Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Are persons high in the need for structure more influential communicators of stereotypes? Conway, Lucian Gideon 1998

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

Item Metadata

Download

Media
831-ubc_1998-0142.pdf [ 2.61MB ]
Metadata
JSON: 831-1.0088436.json
JSON-LD: 831-1.0088436-ld.json
RDF/XML (Pretty): 831-1.0088436-rdf.xml
RDF/JSON: 831-1.0088436-rdf.json
Turtle: 831-1.0088436-turtle.txt
N-Triples: 831-1.0088436-rdf-ntriples.txt
Original Record: 831-1.0088436-source.json
Full Text
831-1.0088436-fulltext.txt
Citation
831-1.0088436.ris

Full Text

ARE PERSONS HIGH IN THE NEED FOR STRUCTURE MORE INFLUENTIAL COMMUNICATORS OF STEREOTYPES? by LUCIAN GIDEON CONWAY, III B.A., Baylor University, 1990 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES Department of Psychology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA March 1998 © Lucian Gideon Conway, III, 1998 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) II A B S T R A C T It was predicted that individuals with a high chronic need for structure (compared to persons with low need for structure) exert more influence in the processes through which individual perceptions of groups coalesce into consensual stereotypes. This prediction emerges from the joint consideration of two hypotheses: (1) Need for structure is hypothesized to influence the use of abstract language when talking about others; (2) More abstract language is hypothesized to exert greater influence on others' beliefs. Two elements of linguistic abstraction were examined: The extent to which language implies characteristics of groups rather than merely characteristics of individuals ("inclusiveness"); the extent to which language implies stable traits, rather than merely episodic behaviors ("implied stability"). To test the hypotheses, participants in dyads were presented with information about members of two novel groups, and engaged in structured interpersonal communication about this information. Study 1 tested the first hypothesis, and found no support. There was no evidence that individual differences in Need for Structure influenced either the "inclusiveness" or "implied stability" of interpersonal communication. Study 2 tested the second hypothesis and found partial support. The "implied stability" of communications had no effects on stereotype formation, but the "inclusiveness" of communication did have an impact. Participants receiving more "inclusive" communications formed more stereotypic beliefs. The latter effect emerged only on stereotypic beliefs about negative characteristics, but not on stereotypes about positive characteristics. Ill TABLE OF CONTENTS Abstract Table of Contents List of Tables Acknowledgment Chapter One Chapter Two Chapter Three The Consensual Nature of Stereotypes Communication and Consensus The Present Framework Why Might the Person High in Need For Structure Be More Influential? "Seizing" and "Freezing" Differences The Effect of Need For Structure on Communication Content The Hypothesized Effect of Communication Content On Stereotype Formation: The "Weight" of Abstract Information Putting It All Together: A Sample Scenario The Present Hypotheses Study 1 Methods Participants Procedures Stimulus Materials Phase 1 VI 1 2 5 6 6 10 11 13 13 13 14 15 IV Chapter Three (cont.) Chapter Four Phases 2-5 15 Content Coding of Communication "Implied Stability" 16 Content Coding of Communication "Inclusiveness" 17 Study 1 Results Chapter Five Bibliography Study 1 Discussion Study 2 Methods Participants Procedures Stimulus Materials Ersatz Notes Phase 1 Phases 2-5 Sentence Descriptions Similarity Measure Distribution Task Study 2 Results Sentence Description Measure: Change Between Time 1 and Time 2 Similarity Measure: Time 2 Only Distribution Task: Time 2 Only Study 2 Discussion General Discussion 18 18 20 20 20 21 22 23 23 23 24 24 25 25 30 31 33 36 51 LIST OF TABLES Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Table 11 Study 1: Correlations of Communication Inclusiveness and Implied Stability Level With Personality Measures 39 Study 1: Inter-Correlations of Personality Measures 39 Study 2: Summary of Ersatz Notes by Condition 40 Study 2: Index of Dependent Variable Names 41 Study 2: Effect of Communication Inclusiveness Level on Change Scores 43 Study 2: Effect of Communication Implied Stability Level on Change Scores 44 Study 2: Effect of Communication Inclusiveness Level and Communication Implied Stability Level on Time 2 Similarity Measure Scores 45 Study 2: Effect of Communication Inclusiveness Level and Communication Implied Stability Level on Time 2 Distribution Task Scores 46 Study 2: Effect of Communication Inclusiveness Level on Time 2 Sentence Description Scores 47 Study 2: Effect of Communication Implied Stability Level on Time 2 Sentence Description Scores 48 Study 2: Inter-Correlations of Personality Measures 49 VI ACKNOWLEDGMENT I would like to thank my wife, Kathrene Renee Conway, without whose unwavering love and support I am sure that this part of the dream would never have come true. 1 CHAPTER ONE The Consensual Nature of Stereotypes Stereotypes have historically been considered as culturally-shared knowledge structures (see Ashmore & Del Boca, 1981) and are still regarded as such by researchers today (Haslam, 1996; Schneider, 1996; Stangor & Jost, 1997). Indeed, there is good reason for emphasizing their consensual nature: the consequences of stereotypes often arise, not because an individual person holds an individual stereotype, but rather because the majority of people in a given culture hold a similar stereotype (Devine, 1989; Haslam, 1996; Steele & Aronson, 1995). In commenting on various stereotypes that have impacted the course of history ~ such as the Nazi stereotype of the Jew - Haslam (1996, p. 119) notes that "the stereotypes became potent because, within the groups who held them, they came to reflect and express a particular world view which dictated that group's collective behavior towards particular targets of oppression." In spite of the importance of the consensual nature of stereotypes, few psychological theories have actually accounted for how they become consensually shared (Gardner, 1993; Haslam, 1996; Stangor & Schaller, 1996). Indeed, most approaches have constrained themselves to how individual people process information about groups, and only address the question of consensuality indirectly by focusing on commonly-held motives in individuals, which, by extension, can lead to the formation of commonly-held beliefs. For example, research on self-esteem motives - within the framework of Social Identity Theory ~ has demonstrated that 2 one consequence of the very common desire to maintain social self-esteem is that stereotypes of other groups will tend to be negative, so that one's own group will be relatively "more" positive (e.g. Brewer & Kramer, 1985; Brown, 1978; Capozza, Bonaldo, & Dimaggio, 1982; Messick & Mackie, 1989; Tajfel 1981, 1982). Another example comes from research on epistemic motives: because of the desire to be accurate, stereotypes may commonly reflect the things that most accurately distinguish one group from another (Ford and Stangor, 1992). However, these strictly individual-level approaches have difficulty in explaining how stereotypes become consensually shared (Boster, 1991; Schneider, 1996). As Stangor and Jost (1997, p. 34) point out, "these models of 'common informational input'. . . are clearly limited, both because individuals develop stereotypes about groups with which they have had no direct contact, and because contact is notoriously ineffective at changing stereotypes." It is becoming increasingly clear that, if we want to understand stereotypes, we will have to more fully understand the processes that account for their status as shared beliefs (Schaller & Latane, 1996; Stangor & Jost, 1997). Communication and Consensus Indeed, many psychologists now argue that individual-level approaches will have to make room for interpersonal-level approaches in order to gain a richer understanding how beliefs such as stereotypes form (Latane, 1996; Moscovici, 1993; Schaller & Latane, 1996). An obvious starting point in the quest for an interpersonal-level approach is to understand how communication processes are important in stereotype consensus. Currently, it is generally assumed that consensus emerges 3 from communication because humans desire to validate their beliefs through other people (Hardin & Higgins, 1996) or that consensus grows out of a need for ingroup belongingness (Haslam, 1997). Thus research has shown that manipulating the perceptions of stereotype consensuality altered participants' endorsement of those stereotypes (Stangor, Sechrist, Jost, & Bar-Tal, 1996) and that, at least when men and women are the reference groups, individuals discuss outgroups more stereotypically and more negatively than ingroups (Harasty, 1997). Most of this research assumes that consensus motivations lead to consensually-held beliefs through communication processes such that, with respect to stereotypes, communication is simply the medium that carries out our desire to achieve consensus; but this approach forgoes any specific analyses of how the communication processes themselves actually work. Little research has accounted for exactly how, in stereotype formation, communication leads to belief consensus, or, more broadly, how communication processes influence stereotype formation. Results reported by Harasty (1997) show how communication patterns differ along various specific, predictable dimensions depending on whether the group discussed is the in-group and the out-group. However, exactly how those communication differences themselves in turn influence the development of stereotypes is a question that, as of yet, no research has specifically addressed. As Harasty (1997, p. 282) states: "The effect of such socially transmitted group information on stereotype formation is yet to be determined." It is clear that to fully develop an interpersonal approach to stereotype formation, we will have to understand the effect that such communication processes ultimately have on how stereotypes form. The purpose of the present research is to help fill in these gaps by (a) articulating a theoretical framework for understanding, from a psychological perspective, how stereotypes come to be shared across different minds through communication processes, and then (b) identifying and testing one particular set of hypotheses, concerning an individual-level factor relevant to communication, that emerges from this theoretical framework. 