UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Depersonalization in structured groups Fritz, Anna Sabine 1987

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1987_A8 F74.pdf [ 3.67MB ]
JSON: 831-1.0097058.json
JSON-LD: 831-1.0097058-ld.json
RDF/XML (Pretty): 831-1.0097058-rdf.xml
RDF/JSON: 831-1.0097058-rdf.json
Turtle: 831-1.0097058-turtle.txt
N-Triples: 831-1.0097058-rdf-ntriples.txt
Original Record: 831-1.0097058-source.json
Full Text

Full Text

DEPERSONALIZATION EN STRUCTURED GROUPS by ANNA SABINE FRITZ B.A. (Psychology), Reed College, 1982  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF  ARTS  in THE FACULTY OF GRADUATE STUDIES Psychology  We accept this thesis as conforming to the required standard  THE UNIVERSITY  OF BRITISH COLUMBIA  December 1986  ANNA SABINE FRITZ, 1986  In p r e s e n t i n g  this  requirements  f o r an  of  British  it  freely  available  understood  that  financial  shall  for reference  and  study.  I  f o r extensive copying of be  her  copying or shall  g r a n t e d by  not  be  "Psj cU.o\o^rj  The U n i v e r s i t y o f B r i t i s h 1956 Main Mall V a n c o u v e r , Canada V6T 1Y3  Date  )E-6  f3/8'l}  of  '  Columbia  make  further this  thesis  head o f  this  my  It is thesis  a l l o w e d w i t h o u t my  permission.  Department o f  the  representatives. publication  the  University  Library  h i s or  gain  the  the  s c h o l a r l y purposes, may by  f u l f i l m e n t of  I agree that  permission  department or  for  in partial  advanced degree a t  Columbia,  agree t h a t for  thesis  written  ABSTRACT  This thesis examined a topic from the field of intergroup relations, namely the consequences of the process of depersonalization. According to Turner, depersonalization is that process  whereby people  come  to perceive  themselves and others  more as  interchangeable exemplars of a social category than as unique personalities defined by their  differences  from  others.  Based  on  research involving simple  or unstructured  groups, he formulated the consequences of depersonalization in his Assimilation-Contrast Model  as  differences  the  enhancement  (contrast)  of  with  intragroup similarities  a  pro-ingroup  bias.  (assimilation) The  and intergroup  generality  of  the  Assimilation-Contrast Model has recently been challenged by Smith's Person-Situation Model of depersonalization, on the grounds that the assimilation-contrast effect may not be observed in complex or structured groups.  The present work derived and tested the validity of three sets of predictions on the behavior of structured and unstructured groups under depersonalized conditions based on the Assimilation-Contrast and Person-Situation Models. To this end, structured and unstructured groups were studied under two levels of depersonalization, one level of  non-explicit  comparison.  This  non-categorized males  outgroup resulted  comparison in  four  in  the  a  second  experimental  individuals functioned  participated  and  experiment  as  level  conditions.  of A  explicit fifth  a control group. Fifty-eight  which  consisted  of  a  outgroup  condition  of  college aged  game-like  procedure  (brainstorming task). Subjects participated in a group (experimental conditions) or alone (control condition). The assimilation-contrast effect was assessed in three ways. Subjects were asked about their overall impressions regarding the group and the individual  ii  ingroup members ('global questionnaire items'), they were asked to rate other ingroup members on a number of personal attributes ('personal attribute ratings'), and thirdly, they were asked to rate the products of the ingroup as well as those of the outgroup ('product ratings'). There were four major Findings: (a) In unstructured groups, ingroup assimilation was more pronounced under higher levels of depersonalization than under lower  levels.  This  was  in  line  with  previous  research  findings  involving  the  Assimilation-Contrast Model.(b) In structured groups, ingroup assimilation was less under higher levels of depersonalization than it was under lower levels. This was predicted by  the  Person-Situation  assimilation-contrast by either  Model.  (c)  Structured  groups  expressed  more  behavior than unstructured groups. This finding was not predicted  model, (d)  In structured as well  as unstructured groups, more outgroup  contrast was observed under high levels of depersonalization  than under low levels.  This was in line with the predictions of the Assimilation-Contrast Model.  The findings showed that all three sets of predictions were found to be useful in describing a certain component of the behavior of the groups under study. It was concluded  that  the  assimilation-contrast  effect  phenomenon than originally suggested by Turner.  iii  may  constitute  less  monolithic  a  TABLE OF CONTENTS Abstract  ii  Acknowledgements  vii  1. THE RESEARCH QUESTION What is a social group? Group behavior Depersonalization and the Assimilation-Contrast Model The Person-Situation Model Predictions based on the models Summary  „. 1 2 4 6 9 12 16  _  2. METHOD Design Subjects Procedure Conditions Dependent measures Predictions Scales, scoring and analyses  18 18 19 19 22 23 25 27  3. RESULTS Manipulation check - categorization procedure Liking-Evaluation indices Global questionnaire items Personal attribute ratings Product ratings Similarity indices Global questionnaire items Personal attribute ratings Product ratings  29 31 32 32 36 37 38 38 39 39  _  4. DISCUSSION Similarity indices versus Liking-Evaluation indices The three types of measures used Spontaneous self-categorization - control condition Conclusion  41 50 51 54 55  References  58  Table Captions  63  Table I  64  Table U  66  Table III  ~  iv  67  Table IV Table V  68 :  69  Figure Captions  70  Figure 1  •.  71  Figure 2  72  Figure 3  73  Figure 4  74  Appendix A -  Instructions for Brainstorming Task  75  Appendix B -  Rating Sheets  77  Appendix C -  Questionnaires I and II  82  Appendix D -  Consent and Debriefing Forms  89  v  ACKNOWLEDGEMENTS  On this page, I wish to express my appreciation for all the help I received on the completion of this thesis. I am  very  grateful to my  supervisor Dr. Philip Smith for his valuable guidance,  support and encouragement at all times. I wish to thank Dr. Carol Martin and Dr. Susan Butt for serving on my thesis cx)rnrnittee, especially Carol's active interest and helpful criticism are much appreciated. My thanks to Suzanne Hala for her good job as co-experimenter and to Jon Druhan for helping me out with the experimental sessions ("for living life before noon"). I am very thankful to Lorraine Ball, Warren Weir and Ed Wishnow for the many ways in which they helped me.  vi  1. THE RESEARCH QUESTION  Depersonalization is a hypothetical concept. It has been used to describe the cognitive process whereby people come to think of themselves and others in terms of representative  group characteristics  rather than individual characteristics.  According to  Turner (1985) who formulated the concept, through depersonalization people come to perceive themselves and others more as interchangeable exemplars of a social category than  as  unique  personalities  defined  by  their  differences  from  others.  He sees  depersonalization to be a central process to group behavior affecting people's thought, feeling and action towards members of their own and other groups. To date, there is much empirical evidence to support the theory of depersonalization as an adequate way of conceptualizing the cognitive components of group behavior.  In the Assimilation-Contrast Model  (Tajfel,  1969;  Turner,  1985), Turner has  described what he believes to be the consequences of depersonalization. Recently, Smith (1985) has questioned Turner's assumptions of the Assimilation-Contrast Model and has proposed an alternative  formulation, the Person-Situation Model  of depersonalization.  The present research studies the consequences of the depersonalization process. The two models,  the  investigated  Person-Situation in  comparison  to  Model one  and another,  the in  Assimilation-Contrast this  way  to  understanding of the consequences of the depersonalization process.  1  Model  contribute  to  are our  1. THE RESEARCH QUESTION / 2 WHAT IS A SOCIAL GROUP?  An entire nation may be called a group. A few people in a face to face interaction may also comprise a group. In both instances, the individuals involved share a common social identification, or, in other words perceive themselves as belonging to the same social group or category (Turner, 1981, 1982).  This definition of social group emphasizes the perceptual and cognitive bases of group formation. It is central to Turner's Theory of Self-Categorization (1985) which is a further development  of Social Identity  Theory (Tajfel, 1978,  1981;  Turner,  1975,  1981). In contrast, traditional approaches to group behavior in social psychology assume groups form as a result of interdependency between individuals. According to them, a group is defined as two or more persons who are psychologically interdependent as expressed in their interactions and influences of one another and their attractions to each other. Turner (1982, 1985) calls this most popular theory of group behavior the Social Cohesion Model or interdependence perspective. We must replace it, he argues, by the alternative Self Categorization Theory (1985) of group behavior which is: a) a more generic theory in that it attempts to explain group phenomena in any social group, from racial or national groups to small face to face groups, and b) in line with empirical Findings of what constitutes the necessary and sufficient antecedents of group formation. The interdependence  perspective  is preoccupied with the  study of  group behavior as expressed in cohesive or solidary relationships between individuals in small groups. It is difficult to imagine that such affective ties could form the basis of people's private acceptance of their group membership in a larger group, for which interactive bases of group belongingness are minimal. On experimental grounds as well,  1. THE RESEARCH QUESTION / 3 the  interdependence  hypothesis  of  group  formation  has  been  seriously  challenged.  Research results have appeared which allow one to conclude that what is sufficient and necessary for group formation is not interdependence, interaction and interpersonal attraction but self-perceived group membership (Turner, 1981, 1982).  What have become  known as the  Social Categorization experiments  are the  most powerful demonstrations of the role of self-categorization in group formation and group behavior. Employing a procedure known as the Minimal Group Paradigm, Tajfel, Flament, Billig and Bundy (1971) conducted first experiments to address the question whether social categorization per se was sufficient  to cause intergroup discrimination.  The participants of the study (school boys) were randomly assigned to one of two distinct groups (eg. either X or Y), group membership was anonymous, there was no goal interdependence,  no interaction between them or any other basis for cohesive  relations, thus a minimal group. The experimental  task was to distribute monetary  rewards to anonymous ingroup and outgroup members, no personal benefit was possible. The results were that individuals systematically chose reward allocation strategies which discriminated  between  ingroup and  outgroup  members,  favoring  the  ingroup. They  showed group behavior by reacting to others in terms of their own and others group membership. The mere recognition of belonging to a group was sufficient  for the  emergence of group behavior.  Other  experiments  modified  the  basic  procedure.  The  effects  of similarity,  rewardingness and interpersonal attraction were investigated to see whether they would compete  with  the  self-categorization  effect  Furthermore, the  dependent  variable,  discrimination, was measured in various ways both at the behaviorial and judgmental  1. THE RESEARCH QUESTION / 4 level.  The overriding finding was  sufficient  that  self-categorization  per se  for group behavior (Billig & Tajfel, 1973; Doise, 1978;  is  necessary and  Lemyre & Smith,  1985; Tajfel, 1978; 1981; Turner, 1978; 1983; Turner, Hogg, Turner & Smith, 1984; Wetherell & Turner, 1984). On the basis of these findings Turner (1982) redefined the concept of social group in cognitive terms. As stated at the beginning of the chapter, we speak of a group when two or more individuals perceive themselves as members of the same social category.  GROUP BEHAVIOR  We may state the following: when people are aware of belonging to a group, they (ingroup) will tend to behave in a discriminatory fashion towards another group (outgroup), given that this group is defined and a relevant one against which to compare their own. Discrimination implies not only a clear differentiation between two groups  but,  furthermore,  a  pro-ingroup  bias.  In  the  Minimal  Group Paradigm  experiments, discrimination was expressed in the reward allocation strategy characterized by maximum ingroup profit plus maximum intergroup differentiation, a strategy termed "competitive  ingroup  favoritism". Turner  and Tajfel  (Tajfel,  1978;  1982;  Tajfel &  Turner, 1979; Turner, 1975, 1981, 1985) explain the discriminatory nature of intergroup behavior in the following way: We define  and evaluate  ourselves in terms of our  group memberships. In order to do so, we compare ourselves  to the outgroup on  relevant dimensions of intergroup comparison, seeking a "positive  distinctiveness"  for  ourselves that will maintain or enhance our positive identity as group members. Turner (1981) first called Identity  this  hypothetical  Theory and has now  explanation  extended  of  intergroup  and renamed it  discrimination Social  Self-Categorization Theory  1. THE RESEARCH QUESTION / 5 (1985). The theory rests on the two basic assumptions that: 1) our self-concept  is  composed of a personal and a social identity, and 2) we have a basic need for positive social identity (self-esteem) that we search for, via positive self-evaluation. The hypothesized  link  between  intergroup  discrimination  and  self-esteem  has  been  experimentally supported (Oakes & Turner, 1980; Turner & Spriggs, 1982; Lemyre & Smith, 1985).  People do not always think of themselves or interact with one another as group members. This is reflected in Turner's above assumption that the  self-concept  comprises a social as well as a personal identity. In behavioral terms, Tajfel and Turner (1979) speak of intergroup and- interpersonal behavior. This is based on Tajfel's (1978) conceptualization of interpersonal and intergroup behavior as the two extremes of a bipolar continuum. The 'interpersonal-intergroup continuum' (Turner & Tajfel, 1979; Tajfel, 1978) locates interpersonal and intergroup behavior at opposite ends of a bipolar continuum. To conceptualize interpersonal and intergroup behavior as located along a continuum clarifies  that  there  are varying degrees of  interpersonal  and intergroup  behavior, yet that at the same time, a definite distinction can be made between the two. Instances of a pure interpersonal and a pure intergroup interaction are more easily imagined in theory than observed in reality, especially in the case of the truly interpersonal  encounter  which  means  it  is  fully  determined  by  the  interpersonal  relationships of the persons involved. A likely example though, could be an interaction between two playmates (particularly if of same sex,  age group and family, which  would all reduce the possibility of their group memberships playing an important role). A  classic  Experimental  intergroup support  encounter for  Tajfel  would  be  a  and Turner's  sports  game  distinction  between  two  teams.  