. This
distinction suggests that the effects of need and beliefs of deserving for a resource
could have different impacts on behavior in a social dilemma. The present author is
currently planning an experiment to explore the need principle in a social dilemma.
Subjects will share a supply of 7-Up (the commons), social relations will be
manipulated with a procedure similar to the one used in the present study, and need
for the resource will be created by manipulating group members’ thirst. The
dependent variable will be the amount of liquid withdrawn from the commons.
Men

A growing number of theorists and researchers in a growing number of
disciplines have predicted a world experiencing a dwindling resource supply and
commons problems on a vastly larger scale than the commons scenario painted by
Hardin (1968). "If present trends continue, the world in [the year] 2000 will be more
crowded, more polluted, less stable ecologically, and more vulnerable to disruption
than the world be now live in. Serious Stresses involving population, resources, and
environment are clearly visible ahead" (Global 2000 Report, p. 11). So concluded the
Global 2000 report that was commissioned by President Jimmy Carter to project world
conditions at the turn of the century. The Worldwatch Institute’s yearly "State of the
world" assessment reached conclusions similar to the Global 2000 report (Brown,
1989). Projections indicate that in the coming years there will be too few resources to
meet projected demand (cf. Meadows et al., 1972). Ophuls (1977) predicts "shortages

of energy and mineral resources (e.g., petroleum and coal) necessary to keep the

72

engines of industrial production running" (p. 8). For example, in the United States,
half of the petroleum reserves were consumed by 1975 and 80% of the supply may be
depleted by 2000. The world will experience a depletion of half of the petroleum
reserves by 2025 and an 80% depletion by 2050 (Hubbert, 1976) (see Miller, 1980).
Confronted with a future in which resource scarcity will play a predominant role--
resource scarcity, for the most part due to increases in demand (i.e., population)--it is
critical to understand psychological, individual, and social reSponses to scarcity. The
present study may be considered as an attempt to understand psychological and social

responses to resource scarcity.

FOOTNOTES

 

 

FOOTNOTES

1Note that the distinction between the characteristics of each class of norms are
not necessarily Sharp and the classes may overlap with each Other. For example, the
general interaction nouns presented below have similarities with social norms (i.e.,
they operate in a similar manner).

2Researchers interested in computing Strength of relationship effecrs may refer to
the ANOVA summary tables to obtain the appropriate information.

3Since the cohesion manipulation in the present study presented feedback
indicating that the group successfully solved the problem, it cannot be stated

conclusively that discussion alone created cohesion.

73

 

 

REFERENCES

REFERENCES

Back, K. (1951). Influence through social communication. 19prn_al pf Abnormal an_d
M Psychology, 4_6, 9-23.

Baird, J. 8., Jr. (1982). Conservation of the commons: Effects of group cohesiveness
and prior sharing. Ml pf Communig Psychology, m, 210-215.

Berkowitz, L. (1954). Group Standards, cohesiveness and productivity. imp

Relations 1, 509-519.

 

Bixentine, V. E., Levitt, C. A., & Wilson, K. V. (1966). Collaboration among six
persons in a prisoner’s dilemma game. mg pf Conflict Resolu_tic_>p, m,
488-496.

Bonner, H. (1959). G_ro_pp dynamics: Principles an_d applications. New York: Ronald
Press.

Brewer, M. B. (1979). In-group bias in the minimal intergroup situation: A

cognitive-motivational analysis. Psychological Bulletin M, 307-324.

 

Brewer, M. B. (1981). Ethnocentrism and its role in interpersonal trust. In M. B.
Brewer & B. E. Collins (Eds), Scientific Main: an_d th_e social sciences. New
York: Jossey-Bass.

Brewer, M. B., & Campbell, D. T. (1976). Ethnocentrism EB intergzoup a_t_t_i_tt_1_t_l_e_§:

gait african evidence. New York: Halstead Press.

74

75
Brewer, M. B., & Kramer, R. M. (1986). Choice behavior in social dilemmas: Effects

of social identity, group size, and decision framing. 193L113; o_f Personalig ppg
_S_O_c_iz_rl_ Psychology, _5_0, 1-7.

Brewer, M. B., & Silver, M. (1978). Ingroup bias as a funtion of task characteristics.
Eurogan Jo_umg o_f _S_O_cl_al_ Psychology, 8, 393-400.

Brown L. (1989). SMatp o_f @ w_orl_d_ 1_9_8_9_. New York: W. W. Norton.

Campbell, D. T. (1958). Common fate, similarity, and other indices of the status of
aggregates of persons as social entities. Behavioral Mop, 3, 14-25.

Cartwright, D. (1968). The nature of group cohesiveness. In D. Cartwright & A.
Zander, (jar-p dynamics: Research an_d W (3rd ed.). New York: Harper
and Row.

Cartwright, D., & Zander, A. (1960). @213. dynamics: Research an_d t_hg)_ry (2nd ed.).
Evanston, 11: Row, Peterson.

Clark,-M. S., & Mills, J. (1979). Interpersonal attraction in exchange relationships.
Jpprfll o_f Personalig m .S_O_C_ifl Psychology, 31, 12-24.

Coleman, J. S. (1961). BM. pp non-market decision-making. New York: Vantage
Press.

Dawes, R. M. (1980). Social dilemmas. Ami _R_e_vi_e_vy o_f Psychology, _3_l_, 169-193.

Deutsch, M. (1975). Equity, equality, and need: What determines which value will be

 

76
Deutsch, M. (1982). Interdependence and psychological orientation. In V. J. Derlega

& J. Grzelak (Eds), Cooperation fli helping behavior: Theories M. w.
New York: Academic Press.

Dion, K. L., Miller, N., & Magnan, M. N. (1971). Cohesiveness and social
responsibility as determinants of group risk taking. J_o_um_al o_f Personality ppg
M Psychology, g), 400-406.

Edney, J. J., & Harper, C. S. (1978). Herosim in a resource crisis: A simulation study.
Environmental Management, _2_, 523-527.

Festinger, L. (1954). A thoery of social comparison processes. Himp w, 1,
1 17-140.

Fox, J ., & Guyer, M. (1978). "Public" choice and cooperation in N-person prisoner’s

dilemma. Journal o_f Conflict Resolution 22, 469-481.

 

French, J. R. P., Jr., & Raven, R. (1959). The bases of social power. In D.
Cartwright (Ed.), Mp _ip S_oc_i_a_l m. Ann Arbor, MI: Institute for Social
Research.