5 CHAPTER TWO The Present Framework Latane has suggested in his Dynamic Social Impact Model that, with respect to the communication of cultural beliefs, not all persons are equally influential (Latane, 1996; see also Schaller & Latane, 1996). But what is it, exactly, that makes a person an influential communicator? Our intuitions might, probably with some accuracy, guess that such influence is in part a function of physical size, intellect, or wealth; and, indeed, Latane suggested these same three factors as important to the "stable, tran-situational" aspect of influence (Latane, 1996, p. 14). But there may perhaps be a subtler kind of "stable" influence that lies in an altogether different, and less obvious, type of individual difference: the need for structure. The need for structure (used here interchangeably with the conceptually and empirically similar construct "need for closure") is "the desire to mentally structure one's environment in a simplified, manageable manner" (Schaller, Boyd, Yohannes, & O'Brien, 1995, p. 545). Although influenced by situations (e.g., Kruglanski & Webster, 1996), need for structure can also be measured as a stable personality attribute, sometimes called "personal need for structure (PNS; Thompson, Naccarato, & Parker, 1992). The basic thesis of the present framework is that when persons of differing needs for structure interact, those persons who are high in PNS will be more influential, effective communicators of stereotypes than persons who are low in PNS. 6 Why Might the Person High in Need for Structure Be More Influential? "Seizing" and "Freezing" Differences Why might the person who is high in PNS be more influential in stereotype formation? The first half of this story follows from the well-established tendencies for PNS to influence two different mental processes: "seizing" and "freezing" (Kruglanski & Webster, 1996). Let us look at these two distinct aspects in turn. Seizing. Seizing, which results from the desire for simple structure, is the cognitive act of quickly converting incoming information into an established belief structure. The desire to seize equates to a "quickened pace and enhanced volume of the informational search under heightened need for closure" (Kruglanski & Webster, 1996, p. 266). Thus, persons chronically high in the need for structure/closure are more likely to engage in the seizing process (Ford & Kruglanski, 1995; Kruglanski & Webster, 1996; Webster, 1993; Webster & Kruglanski, 1994). With respect to stereotypes, this means that high-PNS persons should generally be more likely to "seize" upon group-relevant information to come to a point of belief about that group. Thus, high-PNS and low-PNS persons will -- cognitively speaking - respond differently to the same stereotype-relevant information observed in their social environment (Ford & Kruglanski, 1995; Thompson, Roman, Moscovitz, Chaiken, & Bargh, 1994). Where both observe that Mitch, a mathematician, hit his wife, the low-PNS person may postpone any judgment of the group "mathematicians," whereas the high-PNS person may, driven by this need for simple structure, tend to "seize" the stereotype-relevant information and conclude that the 7 "mathematicians are aggressive." Thus given the same informational input, stereotypes about groups are formed more quickly for the high-PNS versus the low-PNS person. Freezing. Once a belief has crystallized to the point that it is confidently construed, then "freezing," or holding on to that belief even in the face of contradicting evidence, can occur. Freezing often manifests itself as "a reluctance to continue information processing or a resistance to persuasive arguments aimed at undermining one's current closure and effecting cognitive change" (Kruglanski & Webster, 1996, p. 266). And again, because freezing is driven by the need for simple structure (not wanting to change the structure already in place), differences in freezing can be observed in the degree to which high- versus low-PNS persons engage in the freezing process, with the high-PNS persons doing so with more vigor (Ford & Kruglanski, 1995; Kruglanski & Webster, 1996; Webster & Kruglanski, 1994). Seizing, Freezing, and Stereotype influence. This process potentially bears on the question "why are persons high in PNS more influential communicators of stereotypes?" High-PNS persons may form stereotypes more quickly and also be more difficult to persuade from those stereotypes than low-PNS persons. So high-PNS persons may be stubborn and obstinate about their stereotypes (and communicate them as such), while low-PNS persons may be more open to influence by others (including their high-PNS counterparts); thus, those high in PNS may be relatively more influential. But this is only half of the story. 8 The Effect of Need for Structure on Communication Content The other half of the story lies in the hypothesized effects that PNS has on communication, and in the subsequent hypothesized consequences of those effects. What type of effects might PNS have on communication processes? Persons high in PNS may communicate about other people in more abstract terms than persons low in PNS. There are two different conceptualizations for the notion of "communication abstractness" that are relevant to the present research. One conceptualization of abstractness is "implied stability." Implied stability is the degree to which a particular word seems to represent a stable, unchanging state or trait. According to Semin and Fiedler's (1988; 1991) Linguistic Category Model, the most concrete descriptors are descriptive action verbs, which refer to a single behavioral event, like "kick" or "hit." Next on the continuum are interpretive action verbs, which also refer to a single behavioral event but convey a broader meaning that usually carries a positive or negative tone, like "cheat" or "hurt." More abstract are state action verbs, which refer to a general class of behaviors and convey information about the object of the action as well as the subject, as in "he angered the class." More abstract still are state verbs, which also refer to emotional states, but have no clear beginning and end point, such as "hate." Most abstract of all are the adjectives such as "aggressive" and "assertive" (see Semin & Fiedler, 1991, for an extensive review of these distinctions). There is some evidence that persons high in PNS are more likely to use adjectives and state verbs than persons low in PNS in describing behaviors (Webster, Kruglanski, & Pattison, 1997; Rubini & Kruglanski, 1997). 9 The second conceptualization for abstractness is "inclusiveness." Inclusiveness is the extent to which a description refers to a group of people, rather than to a single individual. A "given description can refer to a specific individual, a subgroup of individuals, a social group, or people in general" (Harasty, 1997, p. 272). Although there is no existing research on the relationship between PNS and inclusiveness, it is expected that persons high in PNS will use more inclusive, group-level descriptions, while persons low in PNS will use less inclusive, person-level descriptions. This notion makes sense in light of the previously-discussed tendency for high-PNS persons to "seize" the stereotype-relevant content in any given information: if they believe that Mitch's aggressive behavior reflects on the entire group of mathematicians, then they are more likely to communicate information about the aggressiveness of mathematicians in general than if they believed that information was specific to only to Mitch. So a person high in the need for structure might communicate "mathematicians are aggressive," while a person low in PNS might simply communicate about the tendencies of Mitch. The Hypothesized Effect of Communication Content on Stereotype Formation: The "Weight" of Abstract Communication How might individual differences in abstractness of communication be relevant to the emergence of shared stereotypes? It may be that abstract language will be more likely to influence the rate of stereotype formation because it is difficult to falsify. The more abstract a statement is, the less it can be directly falsified. Because 10 at the most abstract level information is in the "final form" it can potentially take (in the sense that it is at its most simplistic level), it cannot very easily be undone by new information. For example, the high-inclusiveness, high-implied stability statement "mathematicians are aggressive" cannot be directly falsified. Even if I see a clear example of one mathematician behaving quite passively one time, that does not account for the rest of the members within the prototyped group "mathematicians." Nor does that one act necessarily reflect something internal about that member, and thus does not have to be inconsistent with his "aggressiveness." Semin and Fiedler (1991, p. 432) brought the issue to bear directly on stereotypes this way: "It is tempting to theorize that this property of abstract language may explain why social stereotypes are often represented and conserved at the most abstract level of ADJ (traits or dispositions), which are detached from the empirical reality and therefore immunized from direct falsification." Putting It All Together: A Sample Scenario From the current perspective, why would all this mean that persons high in PNS are more influential in stereotype formation? Consider an imaginary scenario where Hyman, who is high in PNS, and Loman, who is low in PNS, each observed the exact same social information: to be consistent, let's say they first saw that "Mitch, a mathematician, hit his wife." Hyman might immediately seize upon the first bit of stereotype-relevant information and conclude that "mathematicians are aggressive," while Loman would be more content to hold off any judgment of mathematicians as a whole. Now imagine that Hyman and Loman sat around and talked about their impressions of mathematicians. Hyman would convey to Loman heavily-weighted abstract information about how mathematicians are aggressive, while Loman would simply communicate an individual bit of information concerning Mitch. The result? Loman would likely be "pulled" in the direction of Hyman's aggressive stereotype, due to the weight the abstract information; and thus the high-PNS Hyman would be more influential in the formation of an ultimately consensually-held stereotype of mathematicians. The Present Hypotheses The framework outlined here leads to some novel predictions. 1.) There will be a positive correlation between personal need for structure and abstractness of communication (along both the implied stability and inclusiveness dimensions) when persons are communicating stereotype-relevant information about group members. 2.) Abstract communication (along the two dimensions) about group members will lead to more rapid stereotype development than non-abstract communication. Accordingly, Study 1 tested hypothesis 1, and Study 2 tested hypothesis 2. In Study 1, participants were presented with information describing the behavior of various members of two different (fictitious) groups. After completing a battery of questionnaires (one of which is the PNS measure), they were paired with another participant and told that they were to write notes to that participant concerning their impressions of the groups, and that these notes would be passed all at once at the end of the session. This, in fact, never occurred. The notes were coded for abstractness along the two conceptualizations outlined earlier, and correlations were computed with individual differences in personal need for structure. In Study 2, participants were presented with the same information describing 12 the behavior of various members of two different groups. However, in this study, rather than being asked to write notes, they were told that they were in a "read notes" condition, during which they would read notes that participants in a previous session had written. These notes, in actuality, were carefully crafted by the experimenter to differ along the two abstractness dimensions of implied stability and inclusiveness. Both before and after reading the "notes," participants were given measures which assess their emerging beliefs about the groups. Expectations were that the notes higher in implied stability and inclusiveness would be more likely to influence the rate of stereotype formation than notes low in implied stability or inclusiveness. 