Citation
Judging Members of African- and Asian-American Groups: A Test of the Role of Relevant and Irrelevant Affect in Intergroup Perceptions

Material Information

Title:
Judging Members of African- and Asian-American Groups: A Test of the Role of Relevant and Irrelevant Affect in Intergroup Perceptions
Creator:
MCNATT, PENNY S. ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
African Americans ( jstor )
Emotional stability ( jstor )
Emotional states ( jstor )
Judgment ( jstor )
Personality psychology ( jstor )
Personality traits ( jstor )
Social groups ( jstor )
Social psychology ( jstor )
Stereotypes ( jstor )
We they distinction ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Penny S. McNatt. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/31/2006
Resource Identifier:
436098748 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

JUDGING MEMBERS OF AFRICANAND ASIAN-AMERICAN GROUPS: A TEST OF THE ROLE OF RELEVANT AND IRRELEVANT AFFECT IN INTERGROUP PERCEPTIONS By PENNY S. MCNATT A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Penny S. McNatt

PAGE 3

ACKNOWLEDGEMENTS First, I would like to thank my advisor, Dolores Albarracn, for her support and mentorship over the last few years. I also owe thanks to the other members of my committee, James Shepperd, Lisa M. Brown and Joel Cohen, for all of their helpful contributions to this dissertation. I would also like to thank my colleagues and fellow graduate students for their support over the years. First, I want to thank the original members of the Perkins crew: Jodi Grace, Amy Mitchell, and Audra Lifka. This experience would simply not have been the same without you. Additionally, I would like to thank Beth Pontari, Cynthia Klein, Nancy Frye, Erika Koch, Pat Carroll, Tarcan and Ece Kumkale, Lisa Neff, Meredith Terry, Marisa Miller, Will Hart, Allison Earl, Hong Li, Kate Dockery, and Harry Wallace. Finally, I would like to thank my parents, Ron and Song Ye McNatt, who taught me to dream big and work hard. Much thanks and love go to my brother JR, Laurel, Sarah and Micah. Last, but certainly not least, I would like to thank my husband, Dalvin Devine for his endless encouragement, support and infinite faith in me. iii

PAGE 4

TABLE OF CONTENTS page ACKNOWLEDGMENTS. iii ABSTRACT.. ix CHAPTER 1 INTRODUCTION. 1 Effects of Affect in Past Research about Affect and Intergroup Judgments and Decisions... 3 Limitations of This Past Research. 5 Effects of Relevant and Irrelevant Affect as a Function of Low, Moderate and High Cognitive Abilities. 6 2 REVIEW OF LITERATURE........ 16 Definition of Affect.... 16 Literatures Relevant to the Review.... 19 Findings in the Reviewed Literature.. 39 Theoretical Interpretation of the Review Findings 43 Limitations of Previous Studies. 62 3 OVERVIEW...... 92 Overview of the Present Study.. 92 Review of the Hypotheses. 95 4 METHOD.. 102 Participants..... 102 Design ....... 102 Materials 102 Dependent Measures. 105 Procedure... 107 5 RESULTS.. 110 Manipulation Checks..... 110 Order Effects.. 110 iv

PAGE 5

Descriptive Statistics.. 111 Hypothesis Tests.... 111 6 DISCUSSION 135 Study Rationale and Summary of Results. 135 Directions for Future Research.. 140 Conclusion. 144 APPENDIX A INSTRUCTIONS FOR ANGRY AFFECT INDUCTION... 146 B INSTRUCTIONS FOR HAPPY AFFECT INDUCTION. 147 C PHOTOGRAPH OF ASIAN-AMERICAN MALE CHARACTER/ NEIGHBOR... 148 D PHOTOGRAPH OF AFRICAN-AMERICAN MALE CHARACTER/ NEIGHBOR... 149 E SOCIAL JUDGMENT TASK INSTRUCTIONS. 150 F ASIAN-AMERICAN MALE’S APPLICATION. 152 G AFRICAN-AMERICAN MALE’S APPLICATION 154 H MEMO FOR NOTES ABOUT EACH APPLICANT... 155 I LOW DISTRACTION MANIPULATION (APPLICANT 1).. 156 J MODERATE DISTRACTION MANIPULATION (APPLICANT 1).... 157 K HIGH DISTRACTION MANIPULATION (APPLICANT 1). 158 L LOW DISTRACTION MANIPULATION (APPLICANT 2).. 159 M MODERATE DISTRACTION MANIPULATION (APPLICANT 2). 160 N HIGH DISTRACTION MANIPULATION (APPLICANT 2). 161 O PROBE FOR SUSPICION.... 162 P RATING OF THE WRITING TASK 163 Q STEREOTYPE ACTIVATION MEASURE. 164 v

PAGE 6

R LIKELIHOOD JUDGMENTS OF APPLICANTS (ALL TRAITS). 165 S STEREOTYPE APPLICATION TRAITS.... 170 T POSITIVE AND NEGATIVE NON-STEREOTYPIC TRAITS (PRIMING EFFECTS) . 171 U ATTITUDES TOWARD APPLICANTS.. 172 V INSTRUCTIONS FOR BOTH STUDIES 173 W FORM TO RECORD STORY/AFFECT INDUCTION... 174 X SHORT INSTRUCTIONS FOR SOCIAL JUDGMENT TASK.. 175 Y MODERN RACISM SCALE 176 REFERENCES.. 177 BIOGRAPHICAL SKETCH. 186 vi

PAGE 7

LIST OF TABLES Table page 2.1 Description of Studies in the Review. 64 2.2 Description of Video Mood Inductions..... 88 2.3 Description of Musical Mood Inductions.. 91 3.1 Predictions for Stereotype Activation (Hypothesis 1)... 97 3.2 Predictions for Stereotype Application (Hypothesis 2). 98 3.3 Predictions for Priming Effects (Hypothesis 3). 99 3.4 Predictions for Effects of Mood as Information on Attitudes (Hypothesis 4 and 5).. 100 3.5 Summary of Mechanisms for Influence of Mood on Stereotyping... 101 5.1 Descriptive Statistics for Measures 121 5.2 Means for Activation of African-American Stereotypes............... 122 5.3 Means for Activation of Asian-American Stereotypes.. 123 5.4 Means for Stereotype Activation (Hypothesis 1)... 124 5.5 Means for Application of African-American Stereotypes. 125 5.6 Means for Application of Asian-American Stereotype..... 126 5.7 Means for Stereotype Application (Hypothesis 2). 127 5.8 Means for Endorsement of Non-Stereotypic Traits for the African-American Candidate (Priming Effects).... 128 vii

PAGE 8

5.9 Means for Endorsement of Non-Stereotypic Traits for the Asian-American Candidate (Priming Effects)... 129 5.10 Means for Priming Effects for African-American Candidate (Hypothesis 3).... 130 5.11 Means for Priming Effects for Asian-American Candidate (Hypothesis 3)................................................................................................ 131 5.12 Means for Attitudes toward African-American Candidate 132 5.13 Means for Attitudes toward Asian-American Candidate... 133 5.14 Means for Effects of Mood as Information on Attitudes (Hypothesis 4 and 5).. 134 viii