between interpersonal and  1. THE RESEARCH QUESTION / 6 intergroup behavior may be found in the following studies. For example, Turner (1978) and  Brown  and  Deschamps  (1980)  observed  that  when  the  salience  of  group  membership varied, social behavior varied. In a 'minimal group' situation involving reward allocations, subjects only discriminated between ingroup and outgroup members when  their  group  de-emphasized,  high  membership levels  of  was  emphasized,  self-favoritism  in  were  conditions  observed,  where  irrespective  it  was  of  the  affiliation of the other recipients. Other indications are that groups have consistently been observed  to be more competitive  than individuals under the  same conditions  (Wilson & Kayatani, 1968; Dustin & Davis, 1970; Doise & Sinclair, 1973; Doise & Weinberger, 1973). Intergroup/interpersonal differences on negotiation and bargaining (Stephenson, Yinon,  1979)  have been reported in research  1978), in studies on aggression (Yaffe &  and in research on gTOup decision  making (Lamm & Myers,  1978;  Wetherell & Turner, 1984). These research findings support what one is often able to observe and conclude from intergroup encounters in the real world: Individuals behave differently when they are part of a group from when they are not  Depersonalization and the Assimilation-Contrast Model  Tajfel continuum'  and  to  Turner  describe  the  (1979)  used  the  varying degrees  to  concept which  of  'interpersonal-intergroup  a person's  self  and other  perception will be influenced by 'personal identification' or 'social identification'. At one extreme  of  the  continuum, individuals' characteristics  and interpersonal  relationships  determine behaviors, at the other extreme, behaviors are fully determined by social group memberships. If a particular group membership is salient for a person, then 'social identification' is predominant for them and their cognitive processes may be  1. THE RESEARCH QUESTION / 7 described as 'intergroup thinking'. The person thinks of themselves and others in terms of  representative  characteristics  that  define  the  perception of self and others is stereotypic. output  of  (Turner,  a  functioning  1982;  self-perception  social  whole,  that  is, their  Expressed in terse form, the cognitive  identification  Brown & Turner,  group as ,a  is  essentially  a  stereotypic  1981). This form of stereotypic  perception  or depersonalized  Turner (1985) refers to as depersonalization. It should be noted that  depersonalization is understood as a state as well as a process. Turner speaks of a state  of  depersonalization  as  well  as  a  process  of  depersonalization.  The  interpersonal-intergroup continuum may be used to illuminate the difference. A state of depersonalization would be a behavioral instance located on the intergroup end of the continuum. The process of depersonalization describes the change from interpersonal to intergroup behavior along the bipolar continuum.  This perceptually category  kind  of  a  interchangeable  characteristics  stereotypic because  and not  perception they  are  their personal  implies  perceived  that in  individuals  terms  of  their  become shared  idiosyncratic nature. Indeed, everyday  examples come to mind where we automatically assign common category characteristics to  all  members  including ourselves,  ie.  "dog lovers  are better  people". Assigning  attributes of some category to all members of the category is a fundamental aspect of categorizing activity (Tajfel, 1959; this 'means  that  interchangeable  group  members  1972; Doise, 1978). Applied to intergroup relations/ perceive  outgroup  members  as  being  similar or  and also other ingroup members as being similar among themselves.  Moreover, it is found that people also overestimate the (perceived) differences between their group and the other group. Thus, when group membership is salient people will subjectively enhance intragroup similarities and exaggerate intergroup differences.  1. THE RESEARCH QUESTION / 8 This Finding stems from early work on the effects of categorization on physical stimuli (Secord, Bevan & Katz, 1956; Razran, 1950; Tajfel, 1959; Tajfel, Sheikh & Gardner, 1964). When a dichotomous classification was superimposed on a series of physical stimuli, perceived similarity of characteristics within a category and perceived differences - between categories were increased. Later on, similar effects were observed in the  realm of social perception. Studies of natural categories  such as English and  Welsh, Blacks and Whites, males and females, Catholics and Protestants, and French Canadians  and English  Canadians, have  found  that  people  overestimate intragroup  homogeneity and intergroup differences (Brewer, 1979; Doise, 1978; Tajfel, 1978). These findings  provided  the  basis  for  Tajfel's  (1969)  Assimilation-Contrast Model  of  stereotyping. He thought that the relatively automatic cognitive processes associated with categorization together with a motivation to enhance the relative value of the ingroup produced the accentuation of intragroup similarities and intergroup differences.  The Minimal Group Paradigm experiments (eg. Billig & Tajfel, 1973; Tajfel et al., 1971; Turner, 1978) introduced earlier played the key role in Tajfel's efforts to study  categorization  effects  and  group relations.  Inasmuch  as  they  represented a  convincing demonstration of the necessary and sufficient constituent for group formation, that  is  categorization  into  groups,  these  experiments  also  provided  insight  into  consequences of categorization. In the Minimal Group Paradigm experiments, what was labelled 'ingroup favoritism' was the most frequently pursued reward allocation strategy. This strategy distributes reward points in such a way that differences between groups are maximized and the ingroup members get more than the outgroup. This constitutes an example of the assimilation-contrast effect at the behavioral level. The Minimal Group Paradigm experiments not only demonstrated how group behavior unfolds and  1. THE RESEARCH QUESTION / 9 what its consequences are, but furthermore provided the impetus for Social Identity Theory  and  integrative  for  the  formulation  development  psychological  process  was  of  the  underlying  the  Assimilation-Contrast Model.  conceptualization  group  behavior.  of  A  depersonalization  Turner,  who  built  further as  on  the  Tajfel's  Assimilation-Contrast Model, argues that the major consequence of depersonalization is the  accentuation  of  perceived  similarity,  ingroup members and exaggeration  equivalence  of differences  and  interchangeability  among  between groups, consistent with a  positive ingroup evaluation. In a recent paper (1985), Turner develops as his central hypothesis that the depersonalization of self-perception is the basic process underlying those  group  phenomena  which  are  interpreted  as  having  an  assimilation-contrast  component He shows how a number of major group phenomena can be derived from this  hypothesis.  stereotyping,  According  group cohesion  to  his  and  theory,  such  ethnocentrism,  intergroup  cooperation  processes  like  social  and altrusim, emotional  contagion and empathy, collective behavior, shared norms and influence processes, are all characterized by a uniformity of thought feeling and action among members of the ingroup and simultaneous differentation from the outgroup.  The Person-Situation Model  An alternative formulation of the effects of depersonalization has recently been put forward by Smith (1985). Central to his Person-Situation Model is the question of whether  the  predominant  homogenization feature  of  the  of  relations  within  consequences of  a  group  depersonalization.  is  always  the  most  The Person-Situation  Model questions the generality of the assimilation-contrast process of depersonalization on the grounds that certain situations in which depersonalization were expected to be  1. THE RESEARCH QUESTION / 10 at its highest do not seem to be characterized by similarity or interchangeability of group members.  Smith mentions,  that members of sports teams, combat units and  musical orchestras, for example, become less, not more interchangeable at the point of intergroup competition. He argues this is related to the different roles or positions group members occupy in these groups.  According to the Person-Situation Model, depersonalization influences a person's thinking by drawing away attention  from the  causal priority of persons and their  dispositions, towards the causal priority of their predicaments or situations. Expressed in a different way, depersonalization changes one's understanding of persons in terms of their personal, idiosyncratic characteristics to an understanding of others in terms of their  shared  'egocentric'  situation and  'de'-personalized idiosyncratic,  as  defined  'ethnocentric'  by  personality  group  judgment  state, respectively.  internal  their  membership.  characterizing  'Egocentric' means characteristics  for  a  Smith  speaks of  personalized  that  attention  judgment  and  is  and  directed to  decision  making.  'Ethnocentric' means that group characteristics, like positions and roles within a group and how well they are fulfilled are important for judgment and decision making. He summarizes the primary implications of the Person-Situation Model in the following form:  "1)  among  members  of  groups  criteria of interpersonal judgment perception,  attribution,  attraction,  in  a relatively  and decision cooperation  interpersonal  state, the  making (including person and  influence)  will  be  egocentric, based on the degree to which others do or do not satisfy the norms of interpersonal complementarity and the requirements of self-esteem, and, 2) among members of groups in a relatively intergroup, depersonalized state, the criteria of judgment and decision making will be ethnocentric,  1. THE RESEARCH QUESTION / 11 based on the perceived importance of various positions and roles within the group  in  achieving  positive  group  status,  and  the  degree  to  which  individuals fulfill the obligations of these positions" (Smith, 1985, p.8).  It is predicted by Smith that depersonalization will lead to accentuation of intragroup similarities only when groups are unstructured as opposed to structured. In the case of unstructured groups in which members do not occupy different positions there exists no criterion relevant to group membership that could serve as a basis for differentiating among members in terms of their impact on the group. In structured groups, on the other hand, with different roles or positions, group members will be looked upon as different and not as similar or interchangeable because the roles will differ with respect to degree of importance to overall benefit of the group, and the extent to which individuals fulfill the requirements of a certain role. In short, while there  continues  group,  the  (perceived  to exist a uniformity of purpose among members of a structured  structural  differentiation  and actual)  among its  in  terms  of  members that  roles  introduces  a  differentiation  will inhibit a general increase in  perceived similarity among members of the same group.  The Person-Situation Model, first presented in the specific context of research on the effects of depersonalization on leadership processes, generates predictions about the effects of depersonalization on intergroup relations involving any kind of group. Building represents  on  the  assimilation-contrast  a generic  findings  explanation of the  in  unstructured  groups,  this  consequences of depersonalization.  model  Like the  Assimilation-Contrast Model, it understands depersonalization as the fundamental process underlying  intergroup  behavior.  However,  it  departs  from  the  assumption  of  the  1. THE RESEARCH QUESTION / 12 Assimilation-Contrast Model intergroup differences Person-Situation  that an increase in perceived intragroup similarities and  is the major consequence  Model,  the  assimilation-contrast  of depersonalization. According to the phenomenon  emerges in intergroup  relations when no differential positions are held within a group that would provide a basis for differential thought, feeling or action on the part of group members, with respect to the common goal of group benefit assimilation-contrast  finding is an artefact  In this sense, Smith argues that the  of the  kind of groups that have been  typically convened in experimental studies of group behavior. Turner and Smith both agree  that  in  a  depersonalized  or intergroup state, interaction between people  is  determined by their group memberships. They differ in what they understand to be the result of depersonalization: Intragroup similarity and simultaneous intergroup contrast (Turner) or ethnocentric judgment (Smith)?  PREDICTIONS BASED ON T H E MODELS  This question is central to the present study. Predictions based on the two models  are tested by comparing the  structured  groups.  Before  turning to  depersonalized interactions  of unstructured and  the  predictions, Turner's  present  design  and  assumptions concerning the Assimilation-Contrast Model require a closer look because two sets of predictions regarding the consequences of depersonalization in structured groups may be derived from them. The first set of predictions is straightforward, the second one requires more inference.  (1) Given the assumption of the Assimilation-Contrast Model that the major consequence  of  depersonalization  is  intragroup assimilation  and  intergroup  contrast  1. THE RESEARCH QUESTION / 13 combined with the fact that Turner has never addressed the issue of group structure, one may directly apply his model to group interactions involving structured groups expecting  'assimiliation-contrast' as the major outcome.  This has been proposed by  Smith (1985) in a recent paper.  (2) While Turner does not acknowledge the specific, though frequently observed case  of  structured  groups  in  his  writings  or  in  the  formulation  of  the  Assimilation-Contrast Model, it is open to interpretation whether he has simply not addressed yet this more complex form of group, or, whether he means to suggest that group structure does not affect  the  otherwise  observed  intragroup assimilation and  intergroup contrast To me, some of his hypotheses (Turner, 1985) imply at least as much as a sensitivity to the fact that certain intragroup circumstances or processes might overshadow the clear manifestation of the assimilation-contrast feature of group behavior. For example, in his recent self-categorization consequences nature  of  paper, Turner expresses the  of depersonalization in the following way: "It follows directly from the depersonalization  that  the  salience  of  an ingroup-outgroup categorization  increases the mentally perceived prototypicality of ingroup members on the stereotypical dimensions defining the ingroup category" (Turner, 1985 p.37). He goes on to address the phenomenon of mutual attraction and develops the hypothesis that the attractiveness of individual persons within a group depends upon their perceived prototypicality in comparison with other ingroup members. As evidence he adduces the data that ". people are more attracted to ingroup members perceived as better than their fellows in some way relevant to or important for the group, for example in terms of relative status  or  prestige,  valued  personality  traits,  their  relative  success,  correctness  or  competence in some group task or activity, and in the general degree to which they  1. THE RESEARCH QUESTION / 14 conform to group norms and role expectations (Lott & Lott, 1965)" (Turner, 1985, p.39). Thus, according to Turner it is conceivable that when persons in a group hold different positions, all in some way salient to the needs and goals of the group, their fulfillment of. these positions as well as the importance of these positions per se will be judged differentially by them and the other group members, at least with respect to intragroup processes, such as mutual attraction.  To  arrive  at  the  concrete  predictions  that  make  up  the  second  set  of  predictions based on the Assimilation-Contrast Model, I built on two major assumptions expressed in the above quotations. First, Turner assumes a positive relationship between the salience of an intergroup contrast and group members' perceived prototypicality of each other along relevant category dimensions. This stems directly from the empirically established finding that perceived  homogeneity  the  more  salient  the  in/out-group  within groups and differences  Brewer & Silver, 1978; Doise et al, 1972;  contrast,  the  greater  between groups (Brewer, 1979;  Turner & Spriggs, 1982,  Wilder, 1978).  