13 CHAPTER THREE Study 1 Methods Participants Thirty-nine undergraduate students at the University of British Columbia (26 female, 13 male) participated for class credit in groups ranging from 2-6 members (the mean was 4.9). Procedures Upon arriving, participants were seated two to a table. They were told that the experiment was about the types of things that people communicate about when they talk to other people ~ particularly when they talk about other people. Participants were instructed that the experimenters first wanted to get some background information on them, and they were given a packet of questionnaires (this packet contained, in the given order, the Personal Need for Structure Questionnaire (Thompson, Naccarato, and Parker, 1992), a "Big 5" questionnaire ~ adapted from McCrae and Costa (1983) - containing ten bipolar items for Conscientiousness, ten bipolar items for Extroversion, and ten bipolar items for Openness to Experience, and the Aspects of Identity lllx Questionnaire (Cheek, 1994). The additional questionnaires were given for two reasons. First, they provided a means for analyzing the degree to which PNS is uniquely related to communication abstractness. Second, they were intended to ensure that the purpose of the experiment was fully veiled to participants, and as such to remove any concerns about demand explanations). After completing this packet, participants 14 were then informed that, over the course of five separate phases, they would receive information about individuals who belonged to one of two groups (the "Red Group" and the "Blue Group"). Further, during phases 2 through 5, they would be asked to communicate with the other person at their table - by writing notes ~ concerning their impressions of the groups; these notes would be passed all at once to that other person near the end of the experiment. Stimulus Materials Over the course of five phases, participants were presented with 18 index cards on which were printed brief concrete descriptions of (male) members of the Red Group and the Blue Group. These descriptions focused on behaviors relevant to aggressiveness and intelligence. For example: Dave is a member of the Red Group. He is 6'2". He scored a 98 on his last microbiology exam and has an "A" in quantum physics. During his summer job as a dishwasher, he frequently and deliberately breaks glasses and dishes as a way to lash out against his boss, whom he does not like. Nine cards described members of the Red Group, and nine described members of the Blue Group. These descriptions were crafted so that, in general, members of the Red Group were portrayed as more aggressive and more intelligent than members of the Blue Group. Six of 9 Red Group members were portrayed as highly aggressive, whereas 3 of the 9 Blue Group members were portrayed as highly aggressive; 6 of the 9 Red Group members were portrayed as highly intelligent, whereas only 3 of the 9 Blue Group members were portrayed as highly intelligent. Importantly, each description was designed to be as concrete as possible on 15 both the implied stability dimension (most were descriptive action verbs; Semin & Fiedler, 1992) and the inclusiveness dimension (all were individual references). The effects predicted here were expected only when the initial social information itself is not abstract. If such initial information about the groups was in a highly abstract form, the weight of that abstract information might overwhelm any effects due to personal need for structure or communication differences. In the first of the 5 phases, participants received 6 cards; in all subsequent phases, they received 3 cards. All participants received the same stimulus cards in the same order. Phase 1 To give participants a chance to form an initial impression about the two groups, phase 1 consisted of participants reading 6 information cards from the experimenter. Participants were told that this phase was a "kind of preliminary phase to let them get acquainted with the groups." Participants also completed a Sentence Descriptions measure adapted from Ford & Stangor (1992), which asked them to write (up to) four sentences about each group, after reading the information1. Phases 2-5 Participants were told that Phases 2 through 5 were the "communication phases." At each of these phases, two distinct steps happened. 1.) Participants received three new cards of information to read and subsequently place in a "Discard Box" (participants were not allowed to remove and re-read things once they were placed in the "Discard Box"). 2.) Second, participants wrote a communication to the 16 other participant. There were two rules that governed the note-writing. First, participants were allowed to only write about one of the two groups at any given phase; and second, each communication had to contain exactly two sentences. Both rules were captured on the "note template," which participants use to write their notes to the other participant; the template read: "Neighbor, here are exactly two sentences which describe the Group." After each participant had finished the communication and placed it in a designated box (dubbed the "Mailbox"), the experimenter proceeded on to the next phase. This pattern continued until all phases were complete. Content Coding of Communication "Implied Stability" Each note written by participants was coded by a rater (blind to the participants' PNS scores) for the total number of group-relevant descriptions that it contained. For all codings, words (e.g., "aggressive") and verb phrases (e.g., "likes to hurt others") were treated as one description. Each of these descriptions were then coded for level of implied stability using a monotonic scale progressing from the most concrete to the most abstract. For this score, descriptive action verbs were coded as "1," interpretive action verbs as a "2," state action verbs as a "3," state verbs as a "4," and adjectives as a "5" (see Webster, Kruglanski, & Pattison, 1997). It should be noted that, although the basic structure behind the Linguistic Category Model (Semin & Fiedler, 1988; 1991) has been well-established, its application in this study is unique and as such presented some unique problems; thus, some adaptations were made to the usual coding scheme. For example, rather than rigidly coding the verb "have" as a "4" (it is a state verb), it was coded according to the 17 broader context, such that "the red group has good grades on tests" was rated as more concrete than "the red group has ambition." (Saying that a group has good grades on tests implies less stability than saying that the red group has ambition.) The final implied stability score was computed by dividing the sum of the implied stability ratings in each note by the total number of descriptions in that note. The resulting score represents the average level of implied stability contained in each description. Content Coding of Communication "Inclusiveness" Each descriptor was also coded for inclusiveness level. To pick up on more of the context of participants' communications, two new categories were added to Harasty's (1997) inclusiveness formulation. Thus, for the inclusiveness score, an individual reference ("one member of the red group") was coded as a "1." A reference to a small part of the group (e.g., "a few members of the red group") was coded as a "2." A reference to part of the group (e.g., "some of the members of the red group") was coded as a "3." A reference to a large part of the group (e.g., "most of the members of the red group") was coded as a "4." A full group reference (e.g., "the red group") was coded as a "5." The final inclusiveness score was computed by dividing the sum of the inclusiveness ratings in each note by the total number of descriptions in that note. The resulting score represents the average level of inclusiveness contained in each description. (For all codings, a second rater coded a sub-sample in an identical manner for interrater reliability purposes. These reliabilities were satisfactory: the overall 18 correlation coefficient between the two coders for implied stability was .91 and for inclusiveness was .97). Study 1 Results Correlations revealed no support for the hypothesis that PNS is positively correlated with abstractness of language when communicating: the correlation between the overall average implied stability and PNS was .06 (p=.70), and the correlation between the overall average inclusiveness and PNS was -.21 (p=.205). Furthermore, no reliable correlation coefficients emerged in any interpretable pattern between abstractness of communication and the three sub-scales of the Aspects of Identity Questionnaire (Collective Identity, Social Identity, and Personal Identity), Extroversion, Conscientiousness, or Openness to Experience. Table 1 presents a summary of this information, and Table 2 presents the inter-correlations between the various personality measures. Study 1 Discussion Study 1 clearly offers no support for the hypothesis. Why did it fail? One possibility is that the instructions given to participants may have (unintentionally) induced high levels of both implied stability and inclusiveness, thus not allowing the variability needed for a correlation. However, inspection of the distribution itself suggested that this was not the case. The means for implied stability (4.0 out of 5) and inclusiveness (3.5 out of 5) both suggested that there was no "ceiling effect" for high scores. Similarly, the standard deviations for implied stability (0.78) and inclusiveness (1.44) suggested that there was indeed enough variability to attain a 19 significant correlation. From one point of view, the failure to find a positive correlation between need for structure and abstractness (at least when abstractness is considered along the implied stability dimension) can be considered a failure to replicate previous findings (Rubini & Kruglanski, 1997; Webster et al, 1997). One potential reason for this failure may be that both prior studies used a "forced choice," rather than the present study's "free response," measure of communication abstractness. While the differences between these two measurements seem superficial and it is admittedly difficult to see exactly how this could have mattered, it is worth keeping in mind that oftentimes apparently superficial differences can have large consequences on psychological processes -- even, as is the case here, when no explanation seems ¥ readily apparent. Another difference between the present study and past research is that the present study extends previous research on individual-relevant communication into the realm of group-relevant communication. Although again hard to discern the exact psychological nature of this potential discrepancy, it may well be the case that the relationship between need for structure and communication abstractness ends with communication about individuals, and as such does not hold for communication in an intergroup context. 