PAGE 9

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy JUDGING MEMBERS OF AFRICANAND ASIAN-AMERICAN GROUPS: A TEST OF THE ROLE OF RELEVANT AND IRRELEVANT AFFECT IN INTERGROUP PERCEPTIONS By Penny S. McNatt May, 2005 Chairperson: Dolores Albarracn Major Department: Psychology The purpose of this dissertation was to explore the ways in which a person’s affective states influence intergroup judgments. Specifically, the goal was to test the influence of several key dimensions of mood on stereotype activation and application, including valence (happy versus angry moods), relevance (relevant versus irrelevant moods), and cognitive ability (low, moderate or high). Participants (N = 79) were induced to be in one of two affective states (happy or angry) elicited by stimuli that were relevant or irrelevant to judging members of different ethnic groups. Following the affect manipulation, participants evaluated an Asian-American and African-American applicant and made a series of judgments about each candidate. Results indicated that controlling for individual differences in prejudice, participants who viewed a (relevant) photograph of an African-American person in relation to the affect manipulation had greater application of African-American stereotypes than those who viewed an (irrelevant) photograph of an Asian-American person in relation to the affect manipulation. With and ix

PAGE 10

without controlling for individual differences in prejudice, participants who saw the relevant photo of an African-American person in relation to the affect manipulation also demonstrated greater application of negative African-American stereotypes and greater endorsement of non-stereotypic negative traits about the African-American candidate. Finally, participants who evaluated the African-American applicant immediately after evaluating the Asian-American applicant endorsed significantly more negative stereotypes about African Americans than participants who evaluated the African-American applicant before evaluating the Asian-American applicant. Future research may explore this apparent negativity bias for African Americans wherein African Americans are hurt by the comparison with an Asian American, although Asian Americans are neither hurt nor helped. x

PAGE 11

CHAPTER 1 INTRODUCTION According to early theories of prejudice, individuals who experience negative emotions are more likely to discriminate against outgroup members than individuals who experience positive emotions (Allport, 1954). For instance, people should be more inclined to use racial slurs when they are frustrated or perceive that the actions of the target are objectionable. The frustration-aggression hypothesis proposed that prejudice was a side effect of channeled aggression that resulted from deprivation (Dollard, Doob, Miller, Mowrer & Sears, 1939). Further, the scapegoat theory of prejudice posited that negative affect toward members of outgroups was the result of displaced hostility and that negative feelings lead to unfavorable evaluations of members of stereotyped groups because these evaluations serve the function of negative state relief (Allport, 1954; Bettelheim & Janowitz, 1950; Hovland & Sears, 1940). Recent perspectives have increasingly examined the different roles of positive and negative emotions in stereotyping and prejudice and have demonstrated that positive mood states can also lead to increased stereotyping and discrimination. Researchers have argued that positive mood states heighten stereotyping because they increase people’s tendencies to process information in a heuristic fashion (Mackie & Worth, 1989). Yet some evidence indicates that these effects may stem from differences in the motivation and ability to think about the target being judged that are associated with various types of emotions, rather than being a direct result of mood valence (Bodenhausen, Mussweiler, 1

PAGE 12

2 Gabriel & Moreno, 2001; Kim & Baron, 1988). One goal of this review is to assess which perspective has received the most support in the literature. In addition to its impact on stereotyping, mood influence other types of intergroup judgments (e.g., general attitudes toward members of a given social group). For instance, people may use their mood as information for deciding whether or not they like French Canadians (Haddock, Zanna & Esses, 1994). In this situation, those who experience negative moods may generate more prejudiced judgments than those who experience positive moods. Furthermore, mood may influence recall of prior knowledge about a given social group (Blaney, 1986; Singer & Salovey, 1988). If this is the case, people who experience negative moods may recall more negative material and end up with a more prejudiced judgment than those who experience positive mood. That is, affect can influence judgments of outgroup targets by directly informing these judgments (e.g., mood as information) or by providing retrieval cues for affect-congruent material stored in memory. Although the fact that affect exerts various influences on intergroup judgments is fairly well established (see Bodenhausen et al., 2001), the conditions that elicit each type of impact are not. Recent research has attempted to clarify the interplay between affective and cognitive processes in the context of stereotyping and prejudice, as well as other mechanisms that directly influence judgments and decisions concerning members of some groups. However, researchers have made little progress in integrating inconsistent findings in the literature or explaining the conditions under which different mechanisms (e.g., mood priming, mood as information) explain the influence of mood on stereotyping. The aims of the present review were to comprehensively survey and

PAGE 13

3 qualitatively synthesize a complex literature on the influence of affect on stereotyping and prejudice. Further, the proposed research attempts to help to clarify some of the ambiguities of past research. In particular, the goals of the current study were to test the influence of several key dimensions of affect, including valence, relevance, and arousal (through its impact on cognitive ability). Effects of Affect in Past Research about Affect and Intergroup Judgments and Decisions The valence of affect indicates the direction (positive or negative) of the affective state. Much research on positive mood states (e.g., happiness) suggests that positive mood is associated with an increased reliance on heuristic processing and heightened stereotyping (Abele et al., 1998; Bless, Schwarz, & Wieland, 1996; Bodenhausen, Kramer & Susser, 1994; Forgas & Fiedler, 1996; Gold, 2002; Lambert et al., 1997; Park & Banaji, 2000; Stepper & Strack, 1993; Stroessner & Mackie, 1992). Other research demonstrates that positive mood can lead to more inclusive judgments of outgroup members and reductions in ingroup favoritism (Burgess, 1993; Dovidio et al., 1995). Results regarding the impact of negative affect on intergroup judgments are also inconsistent. For example, sadness is generally associated with an increase in systematic processing and a reduction in reliance on stereotypes (Bodenhausen et al., 2001; Lambert et al., 1997). However, anger is associated with an increased reliance on stereotypes (Bodenhausen, Sheppard & Kramer, 1994). Positive and negative mood states also influence the mood congruence of judgments via mood priming (Bower, 1981; Isen, 1987) or mood as information (Schwarz & Clore, 1983; Schwarz & Clore, 1988). Research on mood congruence has demonstrated that positive mood leads to more positive judgments and negative moods

PAGE 14

4 lead to more negative judgments. (Abele et al., 1998; Esses & Zanna, 1995; Forgas, 1992b; Forgas, 1992c; Forgas & Moylan, 1991). For example, negative mood leads to negative judgments of specific ethnic minority groups (see Esses & Zanna, 1995). Another critical dimension of affect is arousal level or intensity, which varies across different affective states. For example, among positive mood states, happiness tends to be associated with higher levels of arousal, and contentment tends to be associated with lower levels of arousal (Worth & Mackie, 1987). In contrast, among negative mood states, anger tends to be associated with higher levels of arousal and sadness tends to be associated with lower levels of arousal. In the past, research has not supported that low and high arousal positive mood states differ in their impact on stereotyping (Bodenhausen, Kramer & Susser, 1994). However, among negative mood states, high arousal moods are associated with increased stereotyping and low arousal moods are associated with increased systematic processing (Bodenhausen, Sheppard & Kramer, 1994). Finally, relevance is another key dimension of affect, particularly in the context of stereotyping and prejudice. Intergroup judgments may be influenced by irrelevant or relevant affect. Irrelevant (or incidental) affect is not directly related to the intergroup context (Bodenhausen, 1993). Research has demonstrated that affect that is irrelevant to a judgment context can nonetheless affect those judgments (Isen, 1987; Schwarz & Clore, 1988; Wyer & Srull, 1989). For example, positive affect induced by giving participants a candy bar (Dovidio et al., 1995) can influence subsequent judgments of outgroup members even though the candy provision is irrelevant to the judgment context. Whether these effects are due to mood as information or to priming, however, is not entirely clear.