Gaertner, S. L., Mann, J., Murrell, A., & Dovidio, J. F. (1989). Reducing intergroup
bias: The benefits of recategorization. m o_f Personality app 89%
Psychology, 5‘1, 239-249.

9191311 _2_0_OQ M Q @ president. New York: Penguin.

Goffman, E. (1961). Am; New York: Anchor Books.

Greenberg, J., & Cohen, R. L. (1982). Egpifl app Mp i_n Social Behavior. New

York: Academic Press.

77
Hardin, G. (1968). The tragedy of the commons. Science, 166, 1243-1248.

Hardin, G. (1977). The: m o_f m1} Ag ecologist’s flew 9f su_rv_i_y_a_l.
Bloomington, IN: Indiana University Press.

Homans, C. (1961). _S_o_c_i;rl_ M: lg elemenm limp. Boston: Harcourt.

Hubbert, M. K. (1971). The energy resources of the earth. Scientific Ameripa_n.

Jerdee, T. H., & Rosen, B. (1974). Effects of opportunity to communicate and
visibility of individual decisions on behavior in the common interest. J_o_tr_r_pa_l o_f
mp Psychology, _52, 712-716.

Johnson, D. W., & Johnson, F. P. (1975). Elli—"g m: Group them M g_r9_up
_S_ki_ll_s. Englewood Cliffs, NJ: Prentice-Hall.

Karniol, R., & Miller, D. T. (1981). Morality and the development of conceptions of
justice. In M. J. Lerner & S. C. Lerner (Eds), mjpsiici My; _ip ppgjal
M. New York: Plenum.

Karuza, J., & Leventhal, G. S. (1976, September). Justice judgments: Role demands
and perception of fairness. Paper presented at the meeting of the American
Psychogical Association, Washington, D. C.

Kelley, H. H., & Grzelak, J. (1972). Social interaction basis of cooperators’ and

 

competitors’ beliefs about others. Journal o_f Personality QM. Mal Psychology,
2_1, 190-197.

Kelley, H. H., & Stahelski, A. J. (1970). Social interaction basis of COOperatorS’ and
Competitors’ beliefs about others. 120.1131. pf Personality app S_oc_Lal Psychology,

_1__6_, 66-91.

78

Kellman, H. C. (1958). Compliance, identification, and intematization: Three
processes of attitude change. Jo_um_a_l pf wig Resolution, 2, 51-60.

Kerr, N. L. (1983). Motivation losses in small groups: A social dilemma analysis.
M o_f Personalig ppg M Psychology, 45, 819-828.

Kerr, N. L. (1989). Illusions of efficacy: The effects of group size on perceived
efficacy in social dilemmas. mm}; _o_f_ ExperimeLtal Social Psychology, g,
287-313.

Kerr, N. L. (in press). Norms in social dilemmas. In D. Schroeder (Ed.), m1
dilemmas.

Kerr, N. L., Sattler, D. N., & Kaufman, C. (1990). Efficacy and social identity in
social dilemmas. Manuscript in preparation, Michigan State University.

Knight, G. P., & Dubro, A. P. (1984). Cooperative, competitive, and individualistic
social values: An individualized regression and clustering approach. Jo_u_mpl pf
Personalig a_ng Spcigl Psychology, 4_6,, 98-105.

Komorita, S. (1976). A model of the N-person dilemma-type game. 1mg; _(fi
Exmrimental _S_Qc_i_al, Psychology, 1_2_, 357-373.

Kramer, R.M. & Brewer, MB. (1984). Effects of group identity on resource use in a
simulated commons dilemma. Mal pf Personalig app _S_pc_i_t_rl_ Psychology,
fl, 1044—1057.

Kuhlman, D. M., & Marshello, A. (1975). Individual differences in game motivation
as moderators of preprogrammed strategic effects in prisoner’s dilemma.

Journal o_f Personaligy E9 Social Psychology, 32, 922-931.

79
Lane, I. M., & Messé, L. A. (1971). Equity and the distribution of rewards. Journal

o_f Personality a_ng Sofia] Psychology, M, 1-17.

Latané, B., & Darley, J. M. (1970). The pnresponsive bystander: Why doesn’t h_e
h_el_p? New York: Appleton-Century Crofts.

Latané, B., Williams, K., & Harkins, S. (1979). Many hands make light the work:
The causes and consequences of social loafing. 1m o_f Personality ang
M Psychology, 3_7_, 822-832.

Lenski, G. (1966). _Po_wg m Mpg; New York: McGraw-Hill.

Lerner, M. J. (1975). The justice motive in social behavior: An introducrion. W
o_f 590411 1.8.520. _3_l_, 1-20.

Lerner, M. J. (1977). The justice motive: Some hypotheses as to its origins and forms.
Mum—211 M Personalig, 4_5_, 1-52.

Lerner, M. J ., Miller, D. T., & Holmes, J. G. (1976). Deserving and the emergence of
forms of justice. In L. Berkowitz and E. Walster (Eds), Advances ip
Exmrimental Soc—La; Psychology, (Vol. 9). New York: Academic Press.

Leventhal, G. S., & Anderson, D. (1970). Self-interest and the maintenance of equity.
Mal o_f Personality an_d M Psychology, 15, 57-62.

Leventhal, G. S. (1976). The distribution of rewards and resources in groups and
organizations. In L. Berkowitz & E. Walster (Eds), Advances Q Experimental
M Psychology, (Vol. 9). New York: Academic Press.

Leventhal, G. S., & Weiss, T. (1975). Perceived need and the response to inequitable

distributions of rewards. Unpublished manuscript, Wayne State University.

 

 

 

 

80
Levine, R. A., & Campbell, D. T. (1972). Ethnocentrism: Theories o_f conflict, ethnic

attitudes, fl! grgp behavior. New York: Wiley.

Liebrand, W. B. G. (1983). A classification of social dilemma games. Simulation a_n_g

Games M, 123-138.

 

Liebrand, W. B. G. (1984). The effect of social motives, communication, and group
size on behaviour in an N-person multi-stage mixed-motive game. European
J_o_u_m_al_ o_f m1 Psychology, 14, 239-264.

Luce, R. D., & Raiffa, H. (1957). Games £1191. dilemmas. New York: Wiley.

Maki, J. B., Hoffman, D. M., & Berk, R. A. (1978). A time series analysis of the
impact of a water conservation campaign. Evaluation Quarterly, _2_, 107-118.