20 CHAPTER FOUR Study 2 Methods Participants Eighty-two undergraduate students at the University of British Columbia (56 female, 26 male) participated for class credit in groups ranging from 3-6 members (the mean was 5.5). Procedures Upon arriving, participants were seated two to a table. They were told that the experiment was about the types of things that people communicate about when they talk about other people. Participants were instructed that the experimenters first wanted to get some background information on them, and they were given a packet of questionnaires (this packet contained, in the given order, the Personal Need for Structure Questionnaire (Thompson, Naccarato, and Parker, 1992), a "Big 5" questionnaire - adapted from McCrae and Costa (1984) - containing ten bipolar items for Conscientiousness, ten bipolar items for Extroversion, and ten bipolar items for Openness to Experience, and the Aspects of Identity lllx Questionnaire (Cheek, 1994)). After completing the packet, participants were then informed that, over the course of five separate phases, they would be receiving information about individuals who belong to one of two groups (the "Red Group" and the "Blue Group"). Further, all participants were told that some experimental sessions were in a "write notes" condition, where they wrote notes to be read by participants in another session. Participants were then informed that their experimental session was the other 21 condition, the "read notes" condition, such that, during phases 2 through 5, they would be reading a set of notes written by a participant from a previous session (participants were told that they had been randomly paired with one participant from a previous session). Stimulus Materials Over the course of five phases, participants were presented with 18 index cards on which were printed brief concrete descriptions of (male) members of the Red Group and the Blue Group. These descriptions focused on behaviors relevant to aggressiveness and intelligence. For example: Dave is a member of the Red Group. He is 6'2". He scored a 98 on his last microbiology exam and has an "A" in quantum physics. During his summer job as a dishwasher, he frequently and deliberately breaks glasses and dishes as a way to lash out against his boss, whom he does not like. Nine cards described members of the Red Group, and nine described members of the Blue Group. These descriptions were crafted so that, in general, members of the Red Group were portrayed as more aggressive and more intelligent than members of the Blue Group. Six of 9 Red Group members were portrayed as highly aggressive, whereas 3 of the 9 Blue Group members were portrayed as highly aggressive; 6 of the 9 Red Group members were portrayed as highly intelligent, whereas only 3 of the 9 Blue Group members were portrayed as highly intelligent. Each description was designed to be as concrete as possible on both the implied stability dimension (most were descriptive action verbs; Semin & Fiedler, 1992) and the inclusiveness dimension, (all were individual references). 22 In the first of the 5 phases, participants received 6 cards; in all subsequent phases, they received 3 cards. All participants received the same stimulus cards in the same order. This order was determined on two criteria: first, that the distribution of aggressive and intelligent Red and Blue Group members in the first phase matched the overall distribution, and second, that the first card read by participants of the Red Group was an aggressive and intelligent member, while the first card of the Blue Group was an unaggressive and unintelligent member (this was intended to ensure that any primacy effects were in the direction of the overall distribution). Ersatz Notes The "notes" that participants received were actually crafted in advance by the experimenters. Each note contained two sentences. All participants received four notes total - two concerning the Red Group and two concerning the Blue Group -and they were always received in the same order. The notes were written to vary along both dimensions of abstractness: implied stability and inclusiveness. These two dimensions were crossed to create four different sets of notes, and thus the four different conditions: 1.) low implied stability and low inclusiveness (e.g, "One member knocked down beginner skiers in lift lines trying to get to the front"), 2.) high implied stability and low inclusiveness (e.g., "One member is very ambitious"), 3.) low implied stability and high inclusiveness (e.g., "This group always knocks down beginner skiers in lift lines trying to get to the front"), and 4.) high implied stability and high inclusiveness (e.g, "This group is very ambitious"). Across all conditions, the ersatz notes depicted the Red Group as both aggressive and intelligent, and the Blue Group as both passive and unintelligent (see 23 Table 3 for a summary of all of the notes in each condition). Phase 1 To give participants a chance to form an initial impression about the two groups, Phase 1 consisted of participants receiving 6 information cards from the experimenter and then completing a "sentence description" measure (to be described later). Participants were told that this phase was a "kind of preliminary phase to let them get acquainted with the groups." Phases 2-5 Participants were told that Phases 2 through 5 were the "communication phases." At each of these phases, two distinct steps happened. 1.) Participants received three new cards of information to read and subsequently place in a "Discard Box" (participants were not allowed to remove and re-read things once they were placed in the "Discard Box"). 2.) Second, participants read one "note" that they believed to be written by a participant in a previous session (this note was actually crafted in advance by the experimenters.) Sentence Descriptions One way to conceptualize a stereotype is that it is a belief about a group that is readily retrievable from memory. Accordingly, after Phase 1 (henceforth referred to as "Time 1") and again after Phase 5 (henceforth referred to as "Time 2"), participants were given a modified version of a free response measure developed by Ford and Stangor (1992). Participants were presented with two sheets of paper. On one sheet were four empty boxes and the instructions: "Writing only one sentence 24 per box, write as many sentences as you feel comfortable that accurately describe the Red Group." On the other sheet were four empty boxes and the instructions: "Writing only one sentence per box, write as many sentences as you feel comfortable that accurately describe the Blue Group." Before participants began the sentence description measure, it was verbally emphasized that they were free to write as few or as many sentences as they felt confident actually described the groups. Similarity Measure In addition to the sentence descriptions measure, at Time 2 (only) two additional measures were given to participants. The first ("Similarity Measure") involves simple but broad questions about the perceived similarity of each of the groups' members, for example, "How similar to each other do you think the members of the Red (Blue) Group are?" and "How similar to each other do you think the members of the Red (Blue) Group are on the trait aggressiveness?" Distribution Task The second, and more complex, measure, was a distribution task similar to one that has been used in previous research (Linville, Fischer, & Salovey, 1989; Judd & Park, 1988; Park & Judd, 1990). This task asked participants to imagine that there were 100 people in the Red Group, and then to place how many of those 100 people would fall into each of five different categories. These categories are relevant to two of the specific dimensions being tested: "aggressiveness" and "intelligence." Thus, for the trait "aggressiveness," the categories are "well below average 25 aggressiveness," "below average aggressiveness," "average aggressiveness," "above average aggressiveness," and "well above average aggressiveness." This same pattern is repeated for the Blue Group along both the aggressiveness and intelligence dimensions. Study 2 Results Sentence Description Measure: Change Between Time 1 and Time 2 "Aggressive" as the Focal Trait. Because there is no single best manner to capture stereotype content, the Sentence Descriptions measures were coded in three computationally related, but conceptually distinct, ways. Word Content Scores. One way to conceptualize stereotypes is simply whether or not a particular trait is included in the stereotype. Accordingly, each sentence in the sentence description measure was coded by a rater blind to condition along two dimensions: the total number of aggressive-relevant words describing the Red Group and the total number of unaggressive-relevant words describing the Blue Group. For these codings, verb phrases (e.g., "likes to hurt people") were coded as one word. These two scores were produced by simply adding the total number of words (or phrases) along each dimension. First Sentence Scores. Another way to conceptualize stereotypes is that those things which are most likely to spontaneously spring to mind - and thus to be written down first - are most central to a given stereotype. This approach assumes that initial stereotype-relevant responses are strongly linked in associative memory to the group label (see Ford & Stangor, 1992). Accordingly, for the two Word Content 26 Scores, a separate score was computed -- in an identical manner - that just accounts for the very first sentence written by participants (e.g., Schaller & Conway, 1997). Inclusiveness Scores. Yet another way to conceptualize stereotypes is that a stereotype is something that is believed to apply to a large percentage of the members of a given group. Because stereotypes are by definition represented at the highest inclusiveness level (group description), the degree to which a belief represents a large degree of inclusiveness is a rough measure of its generalizability as a stereotype. Accordingly, each word (or phrase) describing the Red Group as aggressive or the Blue Group as unaggressive, was additionally coded for its inclusiveness level, using the same monotonic scale described in Study 1. Two Inclusiveness Scores were then computed from these codings by multiplying the total number of words (or phrases) along one dimension by the average level of inclusiveness on that same dimension. Thus, the Red Group Aggressive Inclusiveness Score represents the number of aggressive words about the Red Group multiplied by the average inclusiveness of those words, and the Blue Group Unaggressive Inclusiveness Score represents the number of unaggressive words about the Red Group multiplied by the average inclusiveness of those words. (For all codings, a second rater coded a sub-sample in an identical manner for interrater reliability purposes; all reliabilities were satisfactory2.) Z-Score Composites. Finally, because no reasons existed for deeming one conceptualization of stereotypes better than the others, Z-Score composites were also computed for both the Red Group Aggressive and the Blue Group Unaggressive 27 dimensions. Specifically, each composite contained the Word Content, First Sentence, and Inclusiveness Score - all equally weighted - relevant to its dimension (e.g., the Red Group Aggressive Z-Score Composite contained the Red Group Aggressive Word Content, the Red Group Aggressive First Sentence, and the Red Group Aggressive Inclusiveness Scores). Although partially redundant, these two composites serve as a useful way of summarizing a large body of scores, and as such will function as the focal point of the analyses section (though all relevant non-composite information is included in tabular form; see Table 4 for a comprehensive index of variable names.) Finally, each Red Group Aggressive and Blue Group Unaggressive Z-Score Composite was computed for both the Time 1 (after Phase 1) and Time 2 (after Phase 5) Sentence Descriptions measures. To assess the degree to which participants changed their aggressive-relevant stereotypes during the experiment, a Change Score was computed for. both the Red Group Aggressive and Blue Group Unaggressive composites by subtracting each Time 1 score from its corresponding Time 2 score. These Change Scores represent the rate of aggressive-relevant stereotype formation throughout the experiment, such that, the higher the score, the faster an aggressive-relevant stereotype is emerging. Analyses of these Aggressive Change Scores revealed a mean pattern that was consistent with the hypothesis for inclusiveness (see Table 