Secondly, Turner assumes a relationship between a person's contribution to the group's goals and needs and the evaluation of that by fellow group members. Given different roles in a group, this will necessarily result in differential perception and judgment of ingroup members. These two predicted occurrences lead me to hypothesize specific  frame-of-reference  identification  determines  assimilation-contrast following  is  or the  the  most,  consequences  expressed  or  important of  comparison  depersonalization,  overshadowed.  For  for  positive  that  illustration,  that the  is  group whether  consider  the  real world instances of group behavior explained in terms of this Most  Important Comparison hypothesis. Imagine two sports teams at a game: (1) Immediately following the game, both teams are on the field, in close  1. THE RESEARCH QUESTION / 15 vicinity to one another, if not face to face interaction with one another. In this situation, the trainer of a team, or the team members will not make differential judgments about other team members. In front of the other group,  all ingroup members  are seen as  similar and better  than the  outgroup. In this example, the most important comparison is the outgroup. Crucial  for positive  group identification  is the  direct comparison to the  outgroup. Differential role fulfillments are secondary, the assimilation-contrast effect is expressed. (2) It is intermission. Each trainer talks to his team, typically showing such behaviors as giving different  instructions to the different players, making  different evaluative comments to the different players and exchanging one player for another. This is because the players occupy different positions, the different role expectations of which they fulfill The  trainer and other  differently)  and act  team  towards  members them  perceive  differently  to differing degrees.  them based  as on  different (and group  relevant  criteria, not on personal characteristics of the players. In this example, the most important comparison is with respect to the other group members. Crucial for positive group identification is how well a group member meets the  role  Differential  expectations role  in  comparison  fulfillments  are  of  to  the most  other  group  concern  so  members. that  the  assimilation-contrast effect is overshadowed.  Based Turner  on the  regarding  formulated:  In  above  thoughts,  the  consequences  an  intergroup  the of  second  set  of predictions  depersonalization  situation  involving  in  derived from  structured  structured  groups  groups,  is the  1. THE RESEARCH QUESTION / 16 assimilation-contrast effect will either be openly manifested or overshadowed depending on the most important comparison for positive group identification.  In this way, three specific sets of predictions have been generated concerning the behavior of structured groups in a depersonalized state. One set of predictions (Set I) is derived from the Person-Situation Model, and two sets (Set II and Set III) are derived from the Assimilation-Contrast Model. The three differ in their predictions regarding the assimilation-contrast phenomenon in structured groups. Set I predicts that the assimilation-contrast phenomenon will not be observed in structured groups as it would  be  in  unstructured  groups.  Set  II  predicts  that  the  assimilation-contrast  phenomenon will occur in structured groups just as it would in unstructured groups. Set  III  predicts  that  the  assimilation-contrast  phenomenon  will  be  exhibited  in  structured groups in those situations where the most important comparison for the positive self-identification for the group is the outgroup.  S U M M A R Y  Two current models about the consequences of depersonalization will be studied to learn which model might serve best as a representation of the consequences of this process.  Specifically,  we  will  study  the  Person-Situation  Model  and  the  Assimilation-Contrast Model in comparison to one another. The Person-Situation Model challenges the Assimilation-Contrast Model on the grounds that the assimilation-contrast phenomenon  might  not  be  the  major consequence  of  depersonalization  in groups  different from simple groups. The Person-Situation Model hypothesizes that in complex groups, structure introduces a cognitive differentiation among group members that should  1. THE RESEARCH QUESTION / 17 lead  to  differential  assimilation-contrast  judgment  phenomenon  and  decision  making  rather  than  to  the  observed for unstructured groups. Three specific sets  of predictions derived from the two models will be compared. The planned experiment will involve minimal groups, continuing in the tradition of previous research of the Social Categorization approach. If we can continue  to illuminate fundamental group  processes by studying minimal groups (minimal bases for group belongingness) we will have provided further support for the claim that perceptual and cognitive processes might  form the  initial necessary bases for group formation and subsequent group  behavior. Structured and unstructured groups will be studied in a depersonalized state, under conditions which vary the comparison relation to the outgroup. On the basis of the Most Important Comparison hypothesis, two situations of intergroup comparison will be created, modelled after the two aforementioned scenarios involving sports teams. The scenarios  involved situations  which were similar in that they both  situations  in which people  interacted  in a depersonalized  were intergroup  manner, in the  sense of  Turner's theory. The situations  were different with respect to their foci for positive  group identification,  an inwardly  one  had  directed  focus,  the  other  an outwardly  directed focus. Therefore we will look, both, at a situation in which comparison to the outgroup  will  be  the  most  important  comparison  and  at  a  situation  in  which  comparison to other ingroup members will be the most important comparison. The two distinctions  between  structured  and  unstructured  groups,  and  between  the  two  comparison levels will result in four different conditions of groups. In addition, we will include a fifth  condition, non-categorized  condition.  The  expression  observing  the  of  the  persons.  They will function as a control  assimilation-contrast  statements ingroup members  effect  will  will make about each  members of the outgroup, while engaged in intergroup interactions.  be  assessed by  other and about  2. METHOD  DESIGN  Structured and unstructured groups were studied in a depersonalized state, taking into account the explicitness of comparison to the outgroup. The two variables, Group Structure  (Structured  vs.  Unstructured)  Non-explicit comparison) formed a 2x2  and  Outgroup  factorial design  Comparison  (Explicit  to which a fifth  vs.  cell, the  Non-categorized Control condition was added (see Figure 1). The Outgroup Comparison variable could have also been referred to as 'salience of group membership' or 'degree of depersonalization', based on the well-established  findings that intergroup competition  is positively related to salience of group membership or degree of depersonalization. The depersonalized conditions Explicit and Non-explicit outgroup comparison represented states of intergroup behavior. The Non-categorized Control condition represented a state of  interpersonal  continuum,  the  behavior. Control  In terms  condition  of  may  Tajfel be  and Turner's interpersonal-intergroup  seen  as  located  on  the  end marking  personalized interactions. The Non-explicit and Explicit outgroup comparison states may be seen as located on the end marking depersonalized interactions, somewhat apart, reflecting the even stronger depersonalization of the latter state from the former. In the  Explicit conditions,  two groups (ingroup and outgroup) of the  same condition  participated in each other's presence in the same experimental session, in the same room. In the Non-explicit conditions, an outgroup was not actually present but only referred  to. In the Control  condition, the individuals present. in each experimental  session were not categorized into a group but participated independently from another.  18  2. METHOD / 19 SUBJECTS  Fifty-eight college aged males volunteered to participate in this study. The only criterion for participation was that a person would not take part in an experimental session in which they  personally knew  any of  the  others  present  The monetary  compensation for participation was $4.00.  PROCEDURE  The  experiment  consisted  of  a  game-like  procedure  in  which  subjects  participated as groups (experimental conditions) or as individuals (Control condition). The order in which conditions were tested was randomly determined. However, it was not always possible to adhere to this sequence because it was difficult to schedule participants for the double sessions for which 6 subjects were needed. Therefore, the double sessions took place whenever 6 subjects were available at the same time. The experimental  investigations  were  conducted  at  the  Psychology  Department  of  the  University of British Columbia, over a period of four weeks. They took place in a large laboratory room which could be partitioned by a curtain. A square table with a chair on each side was set up in the room. Placed on the table were index cards and pencils. For the double sessions, for which two groups were present at the same time, two  such tables  were  set  up at opposite  ends of the  room.  The general  procedure for all conditions took place as follows.  1. Introduction to the study: Upon arrival at the laboratory room subjects were seated  around the  table.  The experimenter  briefly introduced the  study informing  2. METHOD / 20 participants that they would be working on tasks involving brainstorming and would be asked to answer some questions about their own and others participation in these tasks.  .  2. Brainstorming task: Index cards and a sheet containing the instructions for the  brainstorming  task  (see  Appendix  A)  were  distributed  to  each  person. The  instructions for the task were read aloud by the experimenter. The brainstorming task took 15 minutes during which the experimenter left the room.  3. Rating of ingroup ideas: Each person selected their five best ideas from the ideas that they had come up with in the brainstorming session. Each labelled their ideas 1 through 5 to ensure that in the subsequent rating of ideas an order in which to rate could be maintained. Subjects also put an identifying code on each of their idea cards so that it would be clear which ideas were from the same person. On a rating sheet (see Appendix B), subjects rated their own five ideas on a scale from 1 to 7, T  marking the positive end of the scale. Then they passed their idea cards on  to the person on their left and proceeded to rate the idea cards of the person on their right In this way, idea cards were passed on once more so that everyone rated all five ideas of each person present  4. Questionnaire I (see Appendix C): Subjects were asked to answer questions regarding their group and its members. First they described themselves on a number of personal adjectives (QI.l). Adjective pairs consisting of a personality attribute and its antonym  were  presented  along  a  bipolar  7-point  rating  scale.  After  describing  themselves in terms of these adjective ratings, they described the other two members  2. METHOD / 21 in their group in this fashion (QI.2 to QI.3). Then they were asked a number of questions (again presented in terms of 7-point bipolar rating scales) which aimed, in various ways at a group .member's global satisfaction with the group and its' members, that  is  QI.4  addressed .commitment  to  group,  QI.5  to  QI.7  addressed  liking of  individual ingroup members and QI.8 to QI.11 addressed group performance.  5. Rating of outgroup ideas: A standardized set of outgroup ideas, that is three sets of five idea cards along with the appropriate rating sheet (see Appendix B) were distributed. In the Explicit comparison conditions, subjects were lead to believe that the other group present had produced the ideas they were given to rate as outgroup ideas. In the Non-explicit comparison conditions, subjects were told that another group who had participated in this study earlier had produced the ideas they were given to rate as outgroup ideas. In the Control condition, subjects were told that other individuals who had participated in the study earlier had produced the ideas they were given to rate. Then subjects were asked to proceed in the same way as described under 3.  6. Questionnaire II (see Appendix C): Subjects were asked to answer questions regarding the comparison between ingroup and outgroup. The response format for these questions was also in terms of 7-point bipolar rating scales. QII.1 addressed feelings regarding  the  ingroup, QII.2  addressed  feelings  regarding the  outgroup,  and QII.3  concerned the perceived difference between outgroup and ingroup ideas.  7.  Debriefing  (see  Appendix  D):  After  subjects  were  participation in the research project they were introduced to the  thanked  for  their  assimilation-contrast  phenomenon as it has been observed in the laboratory context and in the real world.  2. METHOD / 22 The objective of the present experiment, that is to study the manifestation assimilation-contrast  effect  in  structured  vs.  unstructured  groups,  based  of the on  the  Person-Situation and Assimilation-Contrast Models, was explained. Finally, subjects were given the opportunity to raise questions or discuss issues of interest to them.  Experimental sessions (steps 1 through 7) lasted 30-45 minutes, depending on condition.  CONDITIONS  The design consisted of four experimental conditions (A,B,C,D) and one Control condition (E) (see Figure 1). The variable, Outgroup Comparison, consisted of the two levels  Non-explicit  outgroup  comparison  and  Explicit  outgroup  comparison, which  marked progressively more depersonalized situations. The dichotomous variable, Group Structure, consisted of the two levels Structured group and Unstructured group. The Non-categorized Control condition could be seen as representing a third level of the variable, Outgroup Comparison, in the' sense that this variable marked progressively less depersonalized situations. For the variable, Group Structure, the Control condition did not represent a third level, but rather marked a different state altogether, that is 'no group'. Therefore, the fifth cell representing the Control condition could not be neatly integrated to yield a complete factorial design.  In all five conditions, three persons participated in one session. In the four experimental conditions (A,B,C,D), the three persons formed a group, whereas in the Control condition (E), the three subjects participated as independent individuals. In the Explicit comparison conditions (A,B) in which two sessions were run concurrently, group formation was determined by asking each of the six persons present to draw a slip of  2. METHOD / 23 paper.  In  the  Non-explicit  comparison conditions  (C,D),  subjects  were  told  they  comprised a group for the purposes of the study. Structured conditions (A,C) differed from the Unstructured conditions (B,D) with respect to the role of group leader. This role was intended as a cognitive differentiation rather than a functional one, to the degree  that  this  is possible.  For this  reason, somewhat  minimal  and open-ended  instructions were given for this role. After one person was selected for it, through drawing slips of paper, the leader was asked to sit on the side of the table which faced the experimenter. The group was simply informed that the role of leader would involve certain responsibilities later on with respect to representing the group and making decisions for the group, but no further definitions of instructions were given concerning the leader role.  