McClintock, C. G. (1972). Social motivation: A set of propositions. Behavioral
m, 1_7, 438-454.

McGrath, J. E. (1984). m: Interaction m performance. Englewood Cliffs, NJ:
Prentice-Hall.

Meadows, D. H., Meadows, M. L., Randers, J., & Behrens, W. W. (1972). TM [i_rrga
_tp w. New York: Signet.

Messé, L. A., Kerr, N. L., & Sattler, D. N. (in press). "But some animals are more
equal than others" Privileged role effects in collective enterprise. In S.
Worchel, W. Wood, J. Simpson (Eds), Perspectives 1p mp productivity a_n_c_l_
processes. Newbury Park, CA: Sage.

Messick, D. M., & Brewer, M. B. (1983). Solving social dilemmas: A review. In L.

Wheller & P. Shaver (Eds), Review o_f pprsonalig apd social psychology, Vol.

81
4, pp. 11-44. Beverely Hills, CA: Sage.

Messick, D. M., & McClelland, C. L. (1983). Social traps and temporal traps.
Personalig anp £0131. Psychology ELM, 2, 105-110.

Messick, D. M., & McClintock, C. G. (1968). Motivational basis of choice in
experimental games. Jo_urn_al_ 91‘ Experimental m Psychology, 4, 1-25.
Messick, D. M., Wilke, H., Brewer, M. B., Kramer, R. M., Zemke, P. E., & Lui, L.
(1983). Individual adaptions and structural change as solutions to social

dilemmas. lou_ma_l o_f Personality aid Social Psychology, 44, 294-309.

Mikula, G. (1974). Nationality, performance, and sex as determinants of reward

 

allocation. LgprLal pf Personality an_g Social Psychology, 22, 435-440.
CA: Wadsworth.

Ophuls, W. Ego—logy an_d fl: Midi-i _(fi W. San Francisco, CA: W. H. Freeman.
Orbell, J., & Dawes, R. M. (1981). Social dilemmas. In G. Stephenson & J.
M. Davis (Eds), m _i_r_r_ appl_ie_d $9.131 psychology, Vol. 1. New York:
Wiley.

Oskamp, S., & Perlman, D. (1966). Effects of friendship and disliking on cooperation

in a mixed-motive game. Journal o_f Conflict Resolution 1_()_, 221-226.

 

Ostrom, E. (1977). Collecrive action and the tragedy of the commons. In G. Hardin

& J. Baden (Eds), Managing t_l_rp commons. San Francisco, CA: W. H.

Freeman.

82

Pepitone, A. (1971). The role of justice in interdependent decision making. M
pi Exmrimental Social Psychology. 1, 144-156.

Perelman, C. (1967). Justice and reasoning. In G. Hughes (Ed.), L4H» w apcl
M9. New York: New York University Press.

Platt, J. (1973). Social traps. American Psychologist, M, 641-651.

Pope, R. M. Jr., Stepp, J. M., & Lytle, J. S. (1975). Effects of price change on the
domestic use of water over time. Clemson University: WRRI Report, _5_6.

Pruitt, D. G. (1972). Methods for resolving differences of interest: A thoeretical

Pruitt, D. G., & Kimmel, M. J. (1977). Twenty years of experimental gaming:
Critique, synthesis, and suggestions for the future. Annual Review o_f
Psychology, _2_8_, 363-392.

Reis, H. T. (1979). The multidimensionality of justice. In R. Folger (Ed.), m as;
_o_f M: §9_c_ig psychological perspectives. New York: Plenum.

Rutkowski, G. K., Gruder, C. L., & Romer, D. (1983). Group cohesiveness, social
norms, and bystander intervention. qu_m_al _o_f Personality apcl _S_pc_ial
Psychology, M, 545-552.

Rutte, C. G., Wilke, H. A. M., & Messick, D. M. (1987). Scarcity or abundance
caused by people or the environment as determinants of behavior in the
resource dilemma. J_Op_m_al pf Exmrimental Mal Psychology, 1;, 208-216.

Sampson, E. E. (1969). Studies of status congruence. In L. Berkowitz (Ed.),

Advances ip Exmrimental Social Psychology, (Vol. 4). New York: Academic

83
Press.

Sampson, E. E. (1975). On justice as equality. km o_f Social Issues, 3_1_, 45-64.
Sattler, D. N., & Kerr, N. L. (in press). Might over morality explored: Motivational
and cognitive bases for social motives. lo_u_m_al pi Personaligy an_d _SLcial

Psychology.

Schelling, T. C. (1971). On the ecology of micromotives. fiblic Interest, 2_5, 61-98.

Schwartz, S. H. (1977). Normative influences on altruism. In L. Berkowitz (Ed.),
Advances i_rl experimental appial psychology, (Vol. 10). New York: Academic
Press.

Schwartz, S. H. & Howard, J. A. (1982). Helping and cooperation: A self-based
motivation model. In V. Derlega & J. Grzelak (Eds), grppgration and helping
M: Th_egy an_d MIC—h, New York: Academic Press.

Schwinger, T. (1981). Steuerung und Rechtfertigung sozialer Prozesse durch
Gerechtigkeitsnormen [Regulation and legitimation of social processes by
justice norms]. In W. Grunwald & H. G. Lilge (Eds), Kooperation und
Konkurrenz in Organisationen (pp. 97-107). Bern: Haupt.

Schwinger, T. (1986). The need principle of distributive justice. In H. W. Bierhodd,
R. L. Cohen, & J. Greenberg (Eds), Ill—SM i_n s_o_g_ial_ glam. New York:
Plenum.

Stern, P. C. (1976). Effects of incentives and education on resource conservation
decisions in a simulated commons dilemma. i_Oppnal pl Personaligy M Spaial

Psychology, 14, 1285-1292.

84
Swingle P. G., & Gillis, J. P. (1968). Effects of the emotional relationship between

protagonists in the prisoner’s dilemma. Jo_urn_al_ pf Personality m Ml
Psychology, 8, 160-165.

Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior.
In S. Worchel & W. Austin (Eds), Psychology pl‘ intergroup relations (pp.
7-24). Chicago: Nelson-Hall.

Thibaut, J. W., & Kelley, H. H. (1959). The social psychology Si gar—ups. New York:
Wiley.

Tomblom, K. J. (1977). Distributive justice: Typology and propositions. fiu_m_ap
Relations, 3_1, 1-24.