5)3. Interestingly, this effect was stronger for the Blue Group than for the Red Group: Z-Score composites revealed that participants in high (versus low) inclusiveness conditions indicated greater change in their beliefs concerning the Blue Group in the direction of 28 aggressive antonyms, F (1,78) = 4.254, p. = .042, and, to a lesser extent, indicated greater change in their beliefs concerning the Red Group in the direction of aggressive synonyms, F (1,78) = 1.869, p. = .176. An important feature of stereotypes is their ability to help us differentiate between groups (Schaller & Conway, 1997; for an explanation of the importance of diagnosticity in stereotypes, see Ford & Stangor, 1992). Thus, a Group Differentiation Aggressive Score was computed by summing the Red Group Aggressive Z-Score Composite and the Blue Group Unaggressive Z-Score Composite, and subtracting from that number the sum of a Red Group Unaggressive Z-Score Composite and Blue Group Aggressive Z-Score Composite (the Red Group Unaggressive and Blue Group Aggressive Scores were coded and computed in an identical manner to the Red Group Aggressive and Blue Group Unaggressive Scores). To the degree that the resulting value is positive, this indicates that the Red Group was perceived to be more aggressive than the Blue Group. This Group Differentiation Score revealed that participants in high (versus low) inclusiveness conditions demonstrated a greater rate of stereotype formation towards perceiving the Red Group as more aggressive than the Blue Group, F (1,78) = 4.555, p = .036. The effects of implied stability on the Aggressive Change Scores were generally consistent with the hypothesis (although note that all three effects may be due to sampling error; see Table 6). Specifically, Z-Score Composites revealed that participants in high (versus low) implied stability conditions indicated greater change in their beliefs concerning the Blue Group in the direction of aggressive antonyms, F (1,78) = 1.178, p = .281, and, to a much larger extent, indicated greater change in 29 their beliefs concerning the Red Group in the direction of aggressive synonyms, F (1,78) = 2.781, p. = .099. A Diagnosticity Z-score composite revealed that participants in high (versus low) inclusiveness conditions demonstrated a greater rate of stereotype formation towards perceiving the Red Group as more aggressive than the Blue Group, F (1,78) = 2.876, p = .094. No interactions between inclusiveness and implied stability were predicted or observed, all F's < .18. "Intelligent" as the Focal Trait. All sentence descriptions were also coded for their intelligent-relevant content and the Red Group Intelligent, Blue Group Unintelligent, and Group Differentiation Intelligent Change Composites were computed in a manner identical to those for the Aggressive Change Scores. The mean patterns of these Z-Score Composites did, at least for the Blue Group, generally conform to the hypothesis (although none of the Z-Score effects obtained conventional levels of significance). Specifically, a Difference Z-Score Composite for the Blue Group revealed that, in high (versus low) inclusiveness conditions, participants indicated greater change in their beliefs concerning the Blue Group in the direction of intelligent antonyms, F (1,78) = 2.967, p = .089. The Group Differentiation Intelligent Score revealed that participants in high (versus low) inclusiveness conditions demonstrated a greater rate of stereotype formation towards perceiving the Red Group as more intelligent than the Blue Group, F (1,78) = 2.436, p = .121. The effect of implied stability on intelligence produced generally weak results on the Difference Scores. However, it should be noted that the mean pattern for the 30 Difference Z-Score composites was inconsistent with the hypothesis (see Table 10). No interactions between inclusiveness and implied stability were predicted or observed, all F's < .40. Similarity Measure: Time 2 Only "Aggressive" as the Focal Trait Consistent with predictions, participants in high inclusiveness conditions rated both Red Group members and Blue Group members as more similar on the trait "Aggressiveness" than participants in low inclusiveness conditions, although there is a reasonable probability that these effects are due to sampling error, (Red Group: F(1,78) = 1.37, p = .245; Blue Group: F(1,78) = 2.665, p. = .107). No effects emerged on "Aggressive" for implied stability, all F's < .30. No interactions between implied stability and inclusiveness were predicted nor obtained, all F's < .600. (See Table 7 for a summary of Similarity Measure Scores). "Intelligent" as the Focal Trait. Consistent with predictions, participants rated the members of the Red Group as more similar with respect to the trait "intelligence" in high (versus low) inclusiveness conditions, F (1,78) = 4.857, p = .030. Looking at the Time 2 stereotype measures together, however, it is hard to know exactly what to make of this number, and indeed, given the sporadic and weak nature of the time 2 intelligent stereotype measures in general, it is tempting to suggest that it is a fluke. This is especially the case given that, for this measure, an interaction between implied stability and inclusiveness also emerged, F (1,78) = 7.756, p < .007, such that - contrary to predictions ~ under conditions where 31 inclusiveness was high, participants reported more similarity for the Red Group on intelligence when implied stability was low (M = 7.86) than when implied stability was high (M = 6.50). In addition, the effect of implied stability on the "Intelligent" similarity ratings of the Red Group was in the wrong direction, although this difference was likely due to sampling error, F(1,78) = 0.881, p.= .351. No effect of inclusiveness was obtained for the "Intelligent" similarity rating of the Blue Group (p. = .506). The effect of implied stability for the same score was nearer conventional levels of significance: specifically, participants in high (versus low) implied stability conditions reported the Blue Group members as more similar to each other on the trait "Intelligence," F(1,78) = 2.048, p. = . 156. (See Table 7 for a summary of Similarity Measure Scores). Distribution Task: Time 2 Only Both the mean and variance were computed from the distribution task (see Linville, Fischer, & Salovey, 1989, for an explanation of these computations). The resulting Mean Scores are measures of the degree to which "aggressive" and "intelligent" were represented in the stereotype. The resulting Variance Scores are measures of dispersion around the mean, and may reflect the percentage of members in a given group to which that stereotype content is applied. These Variance Scores themselves do not directly account for the content of the stereotype: rather, to the degree that a Variance Score is high, it suggests that, whatever content is present in a stereotype, it is weakly held - because it is not applied to a large number of people in a given group. 32 "Aggressive" as the Focal Trait Analyses on the effects of inclusiveness on the Aggressive Mean Scores revealed a mean pattern consistent with predictions: participants in high (versus low) inclusiveness conditions had a higher Mean Scores for the Red Group and lower Mean Scores for the Blue Group on the trait "aggressive." However,.none of these differences attained conventional levels of significance (see Table 8). Similar analyses on the effects of implied stability on the Aggressive Mean Scores revealed a pattern opposite to that predicted, but which was likely due to sampling error (all F's < 1.40). Analyses on the Aggressive Variance Scores suggested no consistent or reliable main effect for either inclusiveness or implied stability (all F's < 1.50; see Table 8). However, for both the Red and Blue Groups Aggressive score, an (unpredicted) interaction between inclusiveness and implied stability emerged, such that, in high inclusive conditions, the Variability Scores were higher for those receiving high (M/s = 433.3 and 397.0) versus low (M's =279.6 and 283.9) implied stability communications, F (1, 78) = 3.656, p. = .060. This result is inconsistent with the present framework, and will be addressed in the Discussion section, below. "Intelligent" as the Focal Trait. Analyses of the effects of inclusiveness on the trait "Intelligent" revealed no consistent or reliable pattern (all F's < 1.0) for either the Mean or Variance Scores. However, analyses of the effects of implied stability on "intelligent" Scores suggested a pattern opposite to that predicted: both the Mean and Variance Scores revealed that the Blue Group was higher and the Red Group 33 lower on the trait "intelligent" when implied stability was high (versus low; see Table 8). In addition, the Intelligent Variance Score for the Red Group revealed an (unpredicted) interaction (p. = .094) between inclusiveness and implied stability such that, in high inclusive conditions, the Variability Score was higher for those receiving high (M = 482.7) versus low (M = 219.2) implied stability communications. Study 2 Discussion The results from Study 2 support the hypothesis that the "inclusiveness" of communication affects stereotype formation. However, this was observed primarily with respect to a negative focal trait (aggressiveness). No consistent effect was found on the formation of stereotypic beliefs about a positive trait (intelligence). It is perhaps noteworthy that the effects of inclusiveness, while consistent across both target groups, were generally stronger for the Blue Group. While this could of course be due to chance, it makes a certain degree of sense: remember that the Blue Group was depicted somewhat more ambiguously (e.g., the majority of members are generally average - or a little less than average -- with respect to the focal traits), and, as such, potentially left more room for the influence of communication. This explanation, though plausible, is of course only speculative, and should be taken with caution. The effects of implied stability are much murkier. Changes in beliefs about aggressiveness revealed mean differences consistent with the hypothesis. On the other hand, mean differences in changes in beliefs about intelligence were fully inconsistent with the hypothesized effects of implied stability. In general, there was 34 little support for the hypothesis pertaining to the effects of implied stability on stereotype formation. Somewhat puzzling are the interactions that emerged on both the Similarity Measure and the Distribution Task Scores: In general, the effects of inclusiveness on the ultimate stereotype formed tended to be weaker in high (versus low) implied stability conditions. This implies either that the presence of communications high in implied stability weakens the effect of high inclusiveness communications on stereotypes, or that the combination of low implied stability and high inclusiveness (e.g., "the Red Group hits their wives") is somehow uniquely powerful with respect to stereotype formation. It is possible that, due to the relative infrequency of high inclusive/low implied stability communication in everyday conversation, its presence is perceived as unusual, drawing attention and increasing its impact. These data do not, of course, bear on this speculation. And, indeed, before speculating too wildly, it is worth keeping in mind that no such pattern emerged on any of the more straightforward Sentence Description measures of stereotype content. The reported analyses do not bear directly on the question, "are people high in PNS more influential communicators of stereotypes?" However, one of the predictions that emerges from the present conceptual analysis is that, once forming an initial belief, persons high in PNS should be more difficult to persuade from that belief than low-PNS persons (cf. Kruglanski et al, 1993). Thus, the effects of highly inclusive language may be moderated by a "listener's" PNS, such that persons high in PNS would be less influenced by the presence of high (versus low) inclusiveness. Consistent with this, regression analyses revealed a weak interaction between PNS 35 and inclusiveness on the various Aggressive Change Scores computed from the Sentence Description measure. Visual inspection of the regression lines suggested that, as predicted, persons high in PNS were generally less influenced by highly inclusive language than persons low in PNS. This effect, however, was not statistically significant at conventional levels (all p.'s > .16), and so should be interpreted with appropriate caution. 