PAGE 15

5 Relevant (or integral) affect is related to the social group to which an individuals responds, and can be either episodic or chronic (Bodenhausen et al., 2001). First, episodic relevant affect refers to affective states that are situationally experienced in intergroup situations, and can be elicited by exposure to members of a group, specifically the group to which the judgment target belongs. For example, positive affect that results from observing the interactions of a gay person in one context can influence subsequent judgments of a different gay person in a different context (Sheppard, 1996). In contrast, chronic relevant affect refers to affect that is chronically associated with members of a group, and is considered a subject variable and cannot be induced (Bodenhausen et al., 2001). For many stereotyped groups, chronic relevant affect is often negative because negative emotions (e.g., fear, anxiety) are frequently associated with members of these groups (Bodenhausen, 1993). Limitations of This Past Research The main limitation of past findings of affect is that they have not provided a clear conceptualization concerning when different effects will occur and how to detect them. For instance, the effects of affect on stereotyping have been inconsistent. Without a clear exploration of the level of ability people have when they make intergroup judgments, it is almost impossible to predict whether affect will increase or decrease stereotyping. As explained presently, the influence of affect on stereotyping is likely to depend on whether low, moderate, or high cognitive ability is necessary to elicit stereotypes. These levels of ability are important because they can precisely model the effects of different levels of arousal of affective states.

PAGE 16

6 As another example, it seems plausible that relevant affect, which is directly linked to a social group, may stimulate the recall of different features of the stereotype of that group depending on its valence. Positive moods may elicit recollections of the musical talents of African Americans, whereas negative moods may elicit associations with illicit behavior. Moreover, both of these effects –which may be less apparent for irrelevant affect—may only occur when individuals have the ability to recall specific aspects of the stereotype. Consequently, without a manipulation of relevant and irrelevant affect and a manipulation of ability level, a determination of these priming driven phenomena is impossible. Effects of Relevant and Irrelevant Affect as a Function of Low, Moderate, and High Cognitive Ability Influences of Affect on Activation and Application of Stereotypes The influences of mood arousal on stereotyping are presumably due to mediating influences on cognitive ability and motivation to think about a target person, which can in turn affect the activation and application of stereotypes. In this regard, there are two perspectives on stereotyping. One implies that the level of distraction of a situation, such as that elicited by the level of arousal of affective states, produces a linear increase in stereotypes. The other implies that ability level should have a nonlinear effect on stereotypes. Specifically, individuals may apply stereotypes when ability is moderate but not when ability is either low or high. Linear effects of cognitive ability. According to Devine (1989), encountering a member of the stereotyped group is sufficient to automatically activate stereotypes about the group. Consequently, people’s ability and motivation to think about a given group

PAGE 17

7 should have no effect on stereotype activation. However, Devine (1989) argues that the stereotype application is a deliberate, controlled type of process that is thus moderated by ability and motivation. With sufficient motivation and ability, people are able to avoid applying the stereotype in a given context. People may be motivated to not apply the stereotype for multiple reasons. First, an individual may not want to appear prejudiced given that contemporary society tends to disapprove of the overt appearance of prejudice (Bodenhausen, Macrae, & Milne, 1998). Second, individuals may have egalitarian values that motivate them to be conscious of situations in which stereotypes are activated and to correct for them whenever possible (Moskowitz, Salomon, & Taylor, 2000). For example, Devine (1989) demonstrated that participants who are low in prejudice do not necessarily have personal beliefs that overlap with the stereotypes, despite their knowledge of these stereotypes. Therefore, low prejudice individuals are apparently more able to inhibit activated stereotype-congruent cognitions than high prejudice individuals. Essentially, even though the stereotype has been activated among all participants, those who are less prejudiced are able to correct for the stereotype, whereas those who are more prejudiced apply the stereotype. Nonlinear effects of cognitive ability. The second perspective about the effects of cognitive ability (and thus of affect-associated arousal) on stereotyping is that cognitive ability is necessary for both stereotype activation and stereotype application. This view asserts that both stereotype activation and application are deliberate processes and therefore demand cognitive capacity. Consistent with this view, Gilbert and Hixon (1991) demonstrated that cognitive busyness can actually prevent stereotype activation when a person encounters a member of the stereotyped group.

PAGE 18

8 To demonstrate that people need ability both to activate a stereotype and to avoid applying it, Gilbert and Hixon (1991) had participants complete a word completion task. The words were consistent or inconsistent with stereotypes of Asian Americans, and the experimenter was Asian-American or European-American. Half the participants repeated digits during the word completion task (and were cognitively busy during activation), whereas half the participants were not required to repeat digits (and were not cognitively busy during activation). Following the word completion task, participants listened to a recording that described the daily life events of the experimenter and were then asked to form an impression of the experimenter. Half the participants engaged in a clicking task whenever a letter appeared on the screen while listening to the recorder (and were cognitively busy during application), whereas the other half did not perform the clicking task (and were not cognitively busy during application). Gilbert and Hixon (1991) found that only participants who were not busy (distracted) during stereotype activation, but were busy (distracted) during stereotype application used the stereotype of Asian Americans as a basis for impressions of the experimenter. Participants who were distracted during the activation phase did not stereotype, nor did people who could after the activation phase (who had the ability to avoid stereotyping). Hypothesis 1. Consistent with Gilbert and Hixon (1991), I predict that high distraction will prevent stereotype activation, (in contrast with Devine’s (1989) prediction that activation occurs when a person merely encounters a member of a stereotyped group). Thus participants in high distraction conditions are expected to have lower scores on the measure of stereotype activation than participants in the moderate or low distraction conditions.

PAGE 19

9 Hypothesis 2 . Further, consistent with Gilbert and Hixon (1991), both high and low levels of ability may prevent the emergence of stereotypes. Thus, participants in high and low distraction conditions are expected to have lower stereotype scores than participants in the moderate distraction conditions. This prediction is in contrast with the pattern predicted by Devine (1989). The evidence about linearity and nonlinearity. There is evidence that high arousal moods lead to increased levels of stereotyping. Most of this research has been conducted by examining positive versus neutral mood, with positive moods being higher in arousal than negative moods. Worth and Mackie (1987) demonstrated that positive moods (e.g., elation) are associated with reduced levels of systematic processing and a greater tendency to be influenced by heuristic cues, as compared to control participants in a neutral mood. For example, people who are in a happy mood are more likely to rely on stereotypes when making a judgment about the guilt of an African-American person in an assault and battery case (e.g., more likely to rely on the stereotype that African Americans are aggressive). In contrast, a sad person is more likely to consider individuating evidence to make a judgment about whether the African-American person is guilty of the assault and battery charge. The association between high arousal affective states and stereotyping has been confirmed in other contexts as well. First, Blessum, Lord, and Sia (1998) found that happy people (who are generally considered to be in a higher arousal mood state) are less likely than their control counterparts to distinguish gay targets based on their stereotypicality. Thus, they viewed even atypical groups members as being typical. Secondly, Park and Banaji (2000) found that relative to participants in a neutral mood,

PAGE 20

10 happy people are less likely to discriminate accurately among different group members of a stereotyped group. Instead they set a lower threshold for drawing stereotypic conclusions about group members, and are more likely to incorrectly recall that specific group members possess stereotypic traits. Furthermore, Forgas and Fiedler (1996) demonstrated that positive moods increase reliance on a simple ingroup favoritism heuristic, as long as personal relevance of the group is low. Thus, a number of studies have demonstrated that positive moods (e.g., happiness) that are considered to be higher in arousal are associated with an increased reliance on heuristic processing and stereotyping. 1 Importantly, Bodenhausen and Kramer (1990) generalized the findings about happiness to other high-arousal emotions with negative valence, demonstrating that the issue may be arousal level (high versus low) rather than simply valence (positive versus negative). They found that participants who experienced high arousal emotions (e.g., happiness and anger) rated stereotypic defendants (e.g., Hispanic student for assault) as more likely to be guilty than did participants in the relatively low arousal, neutral mood conditions. Further, some research has also shown that low arousal emotions (e.g., sadness) promote more careful, systematic analysis of situations than high arousal emotions (Schwarz, 1990). For instance, Lambert et al. (1997) reported that sad 1 However, in one study, Bodenhausen, Kramer, and Susser (1994: study 3) induced positive high arousal (happy-energetic) and low arousal (happy-calm) affect in order to test whether physiological arousal was responsible for the disruption of systematic processing in positive mood states, but this hypothesis was not supported. Both positive mood states led to equivalent levels of stereotyping targets. This study suggests that something other than arousal may be underlying positive moods association with increased heuristic processing.