Tziner, A. (1979). Differential effects of group cohesiveness types: A clarifying
overview. S_oc_ial Behavior Md, Personality, l_Q, 227-239.

Walster, E., Berscheid, B., & Walster, G. W. (1973). New direCtionS in equity
research. m pl Personality an_d M Psychology, g, 151-176.

Walster, E., Walster, G. W., & Berscheid, E. (1978). My: 11mg a_n_d {es—each.
Boston: Allyn and Bacon.

Wilke, H. A. M., de Boer, K. L., & Liebrand, W. B. G. (1986). Standards of justice
and quality of power in a social dilemma situation. Brltisll m o_f marl
Psychology, _2_5, 57-65.

Winer, B. J. (1971). Statistical principles i_n experimental dLsigp. New York:

McGraw-Hill.

85

Wit, A. P. (1990, March). Categorization and cooperation in nested social dilemmas.

Paper presented at the Annual Conference on Social Dilemmas, Saparro, Japan.

APPENDIX A

EXPERIMENTAL MATERIALS

86

DEPARTMENTAL RESEARCH CONSENT FORM
TITLE OF RESEARCH PROJECT: THE GROUP STUDY

1. I have freely consented to take part in a scientific study being conducted by
David N. Sattler, under the supervision of Norbert L. Kerr, Professor of
Psychology.

2. This study is designed to investigate how various conditions affect the
decisions that people make. I may be asked to perform several tasks. One
of these tasks involves deciding how much of a resource to harvest from a
resource pool.

3. I understand that the study does not involve any risks or discomforts.

4. I understand that the results of the study will be treated in strict confidence
and that I will remain anonymous. Within these reStrictions, results of the
study will be made available to me at my request.

5. I understand that my participation in the study does not guarantee any
beneficial results to me.

6. I understand that I have the right to refuse to participate or to withdraw
from this study at any time and without penalty. I may also stop working
temporarily on any task at any time I choose. Because of the scientific
nature of the study, the experimenter may suspend the experiment at any
time.

7. I understand that, at my request, I can receive additional explanation of the
study after my participation is complete.

Date Signed

 

AFTER YOU HAVE READ AND SIGNED THIS FORM, PLEASE SLIDE IT OUT
UNDER THE CURTAIN TO YOUR BOOTH. THEN WAIT QUIETLY FOR
FURTHER INSTRUCTIONS.

87
QUESTIONNAIRE BOOKLET

Your color code:

 

Instructions: Your answers to the open-response questions in this booklet will be used
to help us create a closed-response questionnaire that will be used on campus. Please
answer the questions as completely and in as much detail as you can. You should
spend out 5 minutes on each of the Six questions. Try not to spend too much time on
any particular question. For your convenience, you will be told of the time remaining

try to answer all six questions.

88

1A. In deciding which of two companies to work for, what five factors would be
the most important ones influencing your decision?
13. Why did you select these five? Explain fully.

89

2A. What do you believe will be your job when you finish your education?
2B. What were the major factors influencing your career choice?
2C. What college courses do you believe are the most beneficial to you?

90

3A. Do you agree or disagree with the new semester system? Why?

3B. What do you believe are the advantages and disadvantages of the quarter
system?

3C. What do you believe are the advantages and disadvantages of the semester
system?

91

4A. Do you believe that your ability to learn in a college class is affected by the
class Size? Do you think you would perform better if the class size was
more than 250 students or less than 250 students? Why?

4B. What do you believe is the ideal class size? What are the benefits to a
smaller class than a larger class?

4C. What qualities make a good teacher?

92

5A. Suppose you had to put together a campaign to encourage Michigan State
University Students to participate in Health Week activities (such as
attending talks about eating healthy foods, and participating in various
physical activities). How might you structure the campaign. What key
factors would you focus on?

SB. How would you publicize the events?

5C. Do you think students would participate, even after all your efforts to
encourage them to do so? Why?

93

6A. Do you think freshmen should be required to live on campus in a dorm?
Why?

6B. What are the advantages and disadvantages to living in a dorm, rather than
living in an apartment?

6C. Do you think most students prefer to live in a dorm for their entire college
career? Why?

94
INITIAL RANKING FORM

Instructions
Indicate your solution to the winter survival task below. Rank the items in order
of importance for survival (i.e., 1 = most important, 12 = least important).
__ Ball of Steel wool
Newspapers (one per person)
Compass
Hand ax
Cigarette lighter (without fluid)
Loaded .45-caliber pistol
Sectional air map made of plastic
Twenty-by-twenty fOOt piece of heavy-duty canvas
Extra shirt and pants for each survivor
Can of shortening
Quart of lOO-proof whiskey

Family-size chocolate bar (one per person)

95
FINAL RANKING FORM

Instructions
Indicate your solution to the winter survival task below. Rank the items in order
of importance for survival (i.e., 1 = most important, 12 = least important).
__ Ball of steel wool
Newspapers (one per person)
Compass
Hand ax
Cigarette lighter (without fluid)
Loaded .45-caliber pistol
Sectional air map made of plastic
Twenty-by-twenty foot piece of heavy-duty canvas
Extra shirt and pants for each survivor
Can of shortening

Quart of 100-proof whiskey

Family-size chocolate bar (one per person)

96
WINTER SURVIVAL QUESTIONNAIRE

Your name:

 

Your color code:

 

Please answer the following questions concerning your impressions of the winter
survival problem.

(All responses are confidential and will not be shared with anyone).

1.

To what extent would you like to meet with these same people again?
not at all:_:_:_:_:_:_:_:very much

How much did you enjoy being in a group with these peOple?
not at all:_:_:_:_:_:_:_:very much

How much do you like the peOple in the group?
not at all:_:_:_:_:_:_:_:very much

How much would you like to participate in another experiment with the
same people in the group?

not at all:_:_:_:_:_:_:_:very much

How much would you like to be in a different group in today’s session?
not at all:_:_:_:_:_:_:_:very much

How much do you feel as though you are a member of a group?

not at all:_:_:_:_:_:_:_:very much

97

YOUR PARTNER
DID NOT WORK ON THE QUESTIONNAIRE

TASK AT THE BEGINNING OF THE SESSION

98

YOUR PARTNER
WORKED ON THE QUESTIONNAIRE

TASK AT THE BEGINNING OF THE SESSION

99
QUESTIONNAIRE

Your name:

 

Your color code:

 

Please answer the following questions concerning your impressions of the experiment.

1.