36 CHAPTER FIVE General Discussion From these two studies alone, no conclusion can be drawn concerning the general hypothesis that persons high in the need for structure are more influential in consensual stereotype formation than those low in the need for structure. It is thus useful to ask: What is the status of this hypothesis? A brief look at some additional data may help shed light on this question. In a different study, 24 participants were placed, via a pre-test, in dyads with one high-PNS participant and one low-PNS participant in each session (this was accomplished by a median split). Participants read the same information as in Studies 1 and 2 reported here. After reading these descriptions of Red and Blue Group members, participants were asked to verbally discuss with each other their impressions of the two groups for three minutes (the experimenter left the room during this time). Sentence description measures of stereotype formation were given to participants both before and after this discussion, and from these were derived Z-score composites for the various dimensions in exactly the same manner as in Study 2. Dyad-level analyses on "aggressive" stereotypes of the Red Group demonstrated that some consensus had emerged through communication: Whereas dyad partners' pre-communication stereotypes were largely uncorrelated (r[12] = .17, P = .299), post-communication stereotypes were positively correlated (r[12] = .46, p. = .067). Further analyses also suggested that, consistent with the hypothesis posited 37 here, high-PNS persons were more influential in the emergence of that consensus. Specifically, when focusing on the degree to which the Red Group was perceived as "aggressive," high-PNS persons' pre-communication stereotypes were strongly positively correlated with low-PNS persons' post-communication stereotypes (r [12]=.73, p. = .003); but low-PNS persons' pre-communication stereotypes were much less positively correlated with high-PNS persons' post-communication stereotypes (r [12]= .38, p. = .114). Thus, low-PNS persons' beliefs about the aggressiveness of the Red Group were very similar to their high-PNS counterparts' initial beliefs about the groups, but not vice versa. This suggests that, given that people bring tentative initial beliefs about groups into a conversation, persons high in PNS are more likely to impact - through potentially non-intentional processes - their low-PNS counterparts to share their initial beliefs about groups. Of course, the results of this third study do not address the role of abstractness of communication in the process of stereotype formation. Rather, they merely suggest that, in spite of the null results obtained in Study 1, it may be premature to discard the general theoretical hypothesis that persons high in PNS are more influential communicators of stereotypes than persons low in PNS. Indeed, it is worth considering that PNS can potentially have a variety of influences at various stages of the social-evolutionary process. For example, communication can be broken down, in an overly simplistic fashion, into processes relevant to transmission (e.g., Joe saying something) and processes relevant to reception (e.g., Joe hearing something). Even if it is the case, as suggested by Study 1, that PNS has little impact on the transmission side of the coin, it may well be the case that PNS impacts 38 the way that relevant communications are processed when received. If, for example, once some tentative belief about a group is formed, persons high in PNS are more likely than their low PNS counterparts to disregard communications from others about that group in favor of their initial impressions, then, even without differences in language abstractness, persons high in PNS could be more influential in the process of consensual belief formation. This result could arise simply because ~ as suggested by results reported by Kruglanski et al (1993) - persons high in PNS would be less inclined to be swayed in their initial beliefs about a given group. Although weak and potentially due to sampling error, the regression results reported for Study 2 suggest that this is the case: Persons high in PNS were less likely than persons low in PNS to be influenced by highly inclusive language. Individual differences aside, Study 2 does bear more broadly on the processes involved in the receiving of communication. Specifically, it provides evidence that the inclusiveness of language used by the transmitter impacts the rate that stereotypes form in the receiver ~ at least in regard to stereotypes about negatively-valenced traits. Indeed, although Study 1 provided no support for the general framework, Study 2 is, by itself, an important step in our understanding of the mechanisms by which stereotypes are spread across cultures. This is potentially no small door to open. For, given that high communication inclusiveness leads to a faster rate of stereotype formation, any individual-level or situational variable which impacts the inclusiveness of group language may also impact the rate of stereotype formation, and, by extension, the contents of stereotypes across cultures. It is left in the hands of future research to discover exactly what those variables may be. 39 Table 1 Study 1: Correlations of Communication Inclusiveness and Communication Implied Stability with Personality Measures Personality Measure PNS Col Soc Per Open Cons Ext Average Total Inclusiveness -.21 .12 .03 -.15 .12 -.02 -.03 Average Total Implied Stability .06 .10 .13 .21 .06 .09 .21 n=39 (for collect., n=38). *p<=.05;**p<=.01;***p<=.001; PNS=Personal Need for Structure; CoNAspects of Identity Questionnaire, Collective Identity; Soc=Aspects of Identity Questionnaire, Social Identity; Open=Openness to Experience; Cons=Conscientiousness; Ext=Extroversion. Table 2 Study 1: Inter-Correlations of Personality Measures Personality Measure PNS Col Soc Per Open Cons Ext Personality Measure: PNS ~ .13 .31 .14 -.48** .35* -.27 Collective Identity .13 — 4 4 * * .53** .17 .14 .12 Social Identity .31 .44** — .52** .19 .22 .02 Personal Identity .14 .53** .52** — .42* .32* .09 Openness to Experience -.48** .17 .19 .42* — .26 .54 Conscientiousness .35* .14 .22 .32* .26 .29 Extroversion -.27 .12 .02 .09 .54*** .29 — n=39. (for collect., n=38). *p<=.05;**p<=.01;***p<=.001; PNS=Personal Need for Structure; Col=Aspects of Identity Questionnaire, Collective Identity; Soc=Aspects of Identity Questionnaire, Social Identity; Open=Openness to Experience; Cons=Conscientiousness; Ext=Extroversion. Table 3 Study 2: Summary of Ersatz Notes by Condition Low Implied Stability/Low Inclusiveness: Phase 2: "...about the Red Group: One member knocked down beginner skiers in the ski lift lines trying to get to the front. Another member yelled at his fellow workers." Phase 3: "...about the Red Group: One member scored a 98 on an exam. Another member was a National Award winner for academics." Phase 4: "...about the Blue Group: One member made C's (& a few B's) in college. Another scored failing grades in a gifted and talented class as a child." Phase 5: "...about the Blue Group: One member let lots of people go ahead of him in a crowded parking lot. Another member played with their daughters." High Implied Stability/Low Inclusiveness: Phase 2. "...about the Red Group: One member is very ambitious. Another member is a real go-getter - he'll go after what he wants, no matter who it hurts." Phase 3: "...about the Red Group: One member is really smart. Another member is a great problem solver - no matter how difficult the problems are." Phase 4: "...about the Blue Group: One member is poorly educated. Another member is a bad thinker." Phase 5: "...about the Blue Group: One member is really calm and laid-back. Another member is very passive and mild-mannered." Low Implied Stability/High Inclusiveness: Phase 2: "...about the Red Group: This group always knocks down beginner skiers in the ski lift lines trying to get to the front. They also yell a lot at their fellow workers." Phase 3: "...about the Red Group: This Group always scores well on exams. They also win National Awards for academics." Phase 4: "...about the Blue Group: This group usually makes C's (& a few B's) in college. They also score failing grades in a gifted and talented class as children." Phase 5: "...about the Blue Group: This group always lets people go ahead of them in a crowded parking lot. They also play with their daughters." High Implied Stability/High Inclusiveness: Phase 2: "...about the Red Group: This group is very ambitious. They are real go-getters - they'll go after what they want, no matter who it hurts." Phase 3: "...about the Red Group: This group is really smart. They are also great problem solvers - no matter how difficult the problems are." Phase 4: "...about the Blue Group: This group is poorly educated. They also are bad thinkers." Phase 5: "...about the Blue Group: This group is really calm and laid-back. They are very passive and mild-mannered." 41 Table 4 Study 2: Index of Dependent Variable Names Time 2 Sentence Description measure: Time 2 Word Content Scores: Red Group Aggressive Word Content: aggressive words describing the Red Group. Blue Group Unaggressive Word Content: unaggressive words describing the Blue Group. Group Differentiation Aggressive Word Content: sum of agg./Red Group words and unagg./Blue Group words minus the sum of unagg./Red Group words and agg./Blue Group words. Red Group Intelligent Word Content: intelligent words describing the Red Group. Blue Group Unintelligent Word Content: unintelligent words describing the Blue Group. Group Differentiation Intelligent Word Content: sum of int./Red Group words and unintVBlue Group words minus the sum of unintVRed Group words and int/Blue Group words. Time 2 First Sentence Scores: Red Group Aggressive First Sentence: aggressive words describing the Red Group - first sentence. Blue Group Unaggressive First Sentence: unaggressive words describing the Blue Group - first sen. Group Differentiation Aggressive First Sentence: sum of agg./Red Group words and unagg./Blue Group words minus the sum of unagg./Red Group words and agg./Blue Group words - first sentence. Red Group Intelligent First Sentence: intelligent words describing the Red Group - first sentence. Blue Group Intelligent First Sentence: unintelligent words describing the Blue Group - first sentence. Group Differentiation Intelligent First Sentence: sum of int./Red Group words and unint./Blue Group words minus the sum of unint./Red Group words and int/Blue Group words - first sentence. Time 2 Inclusiveness Scores: Red Group Aggressive Inclusiveness: aggressive words (Red Group) X avg. inclusiveness of words. Blue Group Unaggressive Inclusiveness: unaggress. words (Blue Group) X