PAGE 21

11 participants had a greater tendency to correct for stereotypes than participants in a neutral mood. These results suggest that the extensive body of research suggesting that mood valence, specifically positive mood, is primarily responsible for increased reliance on stereotypes may be overlooking other important dimensions of affect (i.e., arousal level). Although research on arousal generally suggests that higher levels of arousal lead to an increased reliance on heuristic processing and stereotyping, the two studies in the review that addressed mood arousal level (Bodenhausen, Kramer & Susser, 1994; Bodenhausen, Sheppard & Kramer, 1994) provided conflicting results. In the findings by Bodenhausen, Sheppard and Kramer (1994), anger (high arousal mood) was associated with an increased reliance on stereotypes, relative to sadness (low arousal mood). However, Bodenhausen, Kramer and Susser (1994: study 3) found that high arousal (happy-energetic) and low arousal (happy-calm) positive moods led to equivalent levels of stereotyping. Importantly, one reason why the evidence may be equivocal is that curvilinear patterns are likely to be invisible unless one manipulates level of arousal over several levels. This requirement is exceedingly difficult to meet, and thus the question has remained unanswered for many years. In this thesis, I will test the potential influence of arousal by manipulating levels of cognitive ability.

PAGE 22

12 Influence of Affect on Mood Congruent Recall and Interpretation Affect priming models suggest that mood can indirectly influences a person’s cognitive processing in a way that is congruent with their mood state by activating related cognitive categories (Bower, 1981; Isen, 1987). Mood congruent judgments can be facilitated in several ways. First, people may recall, attend to, and interpret information they encounter in a way that is consistent with their mood (Bower, 1981; Isen, 1984). For example, in a study by Forgas (1992c), participants in a positive mood had a better memory for typical targets, whereas participants in a negative mood had a better memory for atypical targets. Second, a person’s mood may influence stereotyping by activating similarly valenced schemas (Clark et al., 1984; Isen, 1987, Isen & Levin, 1982). For example, being in a positive mood state may activate positive stereotypes or schemas about a member of a stereotyped group, rather than negative stereotypes or schemas. Hypothesis 3 . Priming is expected to impact judgments of targets when participants are in low distraction conditions (when ability is high) and recall is easier for both relevant and irrelevant mood. Informational Influences of Affect on Stereotyping The mood as information model proposes that people may use their mood as a direct source of information to be incorporated into the judgment (Schwarz, 1990). According to the model, the impact of mood is a function of its perceived informational value such that, when mood is perceived to be relevant to a social judgment, it can serve as an important source of information (Schwarz, 1990; Schwarz & Clore, 1983; Schwarz & Clore, 1988). People use their mood when they fail to realize that the affective reactions are independent of the target being judged. Correspondingly, they are

PAGE 23

13 unaffected by mood when they attribute it to other sources (Schwarz & Clore, 1983) and also when they do not focus on their affective reactions at all (Albarracn & Kumkale, 2003). The mood as information model assumes a person’s present mood can lead to mood congruent judgments if it has not been discounted as a relevant source of information. In fact, Schwarz and Clore (1983, 1988) state that mood congruency effects arise not because of memory-congruent priming, but because of people misattributing their mood as a reaction to the judgment target (Schwarz & Clore, 1983, 1988). According to a model of affect identification and discounting (Albarracn & Kumkale, 2003), people use mood as information in making judgments by first attending to their affective states and then attempting to determine the informational value of their affective reactions for the judgment in question. If they believe that their feelings are relevant to the judgment, they use them as information; otherwise, they try not to take them into account. Therefore, extraneous, irrelevant affect should have an influence when a person becomes sensitive to it but fails to discount it as irrelevant. However, irrelevant affect is unlikely to have an influence when people discount the affective reactions they identify, or when they do not identify them in the first place. The two-stage model makes innovative predictions about the role of ability and motivation on the informational influence of mood. Decreases in ability and motivation should disrupt affect identification in a similar fashion as they disrupt discounting. Therefore, whereas traditional models of affect as information assume that the influence of affect increases linearly with decreases in processing ability or motivation (e.g., Isbell & Wyer, 1999; Ottati & Isbell, 1996; Petty, Schumann, Richman, & Strathman, 1993),

PAGE 24

14 the model predicts a curvilinear impact of amount of thought on the influence of irrelevant affect as information. Up to a point, decreases in ability and motivation should increase the influence of affect because they prevent affect discounting. Beyond that point, however, decreases in ability and motivation should decrease the influence of affect because people should have difficulty identifying the affect to begin with. Consistent with these predictions, people have been found to use irrelevant affect as information to a greater extent when their ability and motivation to think about it were moderate rather than high or low. In contrast, Albarracn and Kumkale’s (2003) model predicts a linear impact of ability on the influence of relevant mood on attitudes. Initially, higher ability to think about feelings should allow people to realize that their mood associated with a social group is relevant to the judgment of a member of that group. Therefore, in high ability conditions, people should be able to use relevant mood as a basis for their judgments. Moderate levels of ability may partially reduce their ability to realize that the mood is relevant to the judgment at hand, and thus the influence of mood on judgments should be moderate as well. Finally, further decreases in ability (high distraction) should prevent affect identification, and thus reduce the influence of mood altogether. Additionally, although priming, stereotyping and mood as information can all influence attitudes, the effects of priming and stereotyping are mediated by trait valence and specific stereotypic traits. Thus, the only direct effect on attitudes is that of mood as information. Hypothesis 4. In conclusion, consistent with the model of affect identification and discounting (Albarracn & Kumkale, 2003), I predict a curvilinear impact of ability on the influence of irrelevant mood on attitudes. Participants in low distraction conditions

PAGE 25

15 are not expected to use irrelevant mood as information because ability should be sufficiently high to allow them to discount mood. For moderate and high distraction conditions, I expect a decrease in the influence of irrelevant mood from moderate to high levels of distraction. Hypothesis 5 . Further, I predict a linear impact of ability on the influence of relevant mood on attitudes. Initially, I predict that lower levels of distraction will allow people to realize that the mood is relevant to the judgment (in low and moderate distraction conditions), and thus relevant mood will influence judgments. Moderate levels of distraction may reduce the impact of mood by interfering with their ability to judge the mood as relevant, and thus the influence of mood on judgments should be moderate. Finally, high distraction conditions should interfere with affect identification, and thus reduce the influence of mood altogether.