How much money do you think would be fair for yptl to receive for
participating in the study?

$

How much money do you think would be fair for your partner to receive for
participating in the Study?

3
How responsible do you feel for your group’s outcome?

very little:_:_:_._._._._

:very much

How many persons were in your group, including yourself?

 

Do you think that if you made cooperative harvest decisions, then the other
group member would also make cooperative harvest decisions?

other would other would
harvest cooperatively:_:__:_:_:_:_:_:harvest competitively

Do you think that if you made competitive harvest decisions, then the other
group member would also make competitive harvest decisions?

other would other would
harvest cooperatively:_:_:__:__:_:__:_:harvest competitively

There was a fine imposed for harvesting too many acres of lumber. How
much did you want to avoid this fine?

not at all:_:_:_:_:_:_:_:very much
Do you think that you made cooperative or competitive harveSt decisions?
cooperative competitive

 

 

100

One of the things we are interested in is how people in a group form impressions of
other group members. Although you don’t know your fellow group member, you still
may have formed some impressions of him or her just by being a member of the same
group. Please indicate your impressions of your fellow group member on the
following scales. We realize that this may sometimes be difficult; just do the best you
can. (These responses are confidential, and will not be Shared with the Other group
member).

9.

10.

11.

12.

You and your fellow group member had the chance of winning some money

_in today’s experiment. How much do you think your fellow group member

wanted to win this money?
very little:_:_:_:_:_:_:_:very much

How difficult do you think your fellow group member found the task?
very easy:_:_:_:_:_:_:_:very difficult

How responsible do you think your group member felt for your group’s
outcome?

not at all:_:_:_:_:_:_:_:very much

Please indicate your impressions of your group member on the following
scales.

selfish:_:_:_:_:_:_:_: generous
concerned for others:_:_:_:_:_:_:_:concemed only for self
very irresponsible:__:_:___:_:_:_:_:very responsible
honest:_:_:_:_:_:_:_:dishoneSt
friendly:_:_:_:_:_:_:___:unfriendly
trusting:_:_:_:_:_:_:_:not trusting

cooperative:_:_:_:_:_:_:_:competitive

101

Please use the following rating scale for items 13 and 14.
l definitely not applicable
2 not applicable
3 somewhat inapplicable
4 neither applicable nor inapplicable
5 somewhat applicable
6 applicable
7 fully applicable
13. If the size of the forest was 200 acres and you withdrew 32 acres--considered

to be a small harvest--how do you think your fellow group member would
perceive ypp?

A. good __ I. sincere __
B. weak __ J. pliable __
C. aggressive __ K. powerful __
D. dishonest __ L. unjust _—
E. kind __ M. helpful _—
F. timid _ N. wimpy _—
G. forceful O. dominant

H. unfair

7°

11163.11

POP”?

P1

15.

16. Please give us your general impressions of the experiment so far.

THANK YOU VERY MUCH. PLEASE SLIDE THIS FORM OUT UNDER THE
m IQ m §O_OT_H. THEN WAIT QUIETLY. WE WILL CONTINUE

If the size of the forest was 200 and you withdrew 88 acres-considered to be a
large harvest--how do you think your fellow group member would perceive

m?
good

weak

aggressive

dishonest

kind

timid

forceful

unfair

Please describe in your own words what you understand the purpose of this

experiment to be.

SHORTLY.

S"

P.

9.2:

102

sincere
pliable
powerful
unjust
helpful
wimpy
dominant

me an

 

APPENDIX B

INSTRUCTIONS

INSTRUCTIONS

EXPERIMENTER §4Y_S:

"AS YOU REMEMBER WHEN YOU SIGNED UP FOR THE EXPERIMENT,
THE EXPERIMENT LASTS FOR 1 1/2 HOURS BUT YOU ONLY RECEIVE 1
HOUR OF CREDIT. YOU’LL HAVE THE OPPORTUNITY TO EARN MONEY IN
ONE OF THE TASKS TODAY, AND THIS MONEY WILL BE THE WAY IN
WHICH YOU’LL BE COMPENSATED FOR THE 1/2 HOUR THAT YOU DON’T
RECEIVE CREDIT FOR. FOR THE FIRST 1/2 HOUR, YOU WILL NOT GET
ANY EXPERIMENTAL CREDIT, AND THE HOUR OF CREDIT WILL BE FOR
THE LAST HOUR."

[CURTAIN IS CLOSED AND PRE-RECORDED INSTRUCTION TAPE IS
STARTED]
TAPE:

"First of all, I would like to thank each of you for coming today. During today’s
session you will be working on several different tasks. During the session, you will
never be referred to by name and your personal identity will never be revealed or
known to any of the other people. Rather, you will be referred to only by color code.
A sign in your room indicates your own color code.

For the first part of the session, we would like you to complete a battery of

103

104

questionnaires concerning various issues. You will have 30 minutes to work on the

questionnaires. As you work on the questionnaire, the experimenter will inform you
every five minutes of the amount of time remaining to complete the booklet. Please
read the instructions on page one. You may begin working now."

[30 MINUTE PERIOD. EVERY FIVE MINUTES THE EXPERIMENTER SAYS,

"THERE ARE X MINUTES LEFT."

EXPERIMENTER S513:

"OKAY, I WILL NOW COLLECT THE QUESTIONNAIRE BOOKLETS.
PLEASE SIT QUIETLY IN YOUR BOOTH. THE NEXT PART OF THE SESSION
WILL BEGIN SHORTLY."

[2 ADDITIONAL SUBJECTS ARE ESCORTED INTO THE LAB. TIE POSTERS
CONTAINING THE 12 ITEMS IN THE WINTER SURVIVAL TASK ARE PLACED
IN EACH BOOTH]

EXPERIMENTER gags;

"WE ARE NOW READY TO CONTINUE THE SESSION. BEFORE WE
BEGIN, I WOULD LIKE TO TELL YOU THAT THERE ARE NOW FOUR
PEOPLE HERE. THE TWO PEOPLE WHO JUST ARRIVED WILL BE HERE FOR '
ONLY 1 HOUR. TIE OTIER TWO PEOPLE HAVE ALREADY BEEN
PARTICIPATING IN THE STUDY FOR 1/2 HOUR AND DID NOT RECEIVE AND
CREDIT FOR BEING HERE, SO FAR. THIS IS BECAUSE WIEN YOU SIGNED
UP FOR THE STUDY IT WAS WORTH ONLY 1 HOUR OF' CREDIT, BUT SOME

OF YOU ARE HERE FOR 1 1/2 HOURS. BUT FROM THIS POINT ON .