avg. inclusiveness of words. Group Differentiation Aggressive First Sentence: sum of agg./Red Group Inclusiveness and unagg./Blue Gr. Inclusiveness minus the sum of unagg./Red Group Inclusiveness and agg./Blue Indus. Red Group Intelligent Inclusiveness: intelligent words (Red Group) X avg. inclusiveness of words. Blue Group Unintelligent Inclusiveness: unintell. words (Blue Group) X avg. inclusiveness of words. Group Differentiation Intelligent First Sentence: sum of int./Red Group Inclusiveness and unintVBlue Gr. Inclusiveness minus the sum of unint./Red Group Inclusiveness and int./Blue Gr. Inclusiveness. Time 2 Z-Score Composites: Red Group Aggressive: Sum of Red Group Agg. Word Content, First Sentence, and Indus. Scores. Blue Group Aggressive: Sum of Blue Gr. Unagg. Word Content, First Sentence, and Indus. Scores. Group Differentiation Aggressive: Sum of Group Differentiation Aggressive W.C., F.S., and Inc. Scores. Red Group Intelligent: Sum of Red Group Int. Word Content, First Sentence, and Indusiveness Scores. Blue Group Intelligent: Sum of Blue Gr. Unint. Word Content, First Sentence, and Inclusiveness Scores. Group Differentiation Intelligent: Sum of Group Differentiation Intelligent W.C., F.S., and Inc. Scores. Similarity Measure (Time 2 only): Red Group Aggressive Similarity: Response on 1-9 scale to how similar Red Group for "aggressive." Red Group Intelligent Similarity: Response on 1-9 scale to how similar Red Group for "intelligent." Red Group General Similarity: Response on 1-9 scale to how similar Red Group is in general. Red Group Similarity Composite: Sum of Red Agg., Red Int., and Red Geneneral Similarity. Blue Group Aggressive Similarity: Response on 1-9 scale to how similar Blue Group for "aggressive." Blue Group Intelligent Similarity: Response on 1-9 scale to how similar Blue Group is for "intelligent." Blue Group General Similarity: Response on 1-9 scale to how similar Blue Group is in general. Blue Group Similarity Composite: Sum of Blue Agg., Blue Int., and Blue General Similarity. 42 Table 4 (Continued) Study 2: Index of Dependent Variable Names Distribution Task (Time 2 only): Mean Scores: Red Group Aggressive Mean: mean (range 0-100) computed from Red Group Aggressive Distrib. Task. Red Group Intelligent Mean: mean (range 0-100) computed from Red Group Intelligent Distrib. Task. Blue Group Aggressive Mean: mean (range 0-100) computed from Blue Group Aggressive Distrib. Task. Blue Group Intelligent Mean: mean (range 0-100) computed from Blue Group Intelligent Distrib. Task. Variance Scores: Red Group Aggressive Variance: variance computed from Red Group Aggressive Distribution Task. Red Group Intelligent Variance: variance computed from Red Group Intelligent Distribution Task. Blue Group Aggressive Variance: variance computed from Blue Group Aggressive Distribution Task. Blue Group Intelligent Variance: variance computed from Blue Group Intelligent Distribution Task. Sentence Description Measure Change (Time 2 - Time 1) Scores: Word Content Change Scores: Red Group Aggressive Word Content Change: Time 2 Red Agg. W. Content - Time 1 Red Agg. W. C. Blue Group Unaggressive Word Content Change: Time 2 Bl. Unagg. W. Con. - Time 1 Bl. Unagg. W. C. Group Differentiation Aggressive Word Content Change: Time 2 Group Differentiation Agg. Word Content - Time 1 Group Differentiation Agg. Word Content. Red Group Intelligent Word Content Change: Time 2 Red Int. W. Content - Time 1 Red Int. W. Content. Blue Group Unintelligent Word Content Change: Time 2 Bl. Unint. W. Con. - Time 1 Bl. Unint. W. Con. Group Differentiation Intelligent Word Content Change: Time 2 Group Differentiation Int. Word Content -Time 1 Group Differentiation Int. Word Content. First Sentence Change Scores: Red Group Aggressive First Sentence Change: Time 2 Red Agg. F. Sent. - Time 1 Red Agg. F. Sent. Blue Group Unaggressive First Sentence Change: Time 2 Bl. Unagg.F. Sent. - Time 1 Bl. Unagg. F.S. Group Differentiation Aggressive First Sentence Change: Time 2 Group Differentiation Agg. First Sentence - Time 1 Group Differentiation Agg. First Sentence. Red Group Intelligent First Sentence Change: Time 2 Red Int. W. Content - Time 1 Red Int. F. Sent. Blue Group Unintelligent First Sentence Change: Time 2 Bl. Unint. W. Con. - Time 1 Bl. Unint. F. Sent. Group Differentiation Intelligent First Sentence Change: Time 2 Group Differentiation Int. First Sentence - Time 1 Group Differentiation Int. First Sentence. Inclusiveness Change Scores: Red Group Aggressive Inclusiveness Change: Time 2 Red Agg. Inclusive. - Time 1 Red Agg. Inclusive. Blue Group Unaggressive Inclusiveness Change: Time 2 Bl. Unagg. Inclusive. - Time 1 Bl. Unagg. Inc. Group Differentiation Aggressive Inclusiveness Change: Time 2 Group Differentiation Agg. Inclusiveness - Time 1 Group Differentiation Agg. Inclusiveness. Red Group Intelligent Inclusiveness Change: Time 2 Red Int. Inclusive. - Time 1 Red Int. Inclusive. Blue Group Unintelligent Inclusiveness Change: Time 2 Bl. Unint. Inclusive. - Time 1 Bl. Unint. Inclusive. Group Differentiation Intelligent Inclusiveness Change: Time 2 Group Differentiation Int. Inclusiveness -Time 1 Group Differentiation Int. Inclusiveness. Change Z-Score Composites: Red Group Aggressive Change: Time 2 Red Group Aggressive - Time 1 Red Group Aggressive. Blue Group Unaggressive Change: Time 2 Blue Group Unaggressive - Time 1 B.G. Unaggressive. Group Differentiation Aggressive Change: Time 2 Group Differentiation Aggressive - Time 1 Group Differentiation Aggressive. Red Group Intelligent Change: Time 2 Red Group Intelligent - Time 1 Red Group Intelligent. Blue Group Unintelligent Change: Time 2 Blue Group Unintelligent - Time 1 Blue Group Unintelligent. Group Differentiation Intelligent Change: Time 2 Group Differentiation Intelligent - Time 1 Group Differentiation Intelligent. 43 Table 5 Study 2: Effect of Communication Inclusiveness Level on Change Scores (Note: Higher Scores Reflect A Greater Rate of Stereotype Formation) Inclusiveness Level Low High P. Dependent Variable Red Group Sentence Descriptions Aggressive Word Content Change 0.43 0.55 .666 Aggressive First Sentence Change -0.12 0.19 .192 Aggressive Inclusiveness Change 1.81 3.64 .089 Agg. Z-Score Composite -0.12 0.12 .176 Intelligent Word Content Change 0.05 0.10 .883 Intelligent First Sentence Change -0.03 0.05 .787 Intelligent Inclusiveness Change 0.77 1.02 .816 Int. Z-Score Composite -0.01 0.03 .854 Blue Group Sentence Descriptions Unaggressive Word Content Change -0.18 0.40 .067 Unaggressive First Sentence Change -0.30 -0.02 .224 Unaggressive Inclusiveness Change -0.73 2.36 .046 Unagg. Z-Score Composite -0.19 0.18 .042 Unintelligent Word Content Change 0.23 0.45 .205 Unintelligent First Sentence Change 0.03 0.24 .114 Unintelligent Inclusiveness Change 0.88 2.43 .069 Unint. Z-Score Composite -0.18 0.17 .089 Group Differentiation Sentence Descriptions Aggressive Word Content Change 0.03 0.93 .119 Aggressive First Sentence Change -0.45 0.12 .091 Aggressive Inclusiveness Change -0.45 5.52 .034 Agg. Z-Score Composite -0.19 0.19 .036 Intelligent Word Content Change 0.18 0.67 .233 Intelligent First Sentence Change -0.03 0.45 .071 Intelligent Inclusiveness Change 1.36 3.86 .186 Int. Z-Score Composite -0.16 0.16 .121 Table 6 Study 2: Effect of Communication Implied Stability Level on Change Scores (Note: Higher Scores Reflect A Greater Rate of Stereotype Formation) Implied Stability Level Low High P. Dependent Variable Red Group Sentence Descriptions Aggressive Word Content Change 0.34 0.63 .348 Aggressive First Sentence Change -0.12 0.27 .045 Aggressive Inclusiveness Change 1.51 3.29 .216 Agg. Z-Score Composite -0.15 0.15 .099 Intelligent Word Content Chang 0.15 0.00 .578 Intelligent First Sentence Change 0.24 -0.22 .015 Intelligent Inclusiveness Change 0.63 1.18 .641 Int. Z-Score Composite 0.09 -0.08 .377 Blue Group Sentence Descriptions Unaggressive Word Content Change 0.02 0.22 .483 Unaggressive First Sentence Change -0.22 -0.10 .558 Unaggressive Inclusiveness Change -0.17 1.88 .164 Unagg. Z-Score Composite -0.09 0.09 .281 Unintelligent Word Content Change 0.39 0.29 .621 Unintelligent First Sentence Change 0.24 0.02 .102 Unintelligent Inclusiveness Change 1.80 1.54 .818 Unint. Z-Score Composite 0.10 -0.10 .389 Group Differentiation Sentence Descriptions Aggressive Word Content Change 0.15 0.83 .224 Aggressive First Sentence Change -0.39 0.07 .159 Aggressive Inclusiveness Change 0.34 4.88 .096 Agg. Z-Score Composite -0.15 0.15 .094 Intelligent Word Content Chang 0.66 0.20 .264 Intelligent First Sentence Change 0.56 -0.12 .008 Intelligent Inclusiveness Change 3.24 2.05 .550 Int. Z-Score Composite 0.17 -0.16 .118 Table 7 Study 2: Effect of Communication Inclusiveness Level and Communication Implied Stability Level Time 2 Similarity Measure Scores Inclusiveness Level Low High Dependent Variable Red Group Similarity Ratings Aggressive 6.45 6.95 .245 Intelligent 6.38 7.21 .030 General 5.30 6.05 .094 Red Similarity Composite 6.04 6.74 .040 Blue Group Similarity Ratings Aggressive 5.00 5.79 .107 Intelligent 5.98 5.71 .500 General 5.23 5.45 .564 Blue Similarity Composite 5.40 5.65 .483 Implied Stability Level Low High B Red Group Similarity Ratings Aggressive 6.76 6.66 .864 Intelligent 7.00 6.61 .351 General 5.90 5.46 .350 Red Similarity Composite 6.55 6.24 .400 Blue Group Similarity Ratings Aggressive 5.29 5.51 .599 Intelligent 6.15 5.54 .156 General 5.39 5.29 .822 Blue Similarity Composite 5.61 5.45 .663 Table 8 Study 2: Effect of Communication Inclusiveness Level and Communication Implied Stability Level Time 2 Distribution Task Scores Inclusiveness Level Low High p Dependent Variable Mean Scores Red Aggressive 64.6 65.6 .772 Red Intelligent 67.3 69.0 .622 Blue Aggressive 43.9 39.0 .123 Blue Intelligent 50.1 48.9 .707 Variance Scores Red Aggressive 323.7 352.8 .522 Red Intelligent 291.1 344.7 .325 Blue Aggressive 395.0 339.1 .260 Blue Intelligent 292.3 301.4 .740 Implied Stability Level Low High p. Mean Scores Red Aggressive 66.8 63.4 .256 Red Intelligent 71.1 65.2 .039 Blue Aggressive 40.4 42.5 .542 Blue Intelligent 46.9 52.1 .058 Variance Scores Red Aggressive 309.0 368.2 .226 Red Intelligent 239.6 397.4 .011 Blue Aggressive 353.1 379.9 .612 Blue Intelligent 255.2 337.6 .026 Table 9 Study 2: Effect of Communication Inclusiveness Level on Time 2 Sentence Description Scores (Note: Higher Scores Reflect More Stereotyping) Inclusiveness Level Low High E Dependent Variable Red Group Sentence Description Aggressive Word Content 1.65 2.05 .160 Aggressive First Sentence 0.40 0.67 .054 Aggressive Inclusiveness 6.23 9.50 .028 Agg. Z-Score Composite -0.21 0.20 .028 Intelligent Word Content 1.35 1.00 .086 Intelligent First Sentence 0.63 0.25 .118 Intelligent Inclusiveness 4.79 4.33 .619 Int. Z-Score Composite 0.16 -0.13 .126 Blue Group Sentence Description Unaggressive Word Content 1.40 1.95 .051 Unaggressive First Sentence 0.38 0.71 .023 Unaggressive Inclusiveness 5.63 8.55 .042 Unagg. Z-Score Composite -0.24 0.23 .010 Unintelligent Word Content 0.55 0.67 .478 Unintelligent First Sentence 0.25 0.31 .658 Unintelligent Inclusiveness 2.23 3.05 .270 Unint. Z-Score Composite -0.09 0.09 .400 Group Differentiation Sentence Description Aggressive Word Content 2.50 3.57 .054 Aggressive First Sentence 0.65 1.31 .006 Aggressive Inclusiveness 10.40. 