PAGE 26

CHAPTER 2 REVIEW OF LITERATURE Definition of Affect The term affect has been used to refer to a wide variety of emotional and motivational constructs, such as emotion, mood, arousal, incentives, conditioning, and reinforcement (Eagly & Chaiken, 1993). In the literature examining the influence of affect on stereotyping, affective states fall primarily into two broad categories: emotions and mood states. Emotions are object-specific, goal directed, and generally have specific cognitive labels (e.g., sadness versus anger), whereas mood describes more general, diffuse affective states that are not directed at a specific target (Isen, 1984, 1987; Schwarz & Clore, 1988). Affect varies on a number of dimensions, and in this review I will focus on three dimensions: valence, arousal level, and relevance. The first dimension of affect is valence or direction (positive or negative). Although different models include an infinite number of emotions, positive emotions invariably include happiness and joy, whereas negative emotions include anger and sadness. The majority of the research on positive mood (e.g., happiness) has concluded that positive mood is associated with an increased reliance on heuristic processing and thus stereotyping (Bodenhausen et al., 2001; Mackie & Worth, 1989). In contrast, sadness is generally associated with an increase in systematic processing and a reduction in reliance on stereotypes (Bodenhausen et al, 2001; Lambert, Khan, Lickel & Fricke, 16

PAGE 27

17 1997). Like happiness, anger tends to be associated with an increased reliance on stereotypes (Bodenhausen, Sheppard & Kramer, 1994). Examining the impact of valence alone, however, is not sufficient to understand the influence of affect on stereotyping. Another important dimension of affect is arousal level or intensity, which varies across different affective states. For example, even though fear, anger and sadness are all negatively valenced, these emotions differ in the level of associated arousal. Fear is associated with higher levels of arousal, anger is associated with moderate levels of arousal, and sadness is associated with lower levels of arousal. Consequently, each of these three emotions has disparate effects on cognitive processing, and thus on social judgments. Fear and anger tend to lead to decrements in systematic processing (Bodenhausen, 1993; Jepson & Chaiken, 1990), whereas sadness tends to lead to increases in systematic processing (Bless, Bohner, Schwarz, & Strack, 1990; Bodenhausen, 1993). By the same token, elation and contentment are both positively valenced emotions, yet it is likely that they will also have different influences on stereotyping, as elation is associated with a much higher level of arousal than contentment (Worth & Mackie, 1987). Finally, relevance is another critical dimension of affect, particularly in the context of intergroup judgments. Intergroup judgments and perceptions may be influenced by two types of affect: irrelevant and relevant. On the one hand, irrelevant (or incidental) affect or emotions are not directly related to the intergroup context (Bodenhausen, 1993). Research related to irrelevant affect has focused on the extent to which affect in one situation can affect judgments in other, objectively unrelated contexts (Isen, 1987; Schwarz & Clore, 1988; Wyer & Srull, 1989). For example, a woman may

PAGE 28

18 be in a bad mood as a result of an argument with her husband and use that mood as a basis for subsequent judgments about her job satisfaction, even when the mood she is experiencing is objectively irrelevant to her attitudes about her job. On the other hand, Bodenhausen (1993) defined relevant (or integral) affect as the emotions that are elicited by the social group itself and distinguished between two types of relevant affect depending on the duration of the state (Bodenhausen et al., 2001). Episodic relevant affect refers to affective states that are situationally experienced in intergroup situations, whereas chronic relevant affect concerns stable dispositions (Bodenhausen et al., 2001). For instance, it is possible to have chronic negative emotional reactions to a particular ethnic group but have positive episodic affect toward a member of that group based on a specific, pleasant interaction. As Bodenhausen et al. (2001) noted, chronic relevant affect cannot be manipulated and is more appropriately considered a “subject variable,” whereas episodic affect can be easily induced in the laboratory. For most stereotyped groups, chronic relevant affect is often negative because negative emotions (e.g., disgust, anxiety, or anger) are frequently associated with members of these groups (Bodenhausen, 1993). Alternatively, episodic relevant affect is affect that is elicited by exposure to members of a particular group, specifically the group to which the judgment target belongs. For example, if the manager of a company will be making a decision about whether or not to hire a new Asian-American accountant for his company, the bad mood induced by the Asian-American waiter who spilled soup on him at lunch may be judged relevant. Although, like other types of affect, this relevant affect entails diffuse affective reactions, the source of these affective reactions is a member of a target group, thus affect might be

PAGE 29

19 particularly likely to influence judgments about the source of affect, as it coincides with inferences about mood congruent traits of that group. As a result, the manager’s hiring decisions might be influenced by the affective reactions themselves, in addition to the attributions that these reactions are due to the incompetence of the waiter. Literatures Relevant to the Review Papers were included in this review if they examined the influence of affect on intergroup judgments (e.g., prejudice, stereotyping). Studies were only included in the review if they specifically addressed the influence of affect on intergroup judgments (e.g., stereotyping, discrimination). Thus some papers may have had four studies, but only two were included in the review because they fit the criteria for inclusion in the review. The literature search resulted in 31 papers and 56 studies (k = 56), spanning from 1989 to 2002, although the literature search spanned until 2004. To survey the state of the literature on affect and intergroup judgments, I classified each study included in our review according to publication source of the study, language of the paper, study design, participant type, type of affect manipulation (e.g., candy provision or mood inducing video clips), result of the manipulation checks, affect type (e.g., happy or sad), valence (e.g., positive, negative, or neutral), arousal level (e.g., high or low), and relevance (e.g., relevant or irrelevant), type of target of intergroup judgment (general social groups, ethnic/racial groups, or fictitious/hypothetical groups), type of social judgment task (evaluations of a group member or judgments of guilt in a student court context), dependent measures, goal of the study, influence of mood that was demonstrated in the study, and any affect models that were supported by the study. Table 2.1 presents a summary of these characteristics.

PAGE 30

20 Publication Source of the Study The majority of the studies in the review were published in peer reviewed journals (k = 46) and review chapters (k = 5). Studies reported in chapters were included if there was sufficient information to code on the categories of interest. The unpublished studies were from doctoral dissertations (k = 5). Occasionally studies that were published in peer-reviewed journals would also appear in review chapters. In these cases only the studies in peer reviewed journals were included in the review to avoid redundancy. See Table 2.1 for the specific publication source for each study. Language Fifty-three of the studies in the review were written in English, two in Spanish and one in German. A researcher who spoke Spanish and German coded the non-English reprints and I coded the English ones. Participants All but one of the study samples comprised student participants (k = 55). The only non-student sample included participants from gay bars (Gold, 2002: study 2). Additional information about characteristics of the sample (e.g., gender, ethnic background, age, university affiliation) appears in Table 1. Type of Affect In many of the studies of the influence of affect on intergroup judgments, researchers induced or measured a general mood state instead of a specific emotion (k = 28). For example, rather than a happy or sad mood, the mood induction procedure is designed to create a generally positive or negative mood state (Abele, Gendolla & Petzold, 1998: study 1 and 2; Burgess, 1993: study 2; Dovidio, Gaertner, Isen &

PAGE 31

21 Lowrance, 1995; Dovidio, Gaertner & Loux, 2000: study 2, 3, 4, and 5; Esses & Zanna, 1995: study 1, 2, 3, and 4; Forgas, 1992b; Forgas, 1992c: study 1, 2, and 3; Forgas & Fiedler, 1996: study 1 and 2; Haddock et al., 1994; Jackson & Sullivan, 2001; Moreno & Bodenhausen, 2001; Queller, Mackie, & Stroessner, 1996: study 1 and 2; Rodriguez Torres, 1998; Schiff, Esses & Lamon, 1992; Stepper & Strack, 1993; Stroessner, Hamilton & Mackie, 1992: study 1 and 2). In the 28 studies in which a specific emotional state was induced, the emotions were happiness (k = 24), sadness (k = 20), anger (k = 4), gratitude (k = 1), or hypocrisy (k = 1). Mood Induction Procedures Researchers have studied the influence of affect on intergroup judgments using a wide range of affect manipulations, including video mood induction, vivid event recall, Velten mood induction, musical mood induction, false feedback procedure, facial contractions, reading mood inducing stories, candy provision, priming faces with positive and negative adjectives, witnessing an interaction, and a unique hypocrisy induction. Five studies included a combination of two of these mood induction procedures and two studies measured instead of manipulating mood states (Jackson & Sullivan, 2001; Moreno & Bodenhausen, 2001). Researchers in twenty one studies used a videotape to induce various mood states (Abele et al., 1998: study 1 and 2; Burgess, 1993: study 1; Dovidio et al., 2000: study 2, 3, and 4; Forgas, 1992b; Forgas, 1992c: study 1 and 2; Forgas & Fiedler, 1996: study 1; Forgas & Moylan, 1991; Mackie et al., 1989: study 1 and 2; Park, & Banaji, 2000: study 1, 2 and 3, Stroessner et al., 1992: study 1 and 2; Stroessner & Mackie, 1992: study 1 and 2; Stroessner & Mackie, 1993: study 3). Manipulation checks indicated that the video