105
EVERYONE WILL RECEIVE ONE HOUR OF CREDIT. YOU SHOULD BE

AWARE THAT THE PEOPLE WHO WERE HERE FOR THE FIRST 1/2 HOUR
DID NOT RECEIVE ANY CREDTT FOR THAT TIME. INSTEAD, THE WAY

THEY’LL BE COMPENSATED FOR THIS FIRST 1/2 HOUR IS BY EARNING
MONEY IN ONE OF THE TASKS TODAY. HOWEVER, ALL OF YOU WILL
HAVE THE CHANCE TO EARN MONEY IN ONE OF THE TASKS TODAY."
[INSTRUCTION TAPE IS STARTED]

IA_PE:

"Okay, we are now ready to continue the study. The remainder of the Study will
take one hour. All four subjects will receive two credits for participating in the
session. Those of you who arrived earlier and worked 30 minutes will have a chance
of getting cash for the task you performed earlier. I would like to thank those of you
who just arrived for coming today. This study is interested in how people make a
variety of decisions, and in fact, one part of this study is interested in how a number
of different conditions affect decisions that have real monetary significance. During
the session, you will never be referred to by name and your personal identity will
never be revealed or known to any of the other people. Rather, you will be referred to
only by color code. A Sign in your room indicates your own color code.

Today you will be working on several different tasks. We are now ready to
begin the next task. You will be working on a winter survival problem. You will be
using a poster in your booth for this task. I will now describe the survival problem.

You have just crash-landed in the woods of northern Minnesota and southern

106

Manitoba. It is 11:32 am. in mid-January. The light plane in which you and the
others were traveling crashed on a lake. The pilot and COpilot were killed. Shortly
after the crash the plane sank completely into the lake with the pilot’s and cOpilot’s
bodies inside. None of the four passengers are seriously injured and you are all dry.

The crash came suddenly, before the pilot had time to radio for help or inform
anyone of your position. Since your pilOt was trying to avoid a storm, you know the
plane was considerably off course. The pilot announced shortly before the crash that
you were twenty miles northwest of a small town that is the means known habitation.

You are in a wilderness area made up of thick woods broken by many lakes and
streams. The snow depth varies from above the ankles in windswept areas to knee
deep where it has drifted. The last weather report indicated that the temperature
would reach -25 degrees Fahrenheit in the daytime and -40 at night. There is plenty
of dead wood and twigs in the immediate area. You are dressed in winter clothing
appropriate for city wear--suits, pantsuits, street shoes, and overcoats.

While escaping from the plane the four passengers salvaged 12 items. These
items are shown on the poster in your bOOth. Your task is to rank these 12 items
according to their importance to your survival, starting with one for the most
important item and ending with 12 for the least important one.

You may assume that the number of passengers is the same as the number of
persons here today, which is four, and that the group has agreed to stick together."

For the next two minutes, think about how you would rank these 12 items

according to their importance for survival, and indicate your ranking on the form in

107
your booth titled "Initial Ranking Form." You may begin to rank the items now."

[TAPE IS STOPPED. SUBJECTS ARE GIVEN THE "INITIAL RANKING FORM."]
[2 MINUTE PERIOD]
EXPERIMENTER S_AY_S:

"YOU HAVE 30 SECONDS LEFT TO FINISH RANKING THE ITEMS."
[LOW GROUP COHESION CONDITION - NO TALKING CONDITION]
_E_XPE_RIIVIENTER SAYS:

"OKAY, WE WOULD NOW LIKE TO TAKE SOME MORE TIME TO THINK
ABOUT THE WINTER SURVIVAL PROBLEM. PLEASE OPEN THE CURTAIN
TO YOUR BOOTH AND PLACE YOUR CHAIRS IN THE CENTER OF THE
ROOM. YOU SHOULD TAKE THE POSTER AND THE INITIAL RANKING
FORM WITH YOU. [WAIT]. WE WOULD LIKE YOU TO THINK ABOUT THE
SURVIVAL PROBLEM FOR THE NEXT FIVE MINUTES. PLEASE DO NOT
TALK TO THE OTHER PEOPLE HERE. YOU SHOULD THINK ABOUT HOW
BEST TO RANK THE ITEMS."

[EGH GROUP COHESION CONDITION - TALKING CONDITION]
EXPERIMENTER SA_Y_S_:

"OKAY, NOW THAT YOU HAVE HAD A CHANCE TO GET WARMED UP,
WE WOULD LIKE YOU TO DISCUSS THE WINTER SURVIVAL PROBLEM AS
A GROUP. PLEASE OPEN THE CURTAIN TO YOUR BOOTH AND PLACE
YOUR CHAIRS IN A CIRCLE IN THE MIDDLE OF THE ROOM. [GET 1

POSTER]. "WE WOULD LIKE YOU TO DISCUSS TI-E SURVIVAL PROBLEM

108
WITH TI-E OTI-ER GROUP MEMBERS FOR FIVE MINUTES. YOU SHOULD

TRY TO IDENTIFY THE BEST RANKING OF THE 12 ITEMS IN TERMS OF
TI-EIR IMPORTANCE FOR SURVIVAL."
[5 MINUTES PERIOD]
EXPERIMENTER SA_Y_S:
"EXCUSE ME. THERE ARE 30 SECONDS LEFT."
(AFTER THE 30 SECONDS, EXPERIMENTER 84E:

"OKAY, I WOULD NOW LIKE YOU TO RETURN TO YOUR BOOTHS."
[CURTAINS ARE CLOSED AND THE TAPE IS STARTED]

M:

"Okay, at this time we would like you to once again rank the 12 items in terms
of their importance for survival. Please indicate your ranking on the "Final Ranking
Form." Your final ranking will be scored. Your ranking will be compared to the
rankings made by experts and participants who worked on the winter survival task last
term. After you have indicated your ranking, please Slide the "Final Ranking Form"
out under the curtain to your booth. Then wait quietly. We will score your rankings
by determining how close the average group ranking agreed with the expert rankings,
and will give you the results in a few moments."

[TAPE IS STOPPED. SUBJECTS ARE GIVEN THE FINAL RANKING FORM.
AFTER 1 1/2 MINUTES, THE FEEDBACK SHEETS ARE GIVEN TO SUBJECT 8.]