The introduction to the study and the debriefing occurred exactly in the same way for all conditions. Procedural variations occurred at step 2 where formation into groups took place in the experimental conditions and assignment to the leader role in Structured conditions.  Step  4 and step 6 were  varied for the  Control condition:  Questionnaire I was shorter and Questionnaire II was skipped because the respective questions did not apply to the Control group. Aside from these procedural differences, the form in which participants were addressed and referred to in the instructions (i.e. as a group with leader, as a group with no leader, etc.) varied across conditions.  DEPENDENT MEASURES  The categories  assimilation-contrast of questions.  phenomenon  These were  was  assessed  a) global questions  using  three  different  regarding the ingroup and  2. METHOD / 24 outgroup (see Questionnaire I & II in Appendix C), b) descriptions of self and other ingroup members in form of personal attribute ratings (see the evaluation of ingroup and outgroup products (see  Questionnaire I), and c)  Rating Sheets for ingroup and  outgroup in Appendix B). The three question types were named 'global questionnaire items', 'attribute ratings' and 'product ratings', respectively.  The intention behind using these diversified measures was to allow a closer examination of the various forms in which the assimilation-contrast  phenomenon is  expressed. This was done for three reasons. The first one may be linked to Turner's conceptualization  of  the  assimilation-contrast  phenomenon.  According  to Turner,  the  phenomenon manifests itself in a person's thought, feeling and action. Therefore, the present  study  aimed to include such various forms of group related behaviors as  overall attitudes  towards the group and its individual  members, evaluations  of one  another in terms of personal attributes and evaluations of each others' task products. Another reason for the use of different types of questions concerned the role of the variable Structure. Since the intragroup differentiating  effect of structure was under  investigation,  individuals about  it  became  necessary  not  only  to  ask  their  views  regarding the group at large but also regarding individual group members. Thirdly, a further goal was to examine Turner's hypothesis regarding the parallel occurrence of ingroup  and outgroup  stereotyping  in  depersonalized  states.  In his  theory, Turner  assumes that an expression of assimilation is correlated with an equal expression of contrast  Thus, in order to investigate whether the assimilation component and the  contrast component are expressed to the same extent it seemed beneficial to illuminate both processes, those within the group and those directed towards the outgroup, using various types of questions.  2. M E T H O D  / 25  PREDICTIONS  Three sets of predictions regarding the expression of assimilation-contrast were derived from the two models under study. Prediction Set I was derived from the formulation of the Person-Situation Model, Set II and Set III were derived from the Assimilation-Contrast  Model.  The  predictions  stated  how  strongly  the  assimilation-contrast effect was expected to be observed in the various experimental conditions relative to one another. It should be noted that the experimental predictions were stated in terms of relative amounts of assimilation-contrast behavior expressed. When the  models,  in particular the  Person-Situation Model,  were  discussed  in a  theoretical context earlier on, the assimilation-contrast effect was frequently mentioned as either observable or not Turner assumed the assimilation-contrast effect to be the major consequence of depersonalization. In response to this assumption Smith predicted that  an  assimilation-contrast  effect  might  not  be  the  major  consequence  of  depersonalization in structured gToups, or, in other words, that the assimilation-contrast effect  might not be  consequences  of  observed in structured groups. Despite  depersonalization  in  terms  of  a formulation of the  presence  or  absence  of  assimilation-contrast, it is clear on the basis of Depersonalization theory that neither Turner  nor Smith assumed  phenomenon. consequences  Based  on  that  the  the  assimilation-contrast  understanding  that  effect  is  an all-or-none  depersonalization  and  also  its  are a matter of degree, the present experiment tested predictions about  relative amounts of assimilation-contrast in comparison to one another across varying experimental  conditions.  assimilation-contrast  Therefore,  behavior expected  experimental are not  to  predictions be  understood  on as  amounts  of  absolute. For  example, the statement that no assimilation-contrast effect was expected for the Control  2. METHOD / 26 condition does not mean an absolute absence but rather reflects the assumption that the the assimilation-contrast behavior in the Control condition may be called 'none', and relative to that the behavior in the other experimental conditions may be called 'more'.  Common predictions for Set I, Set II and Set III: On the grounds that both models understand depersonalization to be a matter of degree and assume, increases  with  salience  of  group  membership,  all  three  sets  predicted  that it that  the  assimilation-contrast effect would be stronger in the Explicit comparison state than in the  Non-explicit comparison state, in Unstructured groups. For the  non-categorized  condition, the participants of which would presumably be in a personalized state, no assimilation-contrast effect was expected to be observed.  Set I: According to the Person-Situation Model, depersonalization in structured groups should result in less, not more, interchangeability at the point of intergroup competititon.  Therefore,  little  or  no  assimilation-contrast  effect  was  expected  for  Structured groups. Given the assumption that assimilation-contrast effects will vary with respect to the degree of explicitness of outgroup comparison, the effect was predicted to  be  a  minimum  for  the  Explicit  outgroup  comparison  state.  Thus,  if  assimilation-contrast would be observed at all, under Non-explicit comparison conditions, then  even  less  assimilation-contrast  would  be  expected  for  Explicit  comparison  conditions. Following from that, it was predicted that the Structured conditions would demonstrate less assimilation-contrast behavior than the Unstructured groups.  Set II: These predictions represented a direct application of the assumptions of  2. METHOD / 27 the Assimilation-Contrast Model to the case of Structured groups. Thus, for Structured as well as, for Unstructured groups, the assimilation-contrast phenomenon was expected to occur under both outgroup comparison levels, and it was predicted to be more pronounced  under  Explicit  outgroup  comparison  than  under  Non-explicit  outgroup  comparison.  Set III: These predictions were also based on the Assimilation-Contrast Model. It was hypothesized that Group Structure and Outgroup Comparison would interact in such a way that the differing roles of group members in Structured groups would lead to differential judgment under Non-explicit comparison to outgroup but much less so under Explicit comparison to outgroup. Therefore, the assimilation-contrast effect was expected, to be weaker in Structured groups with Non-explicit comparison to outgroup than in Structured groups with Explicit comparison to outgroup. Furthermore, Structured groups under Non-explicit outgroup comparison were expected to show less than the Unstructured groups under comparable outgroup comparison conditions, and therefore also much less than the Unstructured groups under Explicit outgroup comparison.  SCALES, SCORING AND ANALYSES  All responses were given in terms of ratings using 7-point Likert-type scales. Usually, T  marked the positive end of the scale and '7' the negative end, with the  exception of a few personal attribute ratings for which the polarity was reversed to counteract the possibility of a response  set (see  Appendices B & C). In the data  analyses all items were recoded in the positive direction, that is the larger a group mean the more affirmative a response.  2. METHOD / 28 One-tailed t-tests were performed to compare all experimental groups to the control group on the dependent measures to check whether the categorization procedure had  been successful in producing group behavior among the experimental groups. The  data on the dependent measures were analyzed using 1-way (by condition) and 2-way (by structure x outgroup comparison) ANOVA and MANOVA procedures.  3. RESULTS  The dependent variables of the study were subjects' responses to the global questionnaire  items,  their  ratings  on  the  personality  attributes  and . on  the group  products (brainstorming ideas). For the analyses the responses to those different types of questions were assigned descriptive variable labels. Some dependent variables directly represent  the  response  to  a  single  questionnaire  item,  other  dependent  variables  represent a composite score from more than one response.  Dependent variables of the question category of global questionnaire items were (indicated in brackets is how they  relate  to the items of Questionnaire I & II):  Commitment to Group (Q4), Like Group (Q5), Like Other Members ((Q6 + Q7)/2), Group  Performance  ((Q10 + Qll)/2), (Q12-Q13), Outgroup  (Q8),  Own  Performance  (Q9),  Other  Member  Performance  Ingroup Favoritism (Q12), Outgroup Favoritism (Q13), Inoutfavoritism  Outgroup Contrast Evaluation  (Q14),  (Q13 + Q14),  Ingroup  Liking  Evaluation  Difference  (Q4 + Q5 + Q6 + Q7 + Q12),  (Q6-Q7)  and  Performance  Difference (abs(Q9-Q10)+abs(Q9-Qll)+abs(QlO-Qll)).  Dependent variables of the question category of personal attribute ratings were (indicated in brackets is how they relate to the items of Questionnaire I): Liking Ingroup (summary score across 6 personal attribute ratings for entire ingroup), Liking Other Ingroup Members (summary score across 6 personal attribute ratings for other ingroup members), Liking Self (summary score across 6 personal attribute ratings for self), Personality Attribute Difference Index 1 (absolute difference other  two  ingroup members  across  the  7 personal  29  attributes),  between ratings of Personality Attribute  3. RESULTS / 30 Difference Index 2 (sum of absolute differences between the two pairs of ratings of self and one other ingroup member across the 7 personal attributes) and Personality Attribute Difference Index 3 (sum of absolute differences between ratings of the three possible pairs of ingroup members across the 7 personal attributes).  Dependent variables of the question category of product ratings were (indicated in brackets is how they relate to the items of the Rating Sheets): Evaluation Ingroup Ideas (summed score for rating of ideas of entire ingroup), Evaluation Other Ingroup Member Ideas (summed score for rating of ideas of the two other ingroup members), Evaluation Self Ideas (summed score for rating of ideas of self), Evaluation Outgroup Ideas (summed score for rating of ideas of entire outgroup), Variance Ingroup Ideas (variance across idea ratings of entire ingroup), Variance Other Ingroup Members' Ideas (variance across idea ratings of other two ingroup members) and Variance Outgroup Ideas (variance across idea ratings of entire outgroup).  The different types of questions, the global questionnaire items, the personality attribute ratings and the product ratings, were looked at in two different ways. In the conceptualization  of  the  assimilation-contrast  effect,  assimilation  is  characterized by  perceived similarity, homogeneity and interchangeability among ingroup members with a positive  ingroup  homogeneity outgroup  by  bias.  Intergroup  contrast  is  and interchangeabilty and negative members  assimilation-contrast  of  the  phenomenon,  ingroup. the  characterized  by  perceived similarity,  evaluation of the  Therefore,  stereotyped  these  two  perception  members of the  expressions and  the  of  the  evaluative  component, were translated into two ways of assessing the assimilation-contrast effect One way was to look at the similarity of views expressed, the other was to look at  3. RESULTS / 31 the liking and evaluation expressed among the members of a group. These two ways of looking at responses were named Similarity indices and Liking-Evaluation indices, respectively. Before presenting the findings in this light, I will draw attention to the comparison  between  the  subjects  who  participated  as  a  group  and  those  who  participated as individuals. This comparison serves as a manipulation check for the success of the categorization procedure.  MANIPULATION CHECK - CATEGORIZATION PROCEDURE  All  experimental conditions were compared to the Control  condition to test  whether categorization into groups had the intended effect that groups would behave differently  than non-categorized individuals in the ways that have been alluded to.  Table 1 displays the means and t-values  contrasting experimental conditions to the  Control condition on dependent variables. Any ratings of self were not included in these  comparisons  between  all  experimental  conditions  with  the  Control  condition  because self-statements were not expected to differ significantly between categorized and non-categorized individuals. Furthermore, only those of the global questionnaire items appear in Table 1 which were answered by the Control group. The t-tests revealed that the categorization manipulation had the predicted effect for all Liking-Evaluation indices and some of the Similarity indices. Table 1 shows that the means for the experimental conditions on any measures of ingroup liking were significantly higher (p .005 and p .01) than for the Control condition. Interestingly, the mean for Evaluation Outgroup Ideas, was also significantly higher (p .01) in the experimental groups than in the Control group, a finding which is given special attention later. The Similarity indices  for  the  most  part  also  showed  that  the  experimental  conditions  were  3. RESULTS / 32 significantly (p .01 and p .05) less varied in their responses regarding the ingroup and outgroup than were participants in the Control condition. Exceptions were Personality Attribute Difference Indeces 1, 2 and 3 which showed insignificant differences between all  experimental conditions and the Control condition, similarly the measure, Liking  Difference, which only approached significance  (p .10)  in differentiating  between all  experimental groups and the Control group.  Following  below  assimilation-contrast,  are  the  findings  grouped under the  for  the  categories of  dependent  measures  Liking-Evaluation  of  indices and  Similarity indices, in that order. It needs to be mentioned that the present findings are statistically not very strong overall, yet they show a consistent pattern of distribution of means across conditions. Because of its consistency this pattern deserves attention even though it often consists of statistically marginal results.  LIKING-EVALUATION INDICES  Global questionnaire items  The and  group means on the global questionnaire items, which addressed ingroup  outgroup  condition  had  liking are the  displayed in  lowest  means  of  Table 2. all  As expected,  conditions  on  those  the  Non-categorized  items  relevant  to  Non-categorized individuals. This means that these persons expressed the least positive evaluations  towards  the  others  present  in  their  experimental  session. Surprisingly,  Non-explicit comparison, Structured groups consistently showed the highest mean of all groups  on  measures  of  ingroup  liking.  Under increased  Explicitness  of Outgroup  3. RESULTS / 33 Comparison these means were lower again for Structured groups. The opposite occurred in  Unstructured  groups.  Under  Explicit outgroup  comparison,  Unstructured  groups  showed higher mean scores on questions of ingroup liking than under Non-explicit outgroup  comparison.  