PAGE 32

22 mood inductions were successful in all cases. Only one study did not report a mood manipulation check (Stroessner & Mackie, 1993: study 3). See Table 2.2 for a description of all of the video mood inductions. For example, Park and Banaji (2000: study 1 and 2) used three different clips to induce happy, neutral, and sad mood. The happy clip was extracted from the television show, Late Night with David Letterman, and featured a “top 10” list of good things about New York in the summertime and a series of stunts called “stupid human tricks” (e.g., attempting to climb into an advertising balloon without the balloon deflating or bursting). The neutral clip was from the television show Wild, Wild, World of Animals and described the behavior of lions in a non-emotional way. Finally, the sad clip was from Terms of Endearment, a movie in which the female lead character dies from cancer as a young mother. In Experiment 3, Park and Banaji’s (2000) positive mood induction involved video clips from a series of Candid Camera episodes, rather than David Letterman’s comedy performance. The main reason for this change was that stand-up comedy routines often utilize jokes that rely on stereotypes of different social groups. To avoid the association between stand-up comedy and stereotypic humor, Park and Banaji (2000) used the Candid Camera episodes for the third experiment. Another commonly used mood induction is an adaptation of the vivid event recall procedures developed by Strack, Schwarz, and Gschneidinger (1985). In sad conditions, participants are asked to vividly recall an event that made them feel happy, sad, or angry and to describe in detail how the event transpired. In contrast, participants in neutral mood conditions are asked to recall and describe the mundane events of the previous day. The instructions typically ask participants to focus on vivid, concrete, experiential aspects of the recalled event. As can be seen in Table 1, fifteen studies included the vivid event

PAGE 33

23 recall procedure for affect induction (Bless, Schwarz, & Wieland, 1996: study 1 and 4; Bodenhausen, Kramer & Susser, 1994: study 1 and 4; Bodenhausen, Sheppard & Kramer, 1994: study 1; Burgess, 1993: study 1; Esses & Zanna, 1995: study 1; Forgas & Fiedler, 1996: study 3; Gold, 2002: study 2; Jackson, Lewandowski, Fleury, & Chin, 2001; Krauth-Gruber & Ric, 2000; Lambert et al., 1997: study 1, 2, and 3; Sheppard, 1996: study 1). Manipulation checks suggested that the procedure was successful in producing the desired mood effects in each of the studies with two exceptions. In one study (Sheppard, 1996: study 1), although angry and sad participants differed from happy participants, they did not differ from the neutral mood, nor from each other. Similarly, in another study (Burgess, 1993: study 1), the mood ratings by participants in the positive mood condition did not differ significantly from the mood ratings by participants in the neutral mood condition. Another mood manipulation that relies on the affect arousing properties of personal memories was used by Esses and Zanna (1995: study 1). They used a combination of a mood induction procedure developed by Velten (1968) and revised by Polivy and Doyle (1980), and a recall-of-events procedure (Izard, 1972) to induce positive, negative or neutral mood. Participants first read 60 Velten mood induction statements out loud; the statements became increasingly positive or negative depending on whether the participant was in the positive or negative mood conditions. The statements in the neutral mood condition contained no emotional content. After reading the statements, participants were asked to recall and describe a life event that had an affective tone that was similar to the set of Velten statements that they read aloud (Izard, 1972). Dovidio et al. (2000: study 4) also used a combination of the video mood

PAGE 34

24 induction and a modified Velten procedure to induce mood successfully. In all those cases, manipulation checks indicated that the procedures were successful in producing the intended mood states. The researchers in six studies used musical mood induction procedures (Bodenhausen, Kramer & Susser, 1994: study 3; Esses & Zanna, 1995: study 2, 3, and 4; Haddock et al., 1994; Rodriguez Torres, 1998). See Table 2.3 for a description of the songs used for all the musical mood inductions. For instance, to examine moods at different levels of arousal (happy and calm versus happy and excited versus neutral), Bodenhausen, Kramer and Susser (1994: study 3) chose a section of music from Delius’s A Village Romeo and Juliet (“The Walk Through Paradise Garden”), which is particularly relaxing, for the happy and calm condition. For the happy and excited music, they chose a section from Borodin’s Prince Igor, a vigorous dance sequence. For the neutral conditions, they chose “Solar Winds” by Hykes, which seemed to have a minimal effect on listeners. Bodenhausen, Kramer and Susser (1994: study 3) reported that both types of happy music resulted in greater levels of happiness than the neutral music, and that the happy, excited musical selection produced significantly higher levels of excitement than both the happy, calm selection and the neutral selection. Manipulation checks demonstrated that the use of music and imagery produced the expected mood states in all studies except two. In one study (Esses & Zanna, 1995: study 4), positive mood did not differ significantly from neutral mood, and in another study (Haddock et al., 1994), there was still a great deal of overlap in mood across the three mood conditions.

PAGE 35

25 Another method of mood induction was the false feedback procedure (n=3), which generally entailed receiving bogus successful or failure feedback after an assessment of some sort (Forgas, 1992c: study 3; Forgas & Fiedler, 1996: study 2; Stepper & Strack, 1993). For example, Forgas and Fiedler (1996: study 2) used a false feedback procedure in which participants were asked to make size estimations of the surface area of several complex geometric shapes. Participants were assigned to “overestimator” or “underestimator” groups and given manipulated feedback about their performances. In the positive mood conditions, participants were told that their accuracy scores were extremely impressive and that their excellent performance had positive implications for other skills (e.g., capacity to store, process and calculate abstract information). In the negative mood conditions, participants were told that their accuracy scores were quite disappointing (i.e., below average) and had negative implications for other skills. Participants in control conditions were thanked for their participation and told that their responses would be used to establish norms. Forgas (1992c: study 3) used a similar procedure except that participants received feedback after a measure of verbal abilities, and Stepper and Strack (1993) gave participants bogus feedback after a standard educational test. Manipulation checks revealed that in all cases, the false feedback procedures were successful in producing the expected mood states in each of these studies. A different type of procedure used in this domain is based on the assumption that emotions can be induced by facial feedback procedures or facial movements associated with the expression of that emotion (Adelman & Zajonc, 1989; Duclos et al., 1989; Ekman, Levenson, & Friesen, 1983). Three studies in the review reported using mood

PAGE 36

26 induction procedures that invol ved facial contractions (B odenhausen, Kramer & Susser, 1994: study 2; Meiland, 1996; Sc hiff et al., 1992). Bodenhaus en, Kramer & Susser (1994: study 2) used a facial contrac tion procedure in which particip ants were directed to smile, although the term “smile” was never explicitly used. Participants in happy conditions were instructed to perform specific muscle contractions until their expressions looked like a smile, whereas participants in the neutra l mood condition were told to loosely make a fist. All participants were told to main tain their poses while working on the social judgment tasks.1 Additionally, Meiland (1996) asked pa rticipants to read happy, sad or neutral fictional stories while also adopting facial expression s that matched the affect of the story. Male participants read a story w ith a male main character named Bob, whereas female participants read a st ory with a female main character named Beth. Participants who read happy stories were asked to smile, pa rticipants who read sad stories were asked to frown, and participants who read neutral st ories were not asked to adopt any specific facial expressions. Manipulation ch ecks verified that, in all ca ses, the facial contractions were successful in creating the intended mood states. Three studies used stories to induce the intended m ood (Meiland, 1996; Queller et al., 1996: study 1 and 2). Queller et al. ( 1996: study 1 and 2) used a mood induction procedure in which participants read a stor y about old friends gett ing together and the kindness of strangers (positive mood) or a stor y about an economic recovery in Chicago 1 Bodenhausen, Kramer and Susser (1994: study 2) argued that if th e facial feedback procedure resulted in the same level of happi ness as subjects who were made happy with the memory elicitation procedure, but not with greater levels of stereotyping, this would support the idea that distracting happy thought s are responsible for eliciting reliance on stereotypes. However, this idea was not supported.