[LOW GROUP COIESION CONDITION - NO TALKING CONDITION]

EXPERIMENTER SAYS:

109
"THE NUMBER ON THIS SI-EET SHOWS HOW WELL YOU DID ON TI-E

TASK COMPARED TO THOSE LAST TERM. THE NUMBER SHOWS YOU
YOUR PERCENTlLEuTHAT IS, THE PERCENT OF PEOPLE YOU SCORED
HIGHER THAN ."

lflGH GROUP COHESION CONDITION - TALKING CONDITION]
EXPERIMENTER SAYS:

"YOUR GROUP SCORED AT THE 82ND PERCENTILE. THIS MEANS
THAT YOUR GROUP DID BETTER THAN 82% OF THE GROUPS LAST TERM."
[WINTER SURVIVAL QUESTIONNAIRE IS GIVEN TO SUBJECTS]
EXPERIMENTER SAYS:

"PLEASE COMPLETE THIS SHORT QUESTIONNAIRE AND SLIDE TT
UNDER THE CURTAIN WHEN YOU HAVE FILLED IT OUT."

[POSTERS ARE REMOVED FROM BOOTHS AND TAPE IS STARTED].
TALE;

"Okay, the next part of the session is now ready to begin. For this part of the
session we have formed two separate groups with two persons in each group. In order
to provide complete anonymity, you will not know which booth your fellow group
member is in. I need to emphasize that these two groups are not competing against
one another in any way. There will be no interaction between groups and the outcome
of each group depends only on its own performance, not on the performance of the
Other group. The only reason we are running two groups at the same time iS because

we can complete our Study faster with two groups at once, rather than just testing one

110

group at a time.

Your group member is determined at random, and each of you have an equal
chance of being paired with one of the other persons here today. The experimenter
will give you a sheet of paper that indicates whether or not your partner worked on a
questionnaire task at the beginning of the study."

[PAUSE]. [SUBJECTS ARE GIVEN TIE PAPER WITH PARTNER’S
QUESTIONNAIRE STATUS].
ELXPERIMENTER SAYS:

"THIS SHEET TELLS YOU WHETHER OR NOT YOUR PARTNER WAS
IERE FOR TIE FIRST 1/2 HOUR. REMEMBER, THOSE WHO WERE IERE
FOR TIE FIRST 1/2 HOUR DID NOT RECEIVE ANY CREDTT FOR THE FIRST
1/2 HOUR. IN TIE NEXT TASK, EVERYONE WILL HAVE TIE
OPPORTUNITY TO EARN MONEY, BUT FOR THOSE WHO WERE IERE FOR
TIE FIRST 1/2 HOUR, TIE TASK WILL ALSO SERVE AS A WAY YOU ARE
COMPENSATED FOR PARTICIPATING DURING THAT TIME."

[TAPE IS STARTED]
M:

"Each of you have the opportunity to earn money in the next part of today’s
session. In fact, this amount could be rather large. The amount of money you may
earn depends in part on the decisions made by you and by the other member of your
group. I would like to stress that all the money you earn today will actually be paid

to you at the end of the study. Since we are interested in how people make decisions

111

when they know those decisions will affect how much money they earn, it is important
that you really do earn money as a result of your choices.

Now let me describe the task you will perform in the next part of the session and
the rules of the task. We are simulating a situation in which members of a group
harvest acres of lumber from a forest that the group members share use of. Your task
is to harvest as many acres of lumber from the forest for yourself as possible. The
forest contains 200 acres, and each person may harvest any number between 20 and
100 acres during each "season" or trial. Each group member will individually and
privately decide how many acres to harvest.

Since the group members will be harvesting acres of lumber from the same forest
and overharvesting can destroy the forest, the government regulates the harvesting. A
government agency discourages overuse of the resource through a fine system. Since
harvesting is regulated, the number of acres of lumber you finally end up with on each
trial is determined by your own harvest and also by your group member’s harvest.

The govemment’s fine or confiscation system works in the following way. As the
t_ot_a1_ number of acres harvested by your group increases, the government fine
increases. Each member shares M responsibility for the fine. For example, if the
total number of acres harvested by both you and your group member were 104 acres,
then a. fine of 80 acres would be imposed, and each member of your two-person group
must pay half of the fine, or 40 acres. It does not matter which person harvested a
larger share of the 104 acre total. Each group member must pay an equal part of the

fine. As you can see, the number of acres you finally end up with is in part

112

determined by the fine imposed on each trial. Each group member will receive the
number of acres he or she harvests from the forest Ma his or her share of the fine.
If the fine is larger than the number of acres you have, then the total number of acres
you have will become zero. The number of acres of lumber you have cannot go
below zero.

There are two purposes for the harvesting task. First, we are interested in how
people make decisions when money is involved. All subjects will participate in the
harvesting task and all have the opportunity to get money, but for those who also
worked on the 30 minute task at the beginning, this is how you will get compensated
for working at the task. You may earn money in the following way. Each acre of
lumber you accumulate during the study will be worth one penny. At the end of the
study you will be paid, in cash, one penny for each acre of lumber you end up with.
There will be between 5 - 10 trials. I want to stress that the number of acres of
lumber you accumulate and the amount that you are paid for the acres of lumber will“
not be revealed to or known to any of the other people here today. Furthermore, each
of you will be paid in private and will leave the room at separate times.

In a moment four example screens will be presented on the screen on your desk
to familiarize you with the task. The example screens are similar to the screen on the
actual trials and present four hypothetical trials in which members of a two person
group harvest acres of lumber from a forest. Since the example screens are only
examples, you will not receive the amount of lumber indicated on your screen.

EXAMPLE SCREEN 1

113

This is the first example screen. Let’s look at the way the screen is set up. The
first line shows the trial number, located on the left, the clock in the middle, and your
booth color on the right. The second line shows information about m gr_o_up. The
third line Shows information about m M current status. And the fourth line shows
information relating to ypu_r mp member’s status.