Overall,  Unstructured groups  in  the  Non-explicit  comparison  condition had noticeably lower mean scores than Non-explicit, Structured groups. On measures of  outgroup  liking and  differentiation,  such  as  Outgroup Favoritism and  Outgroup Contrast, both Structured and Unstructured groups were less favorable towards the outgroup  under a higher level  of Explicitness of Outgroup Comparison. Again,  Structured groups were more extreme, that is their mean score on Outgroup Favoritism was smaller and on Outgroup Contrast higher than the mean of the Unstructured groups of comparable level of Outgroup Comparison.  First,  univariate  ANOVAs  were performed on  the  responses to  the global  questionnaire items (see Table 2). Two-way ANOVAs testing the effects of Outgroup Comparison and Group Structure and the significant  interaction between the two produced a  main effect for Outgroup Comparison on the item Outgroup Contrast (p  .05), a significant main effect for Group Structure on Outgroup Contrast (p .025) and marginal main effects for Group Structure on Commitment to Group (p .075), Ingroup Favoritism (p. 075)  and Outgroup Contrast (p .075), and and significant interaction  effects on Like Group (p .05), Like Other Members (p .05) and Ingroup Favoritism (p .025).  Before conducting post-hoc comparisons to identify which cells differed due to the  experimental  coefficients  were  variables, calculated  Outgroup among  Comparison and the  above  Group  mentioned  Structure, correlation  dependent  measures  to  3. RESULTS / 34 summarize highly related items under one score for further analyses. To accomplish this,  the  less powerful  correlational procedure  seemed  more justifiable  than factor  analysis, for which the present sample! size would have been tod small. As can be seen in Table 2, the interaction pattern. observed for the means of Structured and Unstructured statistical  groups  significance  under low  and high  levels of  for some items in the  correlated variables into one  score  allowed  Outgroup Comparison reached  two-way ANOVAs.  for the possibility  Combining highly  of strengthening  and  clarifying this pattern. The correlational analyses yielded two sets of highly correlated variables (see  Table 3). One of them consisted  of questions more or less directly  concerned with liking of the ingroup, that is Commitment to Group, Like Group, Like Other Members and Ingroup Favoritism. Correlations among these items were significant beyond  the  .0001  level.  They were added  to  form the  Evaluation for subsequent analyses. The other set consisted  combined of the  score  Ingroup  items Outgroup  Favoritism and Outgroup Contrast which were correlated beyond the .0001  level of  significance. These items, which addressed liking for outgroup and differentiation from outgroup (products),  respectively,  were added to form the combined score Outgroup  Evaluation for subsequent analyses. Two of those variables not included under either Ingroup Evaluation or Outgroup Evaluation were poorly correlated with most of the other questionnaire items. Own  Performance.  Specifically, those two items were Group Performance and  Because  content-wise  Performance, this latter item, although  they  were  related  to  Other  Members  it correlated highly with the other Ingroup  Evaluation items, was not included in the Ingroup Evaluation score. Even though these three performance-related items mostly were significantly correlated among one another, they were not combined to form a third summary score because no significant or marginal main effects or interaction effect were observed for any of them in the  3. RESULTS / 35 two-way ANOVA procedure.  Univariate,  two-way  ANOVAs  on  interaction effect (F(l,40) = 3.973, p = .053) plotted  in Figure 2.  Post-hoc  the  combined  scores  yielded  a marginal  on Ingroup Evaluation. This interaction is  paired contrasts  were conducted  via simple  effects  analysis using mean square within as the best estimator of the error variance. These comparisons  showed  that  the  Unstructured, Non-explicit  comparison  condition  was  significantly lower than the Structured, Non-explicit comparison condition (F(l,19) = 7.084, p = .015). The difference between the means for the Structured conditions under Explicit and  Non-explicit  outgroup  comparison  also  almost  reached  significance  with  F(l,19)=4.269 at p=.052. A glance at Figure 2 in which the standard errors of the means are indicated shows that interaction effect was due to the difference between Unstructured and Structured groups under Non-explicit levels of Outgroup Comparison and  to  the  difference  in means  between  the  Non-explicit  and  Explicit outgroup  comparison, Structured groups.  For  the combined score Outgroup Evaluation, its item Outgroup Contrast was  recoded to parallel Outgroup Evaluation such that the lower a mean the more it expressed  'contrasting' or dislike  produced two main effects. (F(l,43) = 2.891,  p=.096)  of  outgroup.  Two-way  ANOVA  on  this  measure  A just marginal main effect for Outgroup Comparison  and  a  significant  main  effect  for  Group  Structure  (F(l,43) = 6.298, p = .016). Figure 3 demonstrates that the Structured groups were much more negative towards the outgroup than the Unstructured groups, and that there was a tendency  in both types of groups to devalue the outgroup more under Explicit  comparison conditions.  3. RESULTS / 36 In/outgroup contrast was also looked at in terms of the difference between scores on Ingroup Favoritism and Outgroup Favoritism. The measure Inoutfavoritism was formed by subtracting the (lower)  score  for the outgroup from the score  for the  ingroup. When group means on Inoutfavoritism were compared by two-way ANOVA a significant main effect for Group Structure (F(l,43)=9.321, p = .O04)  and a significant  interaction effect between Outgroup Comparison and Group Structure (F(l,43) = 6.541, p=.014) emerged. The Structured groups had a significantly  higher difference  score  between liking of ingroup and outgroup than the Unstructured groups. As was the case on  the previous measure  of contrast, the  Structured groups expressed  the stronger  contrasting behavior. The interaction between Outgroup Comparison and Group Structure on  Inoutfavoritism is plotted in Figure 4.  Post-hoc  comparisons via simple effects  analysis showed the Unstructured, Non-explicit comparison condition to be significantly different  from all other three conditions. In the Non-explicit comparison condition,  Unstructured groups were much weaker in their outgroup dislike than the Structured groups. The F-value for difference in means for the two groups was F(l,21) = 16.823, p=.001.  The  difference  in  outgroup  contrast  between  Non-explicit  and  comparison conditions exhibited by the Unstructured groups was statistically with F(l,21) = 11.233 at p = .0003, paralleling the  Explicit significant  earlier observation for Unstructured  groups on Outgroup Contrast, For Structured groups, the difference on Inoutfavoritism between Non-explicit and Explicit comparison groups was insignificant  Personal attribute ratings  With  respect  to rating ingroup members on the personal attributes, similar  observations were made for the means of Structured and Unstructured groups as were  3. RESULTS / 37 for the global questionnaire items. In the Structured groups, ingroup member liking was lower under Explicit comparison than under Non-explicit, and in the Unstructured groups the reverse was observed. To be noted as different from findings regarding the global questionnaire items is that Unstructured groups were above the Structured groups with their means. Table 4 lists the group means and two-way ANOVA results for the three summary scores for the personal attribute ratings. As expected, the Control group was the least positive of all groups in evaluating other ingroup members as well as self. As already mentioned,  noteworthy  is the  familiar pattern of lower means in  Structured groups and higher means in Unstructured groups under Explicit as compared to Non-explicit outgroup comparison, although the pattern is too weak here to have reached statistical significance.  Product ratings  The means for product ratings appear in Table 5. Regarding the evaluation of the ingroup the Structured groups exhibited yet again a new pattern to earlier, while the Unstructured groups behaved like on the earlier reported ingroup measures. The ratings of outgroup products are in line with earlier observations on outgroup measures.  Two-way ANOVA results were significant for the outgroup ratings but not the ingroup ratings. Nevertheless, conditions is of interest.  the pattern described by the changes in means across  With respect to evaluation of ingroup products, Structured  groups behaved now very similarly to the Unstructured groups, that is with Explicit comparison to outgroup, the  evaluation  of ingroup products was  higher than with  Non-explicit comparison. The lowest ingroup ratings were observed for the Control  3. RESULTS / 38 condition.  On  the ratings for outgroup products, Structured groups showed much lower  means than Unstructured groups in the Non-explicit comparison condition. Two-way ANOVA results yielded a significant main effect for Group Structure (see The  Table 5).  means for both types of groups were lower under Explicit comparison conditions  than they were under Non-explicit comparison conditions, however, this difference was not significant Interestingly, the Control condition had the lowest mean for evaluation of outgroup products.  SIMILARITY INDICES  Global questionnaire items  The  two  Liking Difference  Similarity indices and Performance  constructed  based  Difference,  the  on the  global  questions were  former representing  a  difference  score across liking of other ingroup members and the latter a difference score across the contributions to group performance of all three group members.  On Liking Difference, univariate ANOVA produced a marginal main effect for Group  Structure (F(l,41)=3.599, p = .065). The Structured groups  differentiated  more  than the Unstructured groups between the other two ingroup members on liking, under both levels of Outgroup Comparison. Also, in both types of groups, difference scores were smaller  under high. comparison  than  under low  comparison conditions,  more  drastically so in the Unstructured groups than in the Structured groups. The Control  3. RESULTS / 39 condition had the highest difference score which means they varied the most in their liking of the other two persons present  Two-way  ANOVAs  performed on  Performance Difference  did not produce  significant effects. On this measure, the Control condition reached the highest mean score reflecting most differentiation between ratings of the other individuals present  Personal attribute ratings  Three types of indices were computed here to reflect how similarly all ingroup members were perceived or rated on the personal attributes. One kind  of index  involves the absolute difference in rating the other two ingroup members, a second one involves absolute difference between the pairs of self and one other and self and the third member, and the third type of index is a summation of the first two.  For each of the three indices, two-way MANOVAs were performed for each of the 7 items, individually. Then, two-way ANOVAs were performed on the summary scores, (i.e.  Personality Attribute Difference Index  1, Personality Attribute Difference  Index 2 and Personality Attribute Difference Index 3) of each index across the 7 items. The analyses did not produce significant results.  Product ratings  Similarity  with respect  to product ratings was assessed by looking at the  variances across ratings of the entire ingroup (Variance Ingroup Ideas), ratings of the  3. RESULTS / 40 other two members (Variance Other Ingroup Members Ideas) and ratings of outgroup products (Variance Outgroup Ideas). For these indices, no statistical differences between the means were found.  4. DISCUSSION  These" results give us a more complicated picture of the assimilation-contrast phenomenon than suggested by the Person-Situation and Assimilation-Contrast Models. As we shall see, neither of the models with their derived sets of prediction emerged unambiguously as the best explanation of the present findings. Both models turn out to be useful, though, in explaining certain components of this perhaps not so monolithic phenomenon.  First of all, attention must be drawn to the effect that categorization had. The present  study  categorization  provided, into  once  groups  individuals who were  more,  can induce  categorized  a  demonstration  group  behavior.  into groups (i.e.  of  the  ease  with  which  Comparisons between those  the experimental conditions) and  those who were not (i.e. the Control condition) showed this clearly on most measures of  ingroup assimilation  and  outgroup  contrast  For example,  all  Liking-Evaluation  measures strongly demonstrated the positive ingroup and negative outgroup bias among those  people  in  the  experimental  conditions  compared  to  those  in  the  Control  condition. The Similarity measures also demonstrated this effect. Most although not all, of the Similarity measures reflected significantly more similarity in views among persons in the experimental conditions compared to those in the Control condition. Based on the observed difference between all experimental conditions and the Control condition and the hypothesized distinction between intergroup and interpersonal behavior (Tajfel & Turner, 1971)  it may be concluded that on the whole the experimental conditions  showed more depersonalized interactions than the Control condition.  41  4. DISCUSSION / 42 The results indicate that the Liking-Evaluation measures worked decisively better than the Similarity measures. While on the former measures the groups showed a consistent  pattern  of  differences  in  means,  which  also  often  reached  statistical  significance, on the latter measures they did not differ much. The following discussion focuses  on  the  results  assimilation-contrast  effect  for  Liking-Evaluation  measures  The Similarity measures,  as  specifically  indicators the  role  of  the  of these  measures and their relationship to the Liking-Evaluation measures, are discussed in a later section.  It will be recalled that the assimilation-contrast effect was assessed in three ways. Persons were asked about their overall impressions regarding the entire group and individual group members ('global questionnaire items'), they were asked to rate other ingroup members on a number of personal attributes ('personal attribute ratings'), and thirdly, they were asked to rate the products of the ingroup as well as those of the  outgroup  within-group  ('product processes,  ratings'). was  assessed  The by  assimilation-component, all  three  of  these  contrast-component, which refers to between-group processes,  which question  refers  to  types. The  was assessed via global  questionnaire items and product ratings. The findings for these different measures of intra- and intergroup processes shall now be interpreted in light of the models under study, the Person-Situation Model, and the Assimilation-Contrast Model. Briefly, the Person-Situation  Model  makes  predictions  about  the  depersonalized  interactions  of  structured groups. It questions the generalizability of the assimilation-contrast effect to such complex groups as structured groups. The Assimilation-Contrast Model, on the contrary, assumes the effect to be the major consequence of depersonalized interactions, for any type of group. In the beginning, specific sets of predictions were derived from  4. DISCUSSION / 43 the models. What these entailed will become evident again in the following discussion.  There are four major observations in the research results that seem to be the most noteworthy. To start with the most straightforward one, we will first consider the well-studied  Unstructured  groups.  As  expected,  the  Unstructured  groups,  under  Non-explicit and Explicit levels of outgroup comparison, exhibited lower and higher levels of ingroup assimilation, respectively. measures  of the  global questionnaire  This was observed  items,  consistently  the personal attribute  across all  ratings and the  ratings of the group products. Even though the differences in means between low and high comparison levels were not large enough pattern  occurred  observation  was  consistently made,  on  all  however,  to reach statistical  measures.  