PAGE 37

27 (neutral mood). Manipulation checks also indicated that the stories were successful in creating the intended mood states. In three studies, Dovidio et al. (1995) and Dovidio et al. (1995: study 1, 3 and 5) used an affect manipulation in which participants in the positive mood condition received a candy bar, whereas participants in the neutral mood condition did not. Participants in the positive mood condition were told that the candy bars were left over from another study and that they could have one right then or feel free to take it with them. Manipulation checks supported that participants who received the candy bar were significantly happier than those who did not. Some unique procedures have also been implemented to manipulate affect that is linked to a target (integral affect). In one study (Rodriguez Torres, 1998), positive and negative adjectives were used to prime black and white faces in combination with a musical mood induction in order to induce positive and negative affect. Manipulation checks indicated that the procedure was successful in creating the intended moods. In another study (Sheppard, 1996: study 2), participants witnessed an interaction between the experimenter and two confederates, one of whom could be identified as gay by three criteria confirmed to be associated with stereotypes of gay men: (a) he wore a shirt with a pro-gay statement (i.e., “Don’t assume I’m straight”), (b) he wore a single earring in the right ear (reported as a signal of being a gay man in the United States), and (c) he wore a strong cologne. The dyadic interaction took place in the context of an ostensible dilemma in which participants were told that they would be randomly assigned to either a long session or a short session. In the “happy” condition, the gay male confederate agreed to participate in the long version, essentially doing a favor to the other participants by

PAGE 38

28 volunteering for the long session. In the “angry” condition, the same confederate demanded to be in the short session because of his time constraints, essentially leaving the participant with no choice other than the longer time commitment. In the neutral mood condition, there was no emotionally charged interaction, and participants simply appeared to be randomly assigned to conditions. In the control condition, participants witnessed the angry, happy and neutral mood conditions, but with an apparently non-gay confederate. Manipulation checks provided some mixed support for the mood induction procedures in the directions predicted. Although the one way ANOVAs did not demonstrate significant differences among the happy, angry and neutral conditions for happiness or irritation, some limited evidence (via one-tailed t-tests) suggested mild differences between the angry and happy conditions. One study induced hypocrisy as a way to study the effects of negative affect. Specifically, Son Hing, Li and Zanna (2002) asked all participants to write persuasive essays on why they believed it was important to treat minority students on campus fairly, and were told that their essays might be featured in pamphlets promoting racial equality. After writing those essays, participants in the hypocrisy condition were told that the psychology department wanted to understand more about situations in which Asian students are not treated fairly and that they should write about two situations in which they treated an Asian person in a prejudiced manner. All participants wrote the essay advocating treating Asians fairly, but participants in the control condition did not describe situations in which they discriminated against an Asian person. Manipulation checks demonstrated that participants in the hypocrisy condition experienced more negative affect than those in the control condition.

PAGE 39

29 Finally, in two studies of the influence of affect on intergroup judgments, there was no affect manipulation. Rather, the researchers assessed the mood state of participants prior to the social judgment task (Jackson & Sullivan, 2001; Moreno & Bodenhausen, 2001). Moreno and Bodenhausen (2001) assessed negative affect toward gay men and lesbians in a mass testing procedure prior to participation in the experiment. They found that whenever perceivers were able to misattribute their feelings to some seemingly relevant feature (even if they considered their negative feelings about minority group members to be irrelevant), negative affect toward gay men led to substantially more negative evaluative reactions. In a similar study, Jackson and Sullivan (2001) did not induce affect, but assessed general mood state using the mood adjective checklist and found that negative affect predicted evaluations of gay men and lesbians, but not of heterosexuals. Arousal Level Very little research has examined the influence of affective arousal levels in the context of intergroup judgments, with the exception of two (Bodenhausen, Kramer & Susser, 1994: study 3; Bodenhausen, Sheppard & Kramer, 1994: study 1) producing conflicting results. In order to examine the impact of different types of negative affect, Bodenhausen, Sheppard and Kramer (1994: study 1) induced negative high arousal (angry) and low arousal (sadness) affect. Specifically, they expected anger to heighten reliance on stereotypes (similar to positive mood states), and they did receive support for this hypothesis. In contrast, Bodenhausen, Kramer and Susser (1994: study 3) induced positive high arousal (happy-energetic) and low arousal (happy-calm) affect in order to test whether physiological arousal was responsible for the disruption of systematic

PAGE 40

30 processing in positive mood states, but this hypothesis was not supported. Both positive mood states led to equivalent levels of stereotyping targets. Given the conflicting results of these two studies, it is diffi cult to draw conclusions about the impact of arousal levels on stereotyping. Relevance Another neglected variable in the affect and stereotyping literature is relevance2 (defined as affect elicited by or related to an intergroup context). Only four studies examined relevant affect (Moreno & Bodenhausen, 1996; Rodriguez Torres, 1998; Sheppard, 1996: study 2; Son Hing et al., 2002). Moreno and Bodenhausen (1996) did not manipulate relevant affect , but assessed it using a ques tionnaire that measured antigay affect. As predicted, they found that anti-gay affect only influenced explicit evaluations of gay men and le sbians when situational cues (poor essay quality) were present that seemed to legitimize a negative evaluation. In another study, Rodriguez Torres (1998) induced relevant positive and negative mood by priming black and white faces w ith positive and negative adjectives in combination with a musical mood induction a nd found that relevant negative mood was associated with negative beliefs about the target group. Sheppard (1996: study 2) directly mani pulated positive and negative relevant affect by having participants witness a positive or negative interaction between a gay and non-gay confederate in order to induce ange r or happiness associated with the gay confederate. However, there was no differe ntial impact of mood on judgments of gay 2 Another type of relevance, group releva nce, was examined by Forgas and Fiedler (1996), but given that it differs from relevant affect as defined above, this aspect of the study is not elaborated on in the review.

PAGE 41

31 targets, perhaps because of problems with the mood inductions. Regardless of mood, low prejudice participants tended to respond more negatively to non-gay targets, whereas high prejudice participants responded negatively to both gay and non-gay targets. Son Hing et al. (2002) did not explicitly discuss manipulating relevant mood, although, they induced a hypocrisy experience by having participants write a persuasive essay on why they believed it is important to treat minority students on campus fairly. Participants were ostensibly randomly assigned to write specifically about treating Asian students fairly. Moreover, to create a public commitment to their egalitarian stance, participants were told that their essays may be published in a pamphlet promoting racial equality. After the hypocrisy induction, participants completed a budget reduction exercise intended to detect subtle discrimination against Asians. The results indicated that aversive racists (but not low prejudiced participants) responded to the hypocrisy induction with increased feelings of guilt, discomfort, and reductions in prejudicial behavior. Targets of Intergroup Judgments Rather than limiting targets to members of stigmatized racial groups, researchers have used a variety of types of targets to study the influence of affect on intergroup judgments. Thirty one studies used targets that were members of general social groups that were not based on ethnic group membership. The most common social group not associated with ethnicity or race was gay men and lesbians (n=6; Burgess, 1993: study 2; Gold, 2002: study 2; Jackson & Sullivan, 2001; Moreno & Bodenhausen, 2001; Sheppard, 1996: study 1 and 2). Five studies used a track and field athlete who was accused of either a stereotypical or non-stereotypical offense as the target of the