16.19 .038 Agg. Z-Score Composite -0.26 0.24 .009 Intelligent Word Content 1.50 1.33 .599 Intelligent First Sentence 0.73 0.71 .876 Intelligent Inclusiveness 5.74 5.88 .963 Int. Z-Score Composite 0.03 -0.02 .773 48 Table 10 Study 2: Effect of Communication Implied Stability Level on Time 2 Sentence Description Scores (Note: Higher Scores Reflect More Stereotyping) Dependent Variable Low High Implied Stability Level Red Group Sentence Descriptions Aggressive Word Content 1.96 1.76 .523 Aggressive First Sentence 0.44 0.63 .145 Aggressive Inclusiveness 8.17 7.46 .687 Agg. Z-Score Composite -0.01 0.01 .870 Intelligent Word Content 1.20 1.15 .746 Intelligent First Sentence 0.56 0.46 .449 Intelligent Inclusiveness 4.73 4.38 .695 Int. Z-Score Composite 0.05 -0.03 .642 Blue Group Sentence Descriptions Unaggressive Word Content 1.83 1.54 .330 Unaggressive First Sentence 0.59 0.51 .697 Unaggressive Inclusiveness 7.41 6.83 .752 Unagg. Z-Score Composite 0.07 -0.07 .487 Unintelligent Word Content 0.63 0.59 .789 Unintelligent First Sentence 0.39 0.17 .052 Unintelligent Inclusiveness 2.71 2.59 .912 Unint. Z-Score Composite 0.09 -0.09 .389 Diagnosticity Sentence Descriptions Aggressive Word Content 3.22 2.88 .593 Aggressive First Sentence 0.95 1.02 .653 Aggressive Inclusiveness 13.85 12.88 .799 Agg. Z-Score Composite 0.03 -0.03 .891 Intelligent Word Content 1.59 1.24 .317 Intelligent First Sentence 0.93 0.51 .038 Intelligent Inclusiveness 6.68 4.93 .266 Int. Z-Score Composite 0.15 -0.14 .144 Table 11 Study 2: Inter-Correlations of Personality Measures PNS Col Personality Measure: PNS — .31 Collective Identity .31** — Social Identity .27* .28 Personal Identity .09 .44 Openness to Experience -.30** -.11 Conscientiousness .13 .03 Extroversion -.21 -.09 Personality Measure Soc Per Open Cons Ext .27* .09 -.30** .13 -.21 .28* 44*** -.11 .03 -.09 — .19 .04 -.01 .12 .19 - .17 .30** -.04 .04 .17 ~ .30** .51*** .01 .30** .30** — .27* .12 -.04 .51*** .27* — n=82. *p<=.05;**p<=.01;***p<=.001; PNS=Personal Need for Structure; Col=Aspects of Identity Questionnaire, Collective Identity; Soc=Aspects of Identity Questionnaire, Social Identity; Open=Openness to Experience; Cons=Conscientiousness; Ext=Extroversion. 50 FOOTNOTES 1. Both after Phase 1 and again after Phase 5, participants completed the Sentence Descriptions measure. (They also completed two additional measures after Phase 5.) The primary purpose of these measures was to serve as a crude "baseline" comparison for the results obtained in Study 2 (e.g., what stereotypes do participants develop from the stimulus cards when no additional communication is given to them). As this purpose is not generally relevant to the main hypotheses tested here (and because it is not a true "baseline," because participants were surely influenced by the process of writing notes), analyses of these measures are not reported. However, they are available from Lucian G. Conway, III on request. 2 . For the Red Group at Time 1: Word Content Aggressive Agreement = 100%, Word Content Intelligent Agreement = 100%, Inclusiveness Correlation Coefficient = .86. For the Blue Group at Time 1: Word Content Aggressive Agreement = 86%, Word Content Intelligent Agreement = 86%, Inclusiveness Correlation Coefficient = .96. For the Red Group at Time 2: Word Content Aggressive Agreement = 82%, Word Content Intelligent Agreement = 70%, Inclusiveness Correlation Coefficient = .97. For the Blue Group at Time 2: Word Content Aggressive Agreement = 80%, Word Content Intelligent Agreement = 80%, Inclusiveness Correlation Coefficient = .91. 3 .All reported dimensions for the Change Scores were also analyzed at Time 2 only. Looking at the Time 2 only scores is analogous to asking the question "What were the 'final' stereotypes of participants?" These analyses, not reported here because they are less exact (e.g., they do not take into account participants pre-communication reception belief), mirror the results reported for the change scores. Tables 9 and 10 report this information. 51 BIBLIOGRAPHY Ashmore, G. W., & Del Boca, F. L. (1981). Conceptual approaches to stereotypes and stereotyping. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 1-35). Hillsdale, NJ: Lawrence Erlbaum. Boster, J. S. (1991). The information economy model applied to biological similarity judgment. In L. B. Resnick, J. M. Levine, & S. D. Teasley (Eds.), Perspectives on socially shared cognition (pp. 203-225). Washington, DC: American Psychological Association. Brewer, M. B., & Kramer, R. M. (1985). The psychology of intergroup behavior. Annual Review of Psychology. 36, 219-243. Brown, R. (1978). Divided we fall: An analysis of relations between sections of a factory workforce. In H. Tajfel (Ed.), Differentiation Between Social Groups. London: Academic Press, pp. 395-429. Campbell, D. T. (1974). Evolutionary epistemology. In P. A. Schlipp (Ed.), The philosophy of Karl R. Popper (pp. 413-463). LaSalle, IL: Open Court. Capozza, D., Bonaldo, E., & DiMaggio, A. (1982). Problems of identity and social conflict: Research on ethnic groups in Italy. In H. Tajfel (Ed.). Social Identity and Intergroup Relations. Cambridge: Cambridge University Press, pp. 299-334. Cheek, J. M., & Tropp, L. R. (1994). The aspects of identity questionnaire: History and bibliography. Unpublished manuscript. Dawkins, R. (1989). The selfish oene (New Edition). Oxford, UK: Oxford University Press. (Original edition published 1976.) Ford, T. E. & Kruglanski, A. W. (1995). Effects of epistemic motivations on the use of accessible constructs in social judgment. Personality and Social Psychology Bulletin. 21, 950-962. Ford, T. E., & Stangor, C. (1992). The role of diagnosticity in stereotype formation: Perceiving group means and variances. Journal of Personality and Social Psychology. 63, 356-367. Harasty, A. S. (1997). The interpersonal nature of social stereotypes: Differential discussion patterns about in-groups and out-groups. Personality and Social Psychology Bulletin. 23, 270-284. 52 Hardin, C , & Higgins, E. T. (1996). Shared reality: How social verification makes the subjective objective. In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook of motivation and cognition: Foundations of social behavior. New York: Guilford. Haslam, S. A. (1996). Stereotyping and social influence: Foundations of stereotype consensus. In R. Spears, P. J. Oakes, N. Ellemers, & S. A. Haslam (Eds.), The Social Psychology of Stereotyping and Group Life. Oxford: Blackwell, pp. 119-143. Judd, C. M., & Park, B. (1988). Out-group homogeneity: Judgments of variability at the individual and group levels. Journal of Personality and Social Psychology. 54,778-788. Kruglanski, A. W. (1989). Lay epistemics and human knowledge: Cognitive and motivational bases. New York: Plenum. Kruglanski, A. W., & Webster, D. M. (1996). Motivated closing of the mind: "Seizing" and "Freezing." Journal of Personality and Social Psychology. 103. 263-283. Kruglanski, A. W., Webster, D. M., & Klem, A. (1993). Motivated resistance and openness to persuasion in the presence or absence of prior information. Journal of Personality and Social Psychology. 65, 861-865. Latane, B. (1996). Dynamic social impact: The creation of culture by communication. Journal of communication. 46(4), 13-25. Linville, P. W., Fischer, G. W., & Salovey, P. (1989). Perceived distributions of characteristics of in-group and out-group members: Empirical evidence and a computer simulation. Journal of Personality and Social Psychology. 57. 165-188. McCrae, R. R., & Costa, P. T. (1983). Updating Norman's "adequate taxonomy": Intelligence and personality dimensions in natural language and in questionnaires. Journal of Personality and Social Psychology. 49, 710-721. Messick, D. M., & Mackie, D. M. (1989). Intergroup relations. Annual Review of Psychology. 40,45-81. Moskowitz, G. B. (1993). Individual differences in social categorization: The influence of personal need for structure on spontaneous trait inferences. Journal of Personality and Social Psychology. 53. 431-444. Neuberg, S., & Newsom, J. T. (1993). Personal need for structure: Individual differences in the desire for simple structure. Journal of Personality and Social Psychology. 61, 113-131. 53 Park, B., & Judd, C. M. (1990). Measures and models of perceived group variability. Journal of Personality and Social Psychology. 59, 173-191. Rubini, M., & Kruglanski, A. W. (1997). Brief encounters ending in estrangement: Motivated language use and interpersonal rapport in the question-answer paradigm. Journal of Personality and Social Psychology. 72, 1047-1060. Schaller, M., & Latane, B. (1996). Dynamic social impact and the evolution of culture: A natural history of social representations. Journal of communication. 48, 39-45. Schaller, M., & Conway. L. G. (1997). Communication goals influence the contents of consensually-shared group stereotypes: Support for a psychosocial-evolutionary hypothesis. Manuscript under review. Schaller, M., Boyd, C , Yohannes, J., & O'Brien, M. (1995). The prejudiced personality revisited: Personal need for structure and formation of erroneous group stereotypes. Journal of Personality and Social Psychology. 68, 544-555. Schneider, D. J. (1996). Modern stereotype research: Unfinished business. In C. N. MacRae, C. Stangor, & M Hewstone (Eds.), Foundations of stereotypes and stereotyping (pp. 419-453). New York: Guilford. Semin, G. R., & Fiedler, K. (1988). The cognitive functions of linguistic categories in describing persons: Social cognition and language. Journal of Personality and Social Psychology. 54, 558-568. Semin, G. R., & Fiedler, K. (1991). The linguistic category model: Its bases, applications and range. European Volume of Social Psychology, v.2, 1-30. Stangor, O , & Schaller, M. (1996). Stereotypes as individual and collective representations. In C. N. MacRae, M. Hewstone, & C. Stangor (Eds.), Foundations of stereotypes and stereotyping (pp. 3-40). New York: Guilford. Stangor, C. & Jost, J. T. (1997). Individual, group, and system levels of analysis and their relevance for stereotyping and intergroup relations. In R. Spears, P. Oakes, N. Ellemers, & S. A. Haslam (Eds.), The social psychology of stereotyping and group life. Oxford: Basil Blackwell. Stangor, C , Sechrist, G., Jost, J. T., & Bar-Tal, D. (1996). Diaonosticitv and consensualitv in stereotype change. Unpublished data, University of Maryland. 54 Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology. 69,797-811. Tajfel, H. (1981). Human Groups and Social Categories. Cambridge: Cambridge University Press. Tajfel, H. (1982). Social psychology of intergroup relations. Annual Review of Psychology. 33. 1-39. Thompson, E. P., Roman, R. J., Moscovitz, G. B., Chaiken, S., & Bargh, J. A. (1994). Accuracy motivation attenuates covert priming: The systematic reprocessing of social information. Journal of Personality and Social Psychology. 66, 474-489. Thompson, M. M., Naccarato, M. E., & Parker, K. E. (1989, June). Assessing cognitive need: The development of the Personal Need for Structure and Personal Fear of Invalidity scales. Paper presented at the annual meeting of the Canadian Psychological Association, Halifax, Nova Scotia. Thompson, M. M., Naccarato, M. E., & Parker, K. E. (1992). Measuring cognitive needs: The development and validation of the Personal Need for Structure (PNS) and Personal Fear of Invalidity (PFD measures. Unpublished manuscript, University of Waterloo, Canada. Webster, D. M. (1993). Motivated augmentation and reduction of the overattribution bias. Journal of Personality and Social Psychology. 65, 261-271. Webster, D. M., & Kruglanski, A. W. (1994). Individual differences in need for cognitive closure. Journal of Personality and Social Psychology. 67, 1049-1062. Webster, D. M., Kruglanski, A. W., & Pattison, D. A. (1997). Motivated language use in intergroup contexts: Need-for-closure effects on the linguistic intergroup bias. Journal of Personality and Social Psychology. 72, 1122-1131. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0088436/manifest

Comment

Related Items