the  expression  For of  outgroup  significance,  contrast,  outgroup  contrast  this  the  same  in  both  Unstructured and Structured groups will be discussed later. All three sets of predictions had assumed this occurrence for Unstructured groups. It was expected that the more depersonalized an interaction designed  to  lead  to  and the increasing levels of Outgroup Comparison were  increasing  degrees  of  depersonalization  -  the  more  assimilation-contrast effect would be expressed. This expectation was based on Turner's theory of depersonalization, which in turn, rests on the experimental support obtained for the notion of the 'interpersonal-intergroup' continuum of behavioral interactions (see chapter 1).  Our next observation concerns the Structured groups under Non-explicit and Explicit outgroup comparison levels. As predicted by the Person-Situation Model, their pattern of responses was exactly reversed to that of the Unstructured groups. On the global questionnaire items, Structured groups indicated significantly less ingroup liking  4. DISCUSSION / 44 under Explicit outgroup comparison levels than they did under Non-explicit levels. This was also true for the personal attribute ratings, although here the difference was not statistically  significant  These findings for ingroup liking support the  predictions of  Smith's Person-Situation Model. He predicted that the assimilation-contrast effect would weaken if not disappear under increased depersonalization, that is under Explicit levels of outgroup comparison. It will be recalled that he introduced the terms 'egocentric' and  'ethnocentric'  judgment  to  characterize  personalized  and  depersonalized  states,  respectively. Based on this distinction he argued that group structure provided a group with criteria for their ethnocentric decision-making that would lead to differential liking and evaluation of other ingroup members rather than to Turner's ingroup assimilation. It should be outgroup  mentioned  comparison  assimilation-contrast  that this  condition  measure,  the  weak was  assimilation-contrast not  ratings of  observed  for  effect the  ingroup products. With  in the Explicit third  type  respect  of  to the  ratings of ingroup products the Structured groups behaved parallely to the Unstructured groups. Under conditions of Explicit comparison to outgroup, the members of Structured groups evaluated each others' products higher, thus expressing more assimilation than under Non-explicit comparison to outgroup. However, this difference was not statistically significant  Because this observation is not statistically significant we cannot conclude  that, in the Structured groups, liking each others' products played a different role than liking each other as persons. Also, on no other counts did the product ratings produce different  results  from  the  other  measures  of assimilation-contrast  Still,  in a later  section, the theoretical nature of the three types of measures and in what ways they might have differed, specifically in the Structured groups, will be discussed.  We will now turn to the most surprising result of the present study, namely,  4. DISCUSSION / 45 the  differences  between  assimilation-contrast  effect  the  Structured  expressed.  and  Unstructured  groups  in  Contrary to all predictions it was  amount  of  found that  members of Structured groups expressed significantly more assimilation-contrast behavior than those of the Unstructured groups. The global questionnaire proved  to  be  the  strongest measures  of assimilation-contrast  items once again  Here, the Structured  groups expressed significantly more ingroup assimilation than the Unstructured groups, at least in the Non-explicit comparison conditions. In the Explicit comparison conditions, this difference  disappeared. The groups expressed very similar degrees of assimilation,  due to the fact that here, Structured groups showed less assimilation and Unstructured groups more assimilation than in the Non-explicit comparison conditions. Regarding the reactions to the outgroup, the same unexpected finding was observed. Structured groups engaged in significantly stronger outgroup contrasting than Unstructured groups. On the other  measures, • the  Structured groups  did not show stronger  ingroup assimilation,  however, their tendencies to negatively contrast the outgroup were, once again, more pronounced than in the Unstructured groups. For example, the outgroup product ratings were much lower in the Structured groups than in the Unstructured groups.  How can this finding be explained? It seems so contrary to the expectations of the Person-Situation Model as well as the Assimilation-Contrast Model. The former predicted  relatively  phenomenon  less  or  hardly  in Structured versus  any  expression  of  Unstructured groups (Set  the I).  assimilation-contrast The latter predicted  either no difference at all between these two types of groups (Set II) or differences in line with the predictions of the Person-Situation Model but only under Non-explicit levels of outgroup comparison (Set III). In the present experiment, Structured groups were actually stronger in their expression of the assimilation-contrast phenomenon, than  4. DISCUSSION / 46 Unstructured groups. In the  following  I will  suggest that this might be true for  structured groups in general, under certain conditions. It seems likely that introducing roles  in a group naturally brings with it  an  enhancement  of  salience  of group  membership. Particularly, the role of group leader carries a lot of weight: 'leader' is a connotative expression; it seems inseparable from 'group'. In the present experiment, the Structured groups might have felt 'groupier' than the Unstructured groups, due to the assignment of leader role. In that case we could conclude that Structured groups were more depersonalized than Unstructured groups, throughout the experimental session, and that  this  accounted  for  the  stronger  Structured groups. However, the  present  assimilation-contrast findings  effect  demonstrate  observed  that  in  structure  the  is not  correlated with a positive ingroup bias at all times: under Explicit conditions, ingroup assimilation was lessened in Structured groups. Whereas structure seemed to provide the group with a positive group identification under Non-explicit conditions, as indicated by high ingroup assimilation, structure seemed to become disruptive and work against a positive  group identification  ingroup  assimilation.  structure  in  a  This  group:  under  Explicit  leads to structure  the can  conditions,  following provide  a  as  indicated  assumption group  about  with  a  by the  a lessened effect  stronger  of  group  identification, which, under certain conditions, will result in stronger ingroup favoritism than in a no structure condition. While at this point, this represents no more than an interesting hypothetical assumption that is worth investigating in the future, the present results call for a preliminary explanation for the observed effect of structure on the assimilation-contrast phenomenon.  We shall now come  to the  fourth and last major finding of the  present  research. It concerns the reactions of Structured and Unstructured groups towards the  4. DISCUSSION / 47 outgroup.  These  reactions  assimilation-contrast  effect  are  interpreted  as  the  contrast  For Structured groups the results  component  of  the  for contrasting behavior  turned out to be different from those regarding assimilation, behavior. It was found that Structured groups, like the Unstructured groups, negatively contrasted the outgroup more  in  the  conditions.  Explicit  comparison  conditions  Outgroup Evaluation which  was  than -in the a composite  Non-explicit comparison  score  based  on  questions  regarding the outgroup showed this difference to be statistically significant This finding is  in  line  with  the  predictions  of  the  Assimilation-Contrast  Model.  Increased  depersonalization, that is an increased level of outgroup comparison, was hypothesized to lead to more contrasting against the outgroup in both types of groups. Set II was based on the assumption that Structured and Unstructured groups would not differ at all  in their intergroup relations as a consequence  of increased levels of outgroup  comparison. The assumption underlying Set III was that  in an intergroup situation  involving Structured groups, the assimilation-contrast effect would be openly manifested or  overshadowed  depending  on the  most  important comparison for positive  group  identification. The so-called Most Important Comparison hypothesis was derived from two imagined kinds of situations involving intergroup comparison. In one situation, the most important comparison for the group was the outgroup (comparison between two sports teams in direct competitive interaction). It was hypothesized that in this case the assimilation-contrast  effect  would  be  expressed  within the  group would be secondary. In the  comparison  for the  (comparison  between  group was team  the  members  because  differential  other situation, the  comparison among individual among  role  each  other  fulfillments  most important  ingroup members  because  outgroup  not  immediately present). It was hypothesized that the assimilation-contrast effect would be subdued  because differential  role  fulfillments  would be of  most concern to group  4. DISCUSSION / 48 members.  It is this second set of predictions, Set III, that is of interest now for the interpreting the results regarding the outgroup contrast of the Structured groups. With the  Most Important Comparison hypothesis  in mind one  can  draw an interesting  parallel between the present findings regarding outgroup contrast as well as ingroup assimilation and the two situations of intergroup comparison described above. Let the outgroup contrast results represent instances of direct comparison to outgroup, paralleling the first situation of intergroup comparison just described. This is not hard to imagine, as all these measures of outgroup contrast involved direct evaluations of the outgroup in comparison to the ingroup. In this way the results may be found to support the Most  Important  expressed  Comparison  towards the  hypothesis  outgroup. Next,  because let  the  assimilation-contrast  behavior  was  earlier reported ingroup assimilation  results represent instances of comparison among ingroup members, paralleling the second situation of intergroup comparison just described. Again, this requires no difficult leaps, as instances of intragroup comparison are precisely what all those assimilation measures are. In this case too, the findings support the Most Important Comparison hypothesis because as reported earlier the assimilation effect for Structured groups was subdued.  Can hypothesis  this be  new,  reconciled  originally generated  post-hoc with  the  application  of  original set  by this hypothesis?  the of  Most  Important Comparison  predictions  (Set  III)  that  was  It was shown earlier that Set III was not  supported by the present results. According to the original prediction, the Explicit level of outgroup comparison should have produced a Most Important Comparison situation of  the first kind,  that is comparison to the  outgroup. Similarly, the experimental  4. DISCUSSION / 49 condition of Non-explicit level of outgroup comparison should have produced a Most Important Comparison situation of the second kind, that is comparison among ingroup members. However, this comparison  levels  did  was not found. In general, the not  turn  out  to  represent  Explicit and Non-explicit  strongly  different  levels  of  depersonalization. Instead, direct questions regarding the outgroup versus direct questions regarding  ingroup  members  appear  to  be  much  stronger  and  also  perhaps  less  ambiguous representations of the two situations of intergroup comparison generated by the  Most  Important  Comparison  hypothesis.  At  the  very  least,  this  post-hoc  interpretation of contrast versus assimilation results in terms of the Most Important Comparison  hypothesis  demonstrates  Important Comparison hypothesis  the  potential  of  this  hypothesis.  The Most  is interesting because it makes much sense as an  explanation of real world occurrences. It needs to be investigated  further in future  research.  To summarize the four major observations, 1.  Unstructured  groups  expressed  more  ingroup  assimilation  under  conditions  of  increased depersonalization than under decreased depersonalization. This confirms Set I, Set II, Set III. 2. Structured groups expressed more ingroup assimilation under conditions of decreased depersonalization than under increased depersonalization. This confirms Set I. 3. The assimilation-contrast effect was expressed more strongly in the Structured groups than in the Unstructured groups. This is contrary to all prediction sets. 4. Unstructured  and  Structured  groups  expressed  more  outgroup  contrast  under  4. DISCUSSION / 50 conditions of increased depersonalization than under decreased depersonalization. For the Structured groups, this was contrary to the expression of ingroup assimilation. This confirms Set III.  These are the most important aspects of the present findings. Furthermore, I would like to draw attention to three more issues that emerged from the present results.  One concerns  Liking-Evaluation  the  indices.  differential The second  success of  the  centers on the  Similarity indices role  of  the  versus  three  the  types of  measures used. And finally there is the issue of spontaneous self-categorization among persons in the Control condition. They will be discussed in that order.  SIMILARITY INDICES VERSUS LIKING-EVALUATION  INDICES  It may be stated that the Similarity indices were weak in comparison to the Liking-Evaluation indices. Only for one between significant  groups  be  reported.  For  Similarity index could significant  Liking  Difference  Other  Ingroup  differences  Members a  main effect for Structure was reported. In line with finding three from  above, this Similarity index showed Structured groups to be more pronounced in their ingroup assimilation. No other Similarity indices produced significant differences between groups. Also, for the most part no consistent pattern emerged in which means differed from one another. The t-test results comparing the means of all experimental groups with that of the Control group were also not as strong for Similarity indices as they were for Liking-Evaluation  indices.  With respect  to global questionnaire  items and  product ratings, Similarity indices were successful in differentiating between experimental and  Control groups, however, less strongly so than the Liking-Evaluation measures. The  4. DISCUSSION / 51 t-test  results  for  personal  attribute  ratings  showed  no  differences  between  the  experimental and Control groups in their similarity of views across ingroup members.  A possible explanation for the lack of success of the Similarity indices concerns the type of similarity indices used in this study. Retrospectively, it seems that the Similarity  measures  were  not  appropriate because  they  were  closely  tied  to  the  Liking-Evaluation measures. The Similarity indices were based on ratings which are evaluative measure, per se. Therefore, Similarity indices might have been no more than less  direct  and  thus  weaker  measures  of  liking  and  evaluation  than  the  Liking-Evaluation indices. In this sense, any assessment of the similarity of views expressed in some way other than in terms of an evaluative scale would have been a purer measure of similarity. Not until similarity of views among the members of a group is assessed in a way which is more independent from evaluation, can we draw any  useful  conclusions  about  the  role  of  similarity  in  the  assimilation-contrast  phenomenon.  