Let’s look more closely at the third line. The amount you harvest is Shown at
the left. In this example, you take 65 acres of lumber. Immediately below on the
fourth line shows the amount your group member harvests, which in this example is
55 acres. The total taken by your team is indicated on the second line at the left. In
this example, your group harvests 120 acres. And how much is the team fined? The
second line in the middle shows that the group is fined 100 acres of lumber. The
portion of the fine that you owe is shown below the team fine, on the third line in the
middle, and the portion of the fine that the Other member of your group owes is Shown
below your fine, on the fourth line in the middle. Remember that the group members
share the fine equally. So just how much would you get on this trial? On the third
line on the right, you can see that you would get 15 acres of lumber. The amount that
your partner would receive, five acres, is shown at the far right on the fourth line.
EXAMPLE SCREEN 2

Here is another example screen Notice again that the trial number is indicated
in the upper left comer, and the time remaining for you to indicate how many acres
you would like to harvest is indicated in the middle. When the clock begins you will

have 15 seconds to indicate your harvest decision by rotating the knob on the game

114

paddle on your desk. The clock will count down as is now shown. Your booth color
is indicated on the far right.

In this example, you take 20 acres, as is Shown at the left on the third line, and
your partner also takes 20 acres, as is shown at the left on the fourth line. The
number of acres your group takes, that is, the combined harvest by you and your
group member, is 40 acres, as is shown at the left on the second line. Now look at
the middle of the second line. Here, you can see that the team is fined zero acres.
The middle of the third and fourth lines Show that you and the Other group member
would not be fined. The tOtal amount you would receive, 20 acres, is Shown at the
right Side of the third line, and the amount your partner would receive, 20 acres, is
shown at the right side of the fourth line. In this example, both group members made
small harvests and there was no fine.

EXAMPLE SCREEN 3

In this next example screen, you harvested a lot of acres whereas your group
member harvested a small number of acres. The left side of the third line Shows that
you took 100 acres, and the left side of the fourth line shows that your partner took 20
acres. The left side of the second line shows the group harvested 120 acres, and the
middle of the second line shows the group is fined 100 acres. Because the fines are
equally shared by the group members, both group members are fined 50 acres, as is
shown on the middle of the third and fourth lines. To find out how much you and the
other group member receive, look at the third and fourth lines, respectively. Here, you

would receive 50 acres and the other group member would receive -30 acres. If either

115

you or your partner end up with a negative amount, as did your partner in this
example, you just end up with zero points. The total number of acres you end up with
cannot be negative.

EXAMPLE SCREEN 4

Here’s the last example screen. In this example, both you and the other group
member make large harvests. At the left side of the third line you see that you took
100 acres, and right below on the fourth line you see that your partner took 100 acres.
And the group, that is, you and your group member, took a total of 200 acres, as you
can see at the left of the second line. The middle of the second line Shows that the
team is fined 200 acres, and the middle of the third line shows you are fined 100 acres
and right below on the fourth line you see your partner is fined 100 acres. In this
example, you would get zero acres, as is Shown at the right of the third line, and your
partner would get zero acres, as is shown just below on the fourth line.

Again, the trial number, the time remaining to make your harvest decision, and
your booth color are indicated on the fust line. The second line shows the number of
acres harvested by the group, the group fine, and the number of acres the group
receives. The third line shows ytarl harvest selection, yp_u_r fine, and the number of
acres m receive. The fourth line shows the selection of your group member, the
fine, and the amount the other member receives.

In a few moments, we will begin the actual trials. To indicate how many acres
you would like to harvest, turn the knob on the game paddle on your desk back and

forth. As you turn the knob, the number of acres of lumber you would like to harvest

116

will change on your screen. You will have 25 seconds to indicate the number Of acres
of lumber you would like to harvest from the forest. The amount of time you have
left to indicate this decision is indicated on the first line of your screen. When 25
seconds have elapsed, the computer will record the number of acres indicated on your
screen as your harvest decision.

The computer will signal you on your screen when to make you harvest decision.
Just before each trial begins you will see the number of the trial and a "READY"
Signal on the screen. Then, in a second or two, a "G0" signal will appear on the
screen. You should then make your harvest selection. There will be a brief period
before the next trial begins.

Remember, to indicate the number of acres you would like to harvest, turn the
knob. Again, you will be paid one penny for each acre you finally end up with. The
trials will now begin."

[STOP TAPE]
EXPERIMENTER SAYS:

YOU SHOULD ALSO BE AWARE THAT IN SOME SESSIONS THERE IS A
FEEDBACK CONDITION AND IN OTHER SESSIONS TIERE IS A NO
FEEDBACK CONDITION. YOU ARE IN THE NO FEEDBACK CONDITION AND
WILL NOT RECEIVE ANY INFORMATION ABOUT YOU PARTNER’S
CHOICES.

ALSO REMEMBER THAT TIE TWO PEOPLE WHO WERE IERE FOR

TIE FIRST HALF HOUR AND WHO WORKED ON TIE BATTERY OF

117
QUESTIONNAIRES DID NOT RECEIVE ANY CREDIT FOR THE FIRST 1/2

HOUR. TIE HARVESTING TASK WILL SERVE AS TIE WAY IN WHICH
TIESE TWO PEOPLE ARE COMPENSATED FOR PARTICIPATING DURING
THAT TIME, ALTHOUGH EVERYONE WILL HAVE TIE OPPORTUNITY TO
EARN MONEY DURING TIE TASK.

[HARVEST TRIALS BEGIN .]

"That completes the series of trials on this task. Before continuing, we would
like to quickly get some of your impressions of the study to this point in the session.
In a moment, the experimenter will Slip an envelope containing a Short questionnaire
into your booth. Please fill it out completely. Then put it back into the envelope, Slip

it out under the curtain to your booth, and wait quietly for further instructions."

APPENDIX C
ANOVA RESULTS ON THE GROUP COHESION

MAN IPULATION CHECK ITEMS

ANOVA RESULTS ON THE GROUP COHESION

MAN IPULATION CHECK ITEMS

To what extent would you like to meet with these same peOple again?,

F(l, 143) = 6.13, p < .01. (high, M = 4.68; low, M = 4.15).

How much did you enjoy being in a group with these peOple?, F(1, 143) =
25.63, p < .0001. (high, M = 4.91; low, M = 3.84).

How much do you like the peOple in the group?, 5(1, 149) = 18.55, p < .0001.
(high, M = 4.89; low, M = 4.23).

How much would you like to participate in another experiment with the

same people in the group?, 13(1, 143) = 11.90, p < .001. (high, M = 5.01; low,
M = 4.36).

How much would you like to be in a different group in today’s session?,

5(1, 143) = 7.74, p < .01. (high, M = 3.00; low, M = 3.53).

How much do you feel as though you are a member of a group?, 5(1, 143) =

83.06, p < .0001. (high, M = 4.99; low, M = 2.53).

118

liliiilililliis

 

"I