THE THREE TYPES OF MEASURES USED  The present study purposely employed three different categories of questions to study  the  assimilation-contrast  phenomenon.  The  reasons  for  this  were manifold.  Because the assimilation-contrast phenomenon is understood to be as all encompassing as to include a person's thought,  feeling  and action, the present study aimed to  include various forms of group-related behaviors such as, evaluations of one another in terms of personal attributes and evaluations of each others' task products. Another reason why the use of different types of questions was necessary is related to the role  4. DISCUSSION / 52 of  structure. With  the  major focus  of  the  study  on structure-induced intragroup  differentiation, it became necessary not only to ask individuals about their perceptions and  feelings  regarding the  group  at  large  but  also  regarding individual ingroup  members. Another goal of the study was to symmetrically study within-group processes and between-group processes in order to examine Turner's hypothesis  regarding the  parallel occurrence of ingroup and outgroup stereotyping in depersonalized states. To examine whether the assimilation component and the contrast component are expressed to the same extent, it was important to investigate them both using these different types of questions. To the degree that a reasonable face validity could be maintained, the different questions used to study ingroup processes were mirrored with respect to study of the outgroup.  It needs to be acknowledged that some of the question types or the way in which they were used constitute 'new territory'. For example, assessing views towards individual ingroup members to this detail, in addition to assessing views towards the group at large is a different approach from many previous studies in the field. More often than not, earlier studies have targeted only group members perceptions about the group at large.  Also the product ratings constitute new territory in that they have not been studied  in  controlled  laboratory  settings  as  measures  of  the  assimilation-contrast  phenomenon between groups. In fact, the assimilation-contrast phenomenon has been most often studied in straightforward gain-involving situations in which expression of assimilation-contrast  behavior  was  measured  in  terms  of  reward  allocations.  In  comparison, some of the present measures such as product ratings, personality ratings  4. DISCUSSION / 53 and ratings of the performance contribution of individual others (part of the global questionnaire items), may be seen as more complex measures.  And, finally to symmetrically study ingroup assimilation - and outgroup contrast in a  detailed  manner is  also  different  from  previous  research  in  this  field.  When  examining the results, the category of global questionnaire items stands out compared to the other two in terms of statistical strength. The pattern of differences in group means reported for these questions more often reached statistical significance than the similar pattern for the product ratings and the personal attribute ratings.  It is  suggested here  that  (regarding individual members)  the  novel  character of  the  personality  attributes  and the product ratings compared to the well-tried  character of the global questions might account for why the assimilation-contrast effect was not expressed more strongly on the former measures. It is assuring, though, that the results for the product ratings and the personal attribute ratings mainly supported the  results  for  Structured  groups  assimilation  under  the in  global their  Explicit  questions.  On one  product  evaluations  levels  of  count of  depersonalization  there  the  were  ingroup  than  under  deviations: expressed  the more  Non-explicit, an  observation contrary to those made for global questions and personal attribute ratings. As mentioned earlier, when the second major finding was reported, one cannot draw any conclusions from this regarding the special role product ratings might have played in groups with structure. This would certainly constitute  an interesting question for  future research.  The  results  for these different  measures  of ingroup and outgroup processes  4. DISCUSSION / 54 allow us to state that the assimlation component might not be congruent with the contrast component, at least not in Structured groups. Both the global questionnaire items and the product ratings indicated that Structured groups expressed more outgroup contrast under high depersonalization than under low. This was exactly opposite to the way that Structured groups expressed ingroup assimilation as measured by way of the global  questions.  To arrive  at  a  concluding -statement  about  the  three  different  categories of measures, one may say that the well-tried stood the test again, and that the newly tried measures, at the very least, provided support for the more traditional measures of the assimilation-contrast phenomenon.  SPONTANEOUS SELF-CATEGORIZATION - CONTROL CONDITION  There is indication that persons in the Control condition engaged in what has been  referred to as 'spontaneous  self-categorization'  (Turner,  1985). Spontaneous or  emergent self-categorization means that group formation takes place without an external imposition  of  social  categorization.  Specifically,  one  result  in  the  present findings  suggests that persons in the Control condition, on occasion acted like a group. With respect to the outgroup product ratings the unexpected observation can be made that the control individuals contrasted against the outgroup much more strongly than they contrasted against other individuals present in their experimental session. They rated the products of the other individuals in their experimental session much more favorably than those of the 'outgroup' members. One might make the claim that this could be related to truly existing qualitative differences between the idea products generated by the others present and the (fake) outgroup idea products. However, when the outgroup product ratings of the Control condition are compared to 'those of the other conditions  4. DISCUSSION / 55 the peculiar observation can be made that the Control condition expressed the most negative outgroup product evaluations! It is rather unlikely that a perceived qualitative difference between the products of other Control individuals and those of the outgroup would have lead the Control respondents to give the outgroup products the worst evaluations of all conditions. At least part of the Control condition's differentiation between  'ingroup'  self-categorization  and  'outgroup'  products  must  be  due  to  effect It is quite conceivable that a spontaneous  a  spontaneous  self-categorization  could have occurred in the Control condition. As pointed out by Turner (1985), a factor like physical closeness of persons could already trigger self-categorization, under certain circumstances. Circumstances were probably favorable for this occurrence in the Control  condition of  the present  study in the  following way. At first, when the  ingroup assimilation measures were taken, the three individuals of the Control condition were guided by self-favoritism and no reason to act as a group. On these ingroup assimilation measures no strong evidence of a self-categorization  effect can be noted.  At the point of introduction of outgroup products, however, circumstances probably became favorable to spontaneous self-categorization because the products of three other persons 'who did this in an earlier session' suddenly provided a relevant comparison to the three persons of the Control condition. According to the positive  distinctiveness  principle (see chapter 1), self-favoritism turned into ingroup favoritism. This might well be the reason why the Control condition expressed negative outgroup contrast.  CONCLUSION  The findings differentially  supported certain  aspects of  the Person-Situation  Model and the Assimilation-Contrast Model. The results for the unstructured groups  4. DISCUSSION / 56 supported the predictions of the Assimilation-Contrast Model while the results for the structured groups partially supported the predictions of the Person-Situation Model and to some extent those of the Assimilation-Contrast Model. The present results were statistically not strong enough to allow conlusions about the adequacy of the models. In addition, a few  somewhat  attention  research. Among  in  future  difficult issues arose in this research which require them  were  the  issues  of  usefulness  of  the  Similarity indices versus the Liking-Evaluation indices, the differential strengths of the three types of measures of assimilation-contrast and the manipulation of the structure variable. As discussed earlier, these issues are not just unique to the present study but represent general challenges to future research of intergroup processes. issues  the  present  results  added  important insight  to  the  Despite these  understanding  of  the  depersonalization process. Two respects shall be mentioned in particular in which the present  findings  made  contributions  to  the  understanding  of  this  central  process  underlying intergroup interactions. The predictions of the Person-Situation Model were supported with respect to the occurrence of ingroup assimilation. This indicates that structure within a group seems to work against an expression of ingroup assimilation which threatens the generality of the assimilation-contrast effect This is not only a promising result for the  recent model but also  of  far reaching consequences  for  understanding intergroup processes. The other respect in which this research contributed concerns the  character of the  suggested that the  assimilation-contrast  phenomenon. The present results  assimilation component may not be  exactly  congruent with the  contrast component Overall, outgroup contrast was expressed more strongly than ingroup assimilation. For the Structured groups, the expression of the two components was actually in reversed directions under levels of increased depersonalization. Interestingly, elsewhere similar findings have been reported (Linville & Jones, 1980; Linville, 1982;  4. DISCUSSION / 57 Park & Rothbart, 1982) This group of researchers who do not necessarily share the theoretical assumptions of the Social Categorization approach predict greater perceived homogeneity of outgroup members and more extreme evaluative expressions towards the outgroup. The present  study, based  interpreted  extreme  the  more  on the  expression  assumptions of  outgroup  of  depersonalization theory  contrast  than  of  ingroup  assimilation in terms of the most important intergroup comparison available for positive group identification (Most Important Comparison hypothesis).  As a final conclusion I  wish to state that it might serve us better to think of the assimilation-contrast effect as a complex phenomenon that is sensitive not only to differences between inter- and intragroup processes but also to to the  degree of depersonalization  perhaps even to different modes of expression (thought vs. behavior).  prevalent, and  REFERENCES BILLIG, M.G., & TAJFEL, H. (1973). Social Categorization and Similarity in intergroup behavior. European Journal of Social Psychology, 3, 27-52.  BREWER,  M.B. (1979).  Ingroup  bias  in  the  minimal  intergroup  situation:  a  cognitive-motivational analysis. Psychological Bulletin, 86, 307-24. BREWER,  M.B., & SILVER,  M . (1978).  Ingroup  bias  as  a  function  of  task  characteristics. European Journal of Social Psychology, 8, 393-400. BROWN,  R.J, & DESCHAMPS,  J.C. (1980/81). Discrimination entre individus et entre  groups. Bulletin de Psychologie, 34, 185-95. DOISE, W. (1978). Groups and individuals, explanations in social psychology. Cambridge University Press. DOISE, W., CSEPELI, G., DANN, (1972).  An experimental  H.D., GOUGE, C , LARSEN, K., & OSTELL, A. investigation  into  the  formation  of  intergroup  representations. European Journal of Social Psychology, 2, 202-204. DOISE, W., & SINCLAIR, A. (1973). The categorization process in intergroup relations. European Journal of Social Psychology, 3, 145-57. DOISE, W., & WEINBERGER,  M. (1973). Representations masculines dans differentes  situations de rencontres mixtes. Bulletin de Psychologie, 26, 649-57. DUSTIN,  D.S., & DAVIS,  H.P. (1970).  Evaluative  bias  in group  and individual  competition. Journal of Social Psychology, 80, 103-8. LAMM,  H., & MYERS, behavior,  in  D.G. (1978). L.BERKOWITZ  Group-induced (ed.),  psychology. Vol. 2, New York.  58  polarisation  Advances  in  of attitudes and  experimental  social  References / 59 LEMYRE, L., & SMITH, P.M. (1985) Intergroup discrimination and selfesteem in the minimal group paradigm. Journal of Personality and Social Psychology, 49, 660-670. LINVILLE, P.W. (1982) The Complexity-Extremity effect and age-based stereotyping. Journal of Personality and Social Psychology, 42, 193-211. LINVILLE,  P.W. & JONES,  E E . (1980) Polarized appraisals of outgroup members.  Journal of Personality and Social Psychology, 38, 689-703. LOTT, A.J. & LOTT, B E . (1965) Group cohesiveness as interpersonal attraction: a review  of  relationships  with  antecedent  and  consequent  variables.  Psychological Bulletin, 64, 259-309. OAKES, P.J., & TURNER,  J.C. (1980) Social categorization and intergroup behavior:  Does minimal intergroup discrimination make social identity more positive? European Journal of Social Psychology, JO, 295-301. PARK B. & ROTHBART, M. (1982) Perception of outgroup homogeneity and levels of social categorization  : memory for the subordinate attributes of in-group  and out-group members. Journal of Personality and Social Psychology, 42, 1051-1068. RAZRAN, G. (1950). Ethnic dislikes and stereotypes: A laboratory study. Journal of Abnormal and Social Psychology, 45, 7-27. SECORD, P.F., BEVAN, W., & KATZ, B. (1956). The negro stereotype and perceptual accentuation. Journal of Abnormal and Social Psychology, 53, 78-83. SMITH,  P.M. (1985). Leadership and intergroup relations: Depersonalization and the prospect of security. Paper presented to the British Psychological Association in September 1985.  References / 60 STEPHENSON,  G.M. (1978).  groups,  in  Interparty  and  H.BRANDSTATTER,  interpersonal J.H.DAVIS,  exchange and  in  negotiation  H.SCHULER  (eds.),  Dynamics of group decisions. London: Sage.  TAJFEL;  H. (1959). Quantitative judgment  in  social  perception.  British Journal of  Psychology, 50, 16-29. TAJFEL, H. (1969). Cognitive aspects of prejudice. Journal of Social Issues, 25, 79-97. TAJFEL, H. (1972) Social categorization, English ms of La categorisation sociale. in S.MOSCOVICI (ed.) Introduction a la psychologie sociale. Vol. 1, chapter 8, pp. 272-302, Paris: Larousse. TAJFEL,  H. (1978). Interindividual behavior and intergroup behavior, in H.TAJFEL (ed.) Differentiation between social groups. London: Academic Press.  TAJFEL, H. (1981). Human groups and social categories. Cambridge: University Press. TAJFEL,  H. (1982).  Social  identity  and intergroup relations.  Cambridge: University  Press. TAJFEL,  H., FLAMENT, categorization  C,  and  BILLIG,  M.G., &  intergroup  behavior.  BUNDY,  R.F.  (1971).  Social  European Journal of Social  Psychology, 1, 149-77. TAJFEL, H., SHEIKH, A.A., & GARDNER, R.C. (1964). Content of stereotypes and the inference of similarity between members of stereotyped  groups. Acta  Psychologica, 22, 191-201. TAJFEL, H., & TURNER, J.C. (1979). An integrative theory of intergroup conflict, in W.G.AUSTIN  & S.WORCHEL (eds.) The social psychology of intergroup  relations. Monterey, Ca.: Brooks/Cole.  References / 61 TURNER,  J.C. (1975).  Social  comparison and social  identity:  Some  prospects  for  intergroup behavior. European Journal of Social Psychology, 5, 5-34.  TURNER, J.C. (1978). Social Categorization and social discrimination in the minimal group paradigm, in H.TAJFEL (ed.)  Differentiation hetween social groups:  Studies in the social psychology of intergroup relations. London: Academic Press.  TURNER, J.C. (1981) The experimental social psychology of intergroup behavior, in J.C.TURNER & H.GILES (eds.) Tmergroup hehavior. Oxford: Blackwell.  TURNER,  J.C. (1982). H.TAJFEL  Towards a (ed.)  Social  cognitive  redefinition  identity  and  of  intergroup  the  social  relations.  group, in Cambridge:  University Press. TURNER, J.C. (1983). Some comments on "The measurement of social orientations in the minimal group paradigm". European Journal of Social Psychology, 13, 351-367. TURNER, J.C. (1985). Social categorization and the self-concept: theory  of  group  behavior,  in  E.J.LAWLER  (ed.)  A social cognitive Advances  in  group  processes: Theory and research. Vol 2. Greenwich, Conn.: JAI Press Inc. TURNER,  J.C, HOGG, M.A., TURNER, defeat as  determinants  of  P.J., & SMITH, P.M. (1984). Failure and  group cohesiveness.  British Journal of Social  Psychology, 23, 97-111. TURNER, J.C, & SPRIGGS, D. (1982). Social categorization, intergroup behavior and self-esteem: A replication. Unpublished manuscript WETHERELL,  M., & TURNER,  J.C. (1984). Social identification,  opinion shifts. Unpublished manuscript  differentiation and  References / 62 WILDER, D.A. (1978). Reduction of intergroup discrimination through individuation of the outgroup. Journal of Personality and Social Psychology, 36, 1361-74.  WILSON, W., & KAYATANI, M. (1968). Intergroup attitudes and strategies in games between  opponents of  the  same  or  of  a  different  race,  Journal of  Personality and Social Psychology, 9, 24-30. YAFFE,  Y., & YINON, Y. (1979). Retaliatory aggression in individuals and groups. European Journal of Social Psychology, 9, 177-86.  TABLE CAPTIONS  Table I  Means and t-values from one-tailed  t-tests comparing  experimental  conditions to the Control condition on dependent variables.  Table II  Means and F-ratios for the effects of Comparison x Structure for the  ANOVAs performed on the responses to the global questionnaire items.  Table III  Correlations among questionnaire items (1.4 to 1.11, II.l to II.3).  Table IV  Means and F-ratios for the effects of Comparison x Structure for the ANOVAs performed on the personal attribute ratings.  Table V  Means and F-ratios for the effects of Comparison x Structure for the ANOVAs performed on the ingroup and outgroup product ratings.  63  Table I Means and t - v a l u e s from o n e - t a i l e d t - t e s t s comparing e x p e r i m e n t a l c o n d i t i o n s t o the C o n t r o l c o n d i t i o n on dependent v a r i a b l e s .  VARIABLE  CONDITION  MEAN  T-VALUE  DF  Liking-Evaluation Indices: Tngroup Evaluation  E C  45 11  5.09 4.41  2.83  54  Outgroup Evaluation  E C  45 11  5.22 4.55  2.54'  54  Liking Ingroup  E C  46 11  32.20 29.27  2.82'  55  L i k i n g Other Ingroup Members  E C  47 11  31.90 28.50  3.09  56  Evaluation Ingroup I d e a s  E C  47 11  85.43 78.18  2.56  56  E v a l u a t i o n Other I n g r o u p Members' Ideas  E C  47 11  28.14 25.27  2.64  Evaluation Outgroup Ideas  E C  47 11  75.04 66.36  2.34  64  XXX  56  56  Table I c o n t i n u e d  CONDITION  VARIABLE  MEAN  T-VALUE  DF  S i m i l a r i t y Indices: Liking Difference  E C  45 11  .58 1.00  -1.44  54  Performance Difference  E C  45 11  1.96 3.27  -1.93  54  Personality Attribute Difference Index 1  E C  47 11  6.02 6.82  -.65  56  Personality Attribute Difference Index 2  E C  46 11  12.20 14.27  -1.19  55  Personality Attribute Difference Index 3  E C  46 11  18.26 21.09  -1.09  55  Variance Ingroup Ideas  E C  47 11  1.14 1.66  -2.05  56  V a r i a n c e Other Ingroup Members' Ideas  E C  47 11  1.18 1.72  -1.87  56  Variance Outgroup Ideas  E C  47 11  1.78 2.88  -2.24  p < .005  XK  p  < .01  x  p <.05  65  xx  56  Table II Means and F - r a t l o s for the effects of Comparison x Structure for the ANOVA's performed on the responses to the global questionnaire items. VARIABLE  CONDITION Structured Unstructured by Expl. Non-Expl. Expl. Non-Expl.(Control) Comp.  F-VALUES by Struct.  Commitment to Group  4.82  5.45  4.50  4.55  1.073„  Like Group  5.09  5.55  5.25  4.70  0.035  Like Other Members  4.91  5.41  5.17  4.86  Group Performance  5.00  5.64  5.17  5.09  Own Performance  5.36  5.55  5.33  4.82  (5. 81)  0.506,  Other Member Performance  5.18  5.59  5.13  5.00  (4. 55)  Ingroup Favoritism  5.50  6.08  5.58  Outgroup Favoritism  3.83  3.92  Outgroup Contrast  5.25  4.50  p<.025  w  p<.05  *p <.075  (4. 41)  < t l )  U 4  „  by Comp. x Struct.  3.535*  0.836  1.733  4.254**  0.191  lMl)  0.433  3.802*"  0.561  l l t W  0.249  0.955  2.305  2.028  0.426«,,4., 2.320  1.634  5.18  0.233  5.455***  4.25  4.73  0.725 „,4w 3.534*  4.25  3.09  3-913^,  1|W  (1/tl  j 3.561"  6.244*"  0.369 0.181  Table III Correlations among questionnaire items  Q.4 Q.4 Commitment to Group Q.5 Like Group (Q.6+Q.7)/2  Q.5  (Q.6+Q.7) 2 Q.8  Q.9  (Q.10+Q.11) 2 Q.12  Q.13  Q.14  1.0000  551Q*** 1.0000 ,6026***  .8008*** 1.0000  Like Other Members Q.8 Group Performance  4539  .2156  .1779  Q.9 Own Performance  3203*  .4381**  .3034*  .5039*** 1.0000  (Q.10+Q.ll-)/2 Other Member' Performance  ,5815***  .7337***  .7269***  .3920**  .3419*  Q.12 Ingroup Favoritism  ,5177 *  .5783***  .4816***  .3870**  .3411*  Q.13 Outgroup Favori t Lsm  ,2775  .0663  .2369  .1202  -.0032  .2457  .2057  Q.14 Outgroup Contrast  -.1696  .0436  -.0469  -.2286  .1792  -.0119  -.0054  r  p<.0001  xx  **p<.001  *p<.01,  1.0000  1.0000  .4297** 1.0000  1.0000  -.6357***1.0000  N=45 (except for Commitment to Group, N=44)  Table IV Means and F-ratios for the effects of Comparison x Structure for the ANOVA's performed on the personal attribute ratings.  VARIABLE  CONDITION Structured Unstructured Expl. Non-Expl. Expl. Non-Expl.  (Control)  Liking Ingroup  31.44  32.11  32.94  32.30  (29.27)  n. s.  n. s.  n. s.  Liking Other Ingroup Members  31.21  31.96  32.50  31.95  (28.50)  n.s.  n. s.  n.s.  Liking Self  31.92  32.42  33.83  33.00  (30.82)  n.s.  n.s.  n.s.  by Comp.  F - VALUES by by Struct. Comp. x Struct.  Table V Means and F-ratios for the effects of Comparison x Structure for the ANOVA's performed on the ingroup and outgroup product  VARIABLE  ratings.  CONDITION Structured Unstructured Expl. Non-Expl. Expl. Non-Expl.  (Control)  Evaluation Ingroup Ideas  86.58  84.00  86.67  84.36  (78.18)  n.s.  n.s.  n. s.  Evaluation Other Ingroup Members' Ideas  28.42  27.71  28.50  27.91  (25.27)  n.s.  n.s.  n.s.  Evaluation Self Ideas  29.75  28.58  29.67  28.55  (27.64)  n.s.  n.s.  n.s.  Evaluation Outgroup Ideas  69.92  71.33  78.25  81.18  (66.36)  0.448  7.921  0.055  by Comp.  F - VALUES by by Struct. Comp. x Struct.  as  XX  p <-05  FIGURE CAPTIONS  Figure 1  Experimental Design  Figure 2  Comparison by Structure interaction on Ingroup Evaluation score  Figure 3  Comparison and Structure main effects on Outgroup Evaluation score  Figure 4  Comparison by Structure interaction on Inoutfavoritism score  70  Figure 1 Experimental  Design  Group S t r u c t u r e Struct.  Unstruct.  Explicit  A  B  NonExplicit  C  D  E (Control)  71  FIGURE 2. 30  -i  9 Structure  29 H  O No Structure LU OC  O O  CO  z o  28 H  27-  26  ^ LU  25^  24  non-explicit explicit COMPARISON  FIGURE 3. 12  H  •  11  Structure  O No Structure LU  O  CJ CO  z o  10  9H 8  LU 6  ~  1  1  non-explicit explicit COMPARISON  FIGURE 4.  non-explicit explicit COMPARISON  APPENDIX A - INSTRUCTIONS FOR BRAINSTORMING  75  TASK  Instructions  Imagine  With  the  the  decided life  t o c o n s t r u c t a time  time has  the It  t h a t would  capsule  best  g r a n i t e of will  Its  be  a n u c l e a r war, capsule  the  conserving  C a n a d i a n Government different  o b j e c t s of  show t o someone what c i v i l i z a t i o n  a r e most  Canadian kept  be  about  fill  on  has our  earth  naming one  around  object  When  i t i s your  time  into  the  the  at a time  turn, t e l l  will  at a temperature 1000  be  placed  objects.  deep w i t h i n  a standard  capsule  table  that w i l l  the  bedroom.  with on  o b j e c t s t h a t you earth  in a clockwise  to p l a c e  preserve  years.  r e v e a l i n g o f what c i v i l i z a t i o n going  organisms, only  Shield.  t h a t of  this  living  timecapsule  f o r approximately  i s to  turns  i n the  feel  i s a l l about. direction,  each  capsule.  y o u r g r o u p what o b j e c t you  want  to  put  one  of  the  capsule.  After  you  index  cards  turn,  that  have t o l d you  your  have  idea please  in front on  write  of you.  i s the  person  has  15  min  t o come up  l u c k now  and  l e t your  Your group  good  c o n t a i n any  t h a t the  s e a l e d and  will  task  Take  the  inside  size  Your  cannot  been d e c i d e d  objects  Ok,  danger of  like.  The It  task  following:  impending  here  was  for brainstorming  your  Then  i t down on i t i s the  next  person's  left.  with  as  many good  imagination  76  run  i d e a s as p o s s i b l e .  wild!  APPENDIX B - RATING SHEETS  77  Racing  Sheet  Now p l e a s e r a t e e a c h one o f y o u r own 5 b e s t by c i r c l i n g  ideas on a s c a l e  f r o m 1 t o 7,  a number:  1. i d e a : very  1 good  2  3  4  5  6  7 v e r y bad  2.  idea:  1  2  3  4  5  6  7  3.  idea:  1  2  3  4  5  6  7  4. i d e a :  1  2  3  4  5  6  7  5.  1  2  3  4  5  6  7  idea:  Now p a s s y o u r i d e a c a r d s o n t o g r o u p member Next, rate  t h e i d e a s o f g r o u p member  1.  idea:  2.  idea:  1  3.  idea:  1  4. i d e a :  1  5.  1  idea:  1 v e r y good  2  2  2  2  2  on y o u r you j u s t  left.  received.  3  4  5  6  7 very bad  3  4  5  6  7  3  4  5  6  7  3  4  5  6  7  3  4  5  6  7  78  Now pass these idea cards on to group member Next, rate the ideas of group member  U  i  d  e  a  :  2. idea:  very gold 1  2 2  you just received.  3  4  5  6  3  4  5  6  3 . idea: 4.  idea:  5.  idea:  on your l e f t .  79  ^7  ^ 7  Rating  Sheet  Now p l e a s e r a t e e a c h one o f t h e 5 b e s t  ideas that  t h e members o f t h e o t h e r  came up w i t h .  1.  idea: very  1 good  2  3  4  5  6  7 v  e  r  y  b  2.  idea:  1  2  3  4  5  6  7  3.  idea:  1  2  3  4  5  6  7  4. i d e a :  1  3  4  5  6  7  5.  1  3  4  5  6  7  idea:  Now p a s s t h e s e  2  2  idea cards  o n t o t h e g r o u p member on y o u r  Next r a t e t h e ideas you j u s t  1. i d e a :  2.  idea:  3.  idea:  4.  idea:  5. i d e a :  1 v e r y good  2  a  d  left,  received.  3  4  5  6  6  1  2  3  4  5  1  2  3  4  5  80  7 v e r y bad 7  group  Now pass these idea cards on to the group member on your l e f t . Next rate the ideas you just  1.  idea:  1 very good  received.  2  3  4  5  6  7 very  2. idea:  1  2  3  4  5  6  7  3. idea:  1  2  3  4  5  6  7  4. idea:  1  2  3  4  5  6  7  5. idea:  1  2  3  4  5  6  7  81  APPENDIX C - QUESTIONNAIRES I AND II  82  2.  Next, please rate member  1 *  f a i  2  of your group on these adjectives:  3  4  5  6  7 unfair  1 unfriendly  2  3  4  5  6  7 friendly  1 warm  2  3  4  5  6  7 cold  2  3  4  5  6  7 hard-working  1 self-assured  2  3  4  5  6  7 hesitant  1 pleasant  2  3  4  5  6  7 unpleasant  1 competitive  2  3  4  5  6  7 cooperative  1 l  a  z  v  84  3.  Next, please rate member  of your group on these adjectives:  1  7 unfair  fair  1  unfriendly  friendly  1  7 cold  warm  1  7 hard-working  lazy  1  self-assured  hesitant  1  pleasant  unpleasant  competitive  cooperat ive  8.5  4.  How much do y o u want t o c o n t i n u e w o r k i n g  1  i n t h i s group  6  2  v e r y much  5.  How much do y o u l i k e  1  the people  i n your  6.  How much do y o u l i k e  1  How much do y o u l i k e  1  7 very  little  7 very  little  very  little  very  little  member  2  v e r y much  7.  task?  group?  6  2  v e r y much  f o r the next  member  2  v e r y much  86  8.  How w e l l  do you t h i n k y o u r g r o u p  1 2 very well  9.  2  How much d i d member  1 v e r y mtmh  11.  3  4  How much d i d you c o n t r i b u t e t o y o u r g r o u p ' s  1 v e r y much  10.  performed?  4  6  2  3  4  5  6  5  3  4  87  5  7 very  little  performance?  6  c o n t r i b u t e t o your group's  2  7 very poorly  performance?  c o n t r i b u t e t o your group's  How much d i d member  1 v e r y much  3  5  7 very  little  performance?  6  7 very  little  Questionnaire 2  1.  How f a v o r a b l e a r e y o u r f e e l i n g s about y o u r group? 1  2  3  very favorable  2.  5  4  6  not a t a l l favorable  How f a v o r a b l e a r e y o u r f e e l i n g s about the o t h e r group? 1  2  3  4  5  6  very favorable  3.  7  7  not at a l l favorable  How d i f f e r e n t d i d the o t h e r groups i d e a s seem from y o u r g r o u p ' s i d e a s ' 1  2  3  4  5  very different  6  7  not at a l l different  88  APPENDIX D - CONSENT AND  89  DEBRIEFING FORMS  Consent Form The p r e s e n t s t u d y about " B r a i n s t o r m i n g " i s conducted by Anna F r i t z and Dr. P h i l i p Smith from the P s y c h o l o g y Department a t  UBC.  In our study you w i l l be asked t o p a r t i c i p a t e i n t a s k s t h a t i n v o l v e b r a i n s t o r m i n g and , a l s o , t o answer  some q u e s t i o n s about y o u r  own  and o t h e r s p a r t i c i p a t i o n . The e n t i r e s e s s i o n w i l l l a s t a p p r o x i m a t e l y 45 min. No p e r s o n a l i d e n t i f i c a t i o n of any k i n d w i l l ever be r e q u i r e d o f you as we want t o p r e s e r v e complete  anonymity.  S h o u l d you have any q u e s t i o n s about the p r o c e d u r e p l e a s e f e e l f r e e t o ask them so t h a t i t i s u n d e r s t o o d f u l l y by  everyone.  Y o u r p a r t i c i p a t i o n i n t h i s study i s c o m p l e t e l y v o l u n t a r y . I t i s your r i g h t t o w i t h d r a w from i t a t any time without" e x p l a n a t i o n and w i t h no p e n a l t y . A complete, d e t a i l e d d e s c r i p t i o n o f the study w i l l be g i v e n a t the end of the s e s s i o n .  I have read and u n d e r s t o o d the above, and I agree t o p a r t i c i p a t e i n t h i s s t u d y . I have r e c e i v e d a copy of t h i s c o n s e n t form.  (date)  (signed)  90  Debriefing  Statement  We would l i k e t o thank you f o r p a r t i c i p a t i n g i n t h i s s t u d y and t e l l you why we a r e c o n d u c t i n g t h i s study. O f t e n , when groups a r e i n i n t e r a c t i o n w i t h one a n o t h e r , a so-called  ' a s s i m i l a t i o n - c o n t r a s t * e f f e c t may be observed.  By " a s s i m i l a t i o n - c o n t r a s t ' i s meant the tendency i n people to enhance the s i m i l a r i t i e s among people i n t h e i r own group and  t o exaggerate the d i f f e r e n c e s between t h e i r own and the  o t h e r group. Ingroup members a r e perceived equivalent perceived The  and e v a l u a t e d  as s i m i l a r and  p o s i t i v e l y , and outgroup members a r e  as d i s s i m i l a r from the ingroup and e v a l u a t e d  negatively.  a s s i m i l a t i o n - c o n t r a s t phenomenon can o f t e n be observed i n the  r e a l w o r l d , f r o example when d i f f e r e n t r a c i a l o r e t h n i c groups i n t e r a c t , o r two s p o r t s teams. I n l a b o r a t o r y s t u d i e s , t o o , i t i s a w e l l - e s t a b l i s h e d f i n d i n g . So f a r , however, i t has o n l y involved simple,  unstructured  groups.  I n t h e p r e s e n t study we a r e t e s t i n g whether the a s s i m i l a t i o n c o n t r a s t tendency i s a l s o observed i n groups w i t h  structure.  A group w i t h s t r u c t u r e i s one i n which the members occupy d i f f e r e n t r o l e s o r p o s i t i o n s . The q u e s t i o n  i s whether the  d i f f e r e n t p o s i t i o n s people occupy i n a s t r u c t u r e d group have an i n f l u e n c e on the a s s i m i l a t i o n - c o n t r a s t t e n d e n c i e s such t h a t people w i t h i n one group w i l l be p e r c e i v e d  as d i f f e r e n t r a t h e r  than s i m i l a r . To t e s t t h i s we a r e examining how people i n a s t r u c t u r e d group v e r s u s i n an u n s t r u c t u r e d  group respond t o  o t h e r i n g r o u p members and t o members o f the outgroup. So you e i t h e r j u s t p a r t i c i p a t e d i n a s t r u c t u r e d group, i n an group o r as a c o n t r o l group p a r t i c i p a n t .  91  unstructured  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items