PAGE 42

32 intergroup judgments (Bodenhausen, Kramer & Susser, 1994: study 1, 2, 3 and 4; Bodenhausen, Sheppard & Kramer, 1994: study 1). In four studies, participants were asked to make judgments of typical and atypical students from the following groups: Christians, college types, trendies, engineers, radical feminists, intellectuals, lazy bludgers, sporty types (Forgas, 1992b; Forgas, 1992c: study 1, 2 and 3). The remainder of targets in this category varied widely, including a Greenpeace representative or a manager of BASF, a large chemical company (Bless, Schwarz, and Wieland, 1996: study 1 and 4), different academic majors (Burgess, 1993: study 1), liberals or conservatives (Dovidio et al., 2000: study 3; Lambert et al., 1997: study 2), pro-life or pro-choice individuals (Dovidio et al., 2000: study 4), advertisers or accountants (Dovidio et al., 2000: study 5), a skinhead or a priest (Krauth-Gruber & Ric, 2000), sorority group members (Lambert et al., 1997: study 1; Stroessner & Mackie, 1992: study 2), attractive women in their twenties (Lambert, Khan, Lickel & Fricke, 1997: study 3), an accountant, construction worker, lawyer, or policeman (Mackie et al., 1989: study 1 and 2), surfers (Queller et al., 1996: study 1), members of Big Brother organizations (Queller et al., 1996: study 2), and East German or West German candidates (Stepper & Strack, 1993). As might be expected, however, some researchers investigated the influence of affect on the judgments of members of ethnic minority groups (n=15) or ethnic/racial groups with a defined social history. In five studies, participants made judgments of targets who were African-American and/or European-American (or Black and White) (Jackson et al., 2001; Park & Banaji, 2000: study 1, 2, and 3; Rodriguez Torres, 1998). In another five studies, participants made judgments of Chinese, Native Indian, Arabic, Pakistani, and Jewish targets (Esses & Zanna, 1995: study 1, 2, 3, and 4; Schiff et al.,

PAGE 43

33 1992). In the remaining studies, targets were Asian and Caucasian (Forgas & Moylan, 1991), French Canadian and Pakistani (Haddock Zanna & Esses, 1994), Hispanic (Bodenhausen, Kramer & Susser, 1994: study 1; Bodenhausen, Sheppard & Kramer, 1994: study 1), or Asian (Son Hing et al., 2002). Finally, twelve studies used target members of fictitious or hypothetical groups. For example, in six studies, researchers used the minimal group paradigm and assigned participants to an in-group (“overestimators”) or an out-group (“underestimators”) (Abele et al., 1998: study 1 and 2; Dovidio et al., 2000: study 2; Forgas & Fiedler, 1996: study 1, 2, and 3). In five studies, participants were asked to evaluate target members of a hypothetical group A or B (Meiland, 1996; Stroessner et al., 1992: study 1 and 2; Stroessner & Mackie, 1992: study 1; Stroessner & Mackie, 1993: study 3), and in yet another study, participants judged members of another group who happened to be working on a similar task (Dovidio et al. 1995). Social Judgment Task The types of judgment/evaluation tasks can be divided into three subcategories: judgments or evaluations of members of a hypothetical or minimal group, judgments of members of specific social groups, and stereotypic judgments of members of specific social groups. The types of judgments included likability ratings, evaluations of targets on general characteristics (e.g., introverted/extroverted, intelligent/unintelligent), or evaluations of the target in a specific context (e.g., career mindedness, consumer orientation). Seven studies required that participants make judgments or evaluations of members of a hypothetical or minimal group (Abele et al., 1998: study 1 and 2; Meiland, 1996; Stroessner et al., 1992: study 1 and 2; Stroessner & Mackie, 1992: study 1;

PAGE 44

34 Stroessner & Mackie, 1993: study 3). For example, in one study, participants were categorized as overestimators or underestimators based on a dot estimation task and then asked to rate the likeability of other ingroup and outgroup members (Abele et al., 1998). In twenty one studies, participants were asked to make similar judgments, but of specific social groups (Bless, Schwarz & Weiland, 1996: study 1 and 4; Burgess, 1993: study 1 and 2; Forgas, 1992b; Forgas, 1992c: study 1, 2, and 3; Forgas & Moylan, 1991; Jackson & Sullivan, 2001; Lambert et al., 1997: study 1 and 2; Mackie et al., 1989: study 1 and 2; Moreno & Bodenhausen, 2001; Queller et al., 1996: study 1 and 2; Rodriguez Torres, 1998; Sheppard, 1996: study 1 and 2; Stroessner & Mackie, 1992: study 2). For example, using an impression formation task, Queller, Mackie et al. (1996: study 1) told participants that they would be participating in a study about perceptions of social groups and were asked to read descriptions of eight surfers. In order to activate the stereotype for surfers, participants were also told that they would be reading traits that were stereotypic for surfers. The descriptions consisted of 48 behavioral statements, twelve statements that reflected stereotypes about surfers (“casual” and “clique-ish”), twelve statements that reflected counterstereotypic traits (“success oriented” and “intellectual”), and 24 statements reflecting 12 neutral traits. Half of the participants in each mood condition were presented with stereotype inconsistent information that was concentrated in two of the eight group members, whereas the other half was presented with the same stereotype inconsistent information distributed across six of the eight group members. Thus participants received either concentrated inconsistent information or dispersed inconsistent information, and then judged targets.

PAGE 45

35 In eleven studies, rather than general judgments of targets, participants made stereotypic judgments of members of specific social groups (Esses & Zanna, 1995: study 1, 2, 3 and 4; Gold, 2002: study 2; Jackson et al., 2001; Park & Banaji, 2000: study 1, 2, and 3; Schiff et al., 1992). For example, Park and Banaji (2000: study 1) asked participants to recognize the name of ostensible criminals and politicians. They presented a list of 68 pre-tested European-American and African-American names and asked participants to indicate whether each person was a criminal (stereotypical for African Americans) or a politician (stereotypical for European Americans). The expectation was that participants would misidentify African-American names as criminal and European-American names as politicians. In another stereotypical judgment task, Schiff et al. (1992) asked participants to describe members of six ethnic groups: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish. Specifically, they were asked to list characteristics that described typical members of each group, to assign a valence to each characteristic, and to indicate the percentage of group members to which each characteristic applied. In a similar task, but using a different context, Gold (2002: study 2) asked participants to read brief descriptions of gay men and to make estimates about the likelihood that the men were HIV positive. Participants read the descriptions and made estimates of 7 targets quickly. There were 14 targets in total, and participants only evaluated half of the targets. The targets were described as (a) being of high or low intelligence, (b) having a pleasant or unpleasant personality, (c) preferring saunas and sex clubs or nightclubs and bars, (d) preferring insertive anal intercourse or both insertive and receptive anal intercourse, (e) being a finance manager or a doctor, (f) self identifying as

PAGE 46

36 gay or bisexual, and (g) usually wearing fashion labels or leather. Stereotypical portrayals of gay men resulted in higher estimates that the target was HIV positive. Six of the studies used a “student court” context in which participants make judgments about the likelihood that a student is guilty of an offense that either is or is not stereotypically associated with members of the student’s group (Bodenhausen, Kramer & Susser, 1994: study 2, 3, and 4; Bodenhausen, Sheppard & Kramer, 1994: study 1; Krauth-Gruber & Ric, 2000). For example, Bodenhausen, Kramer and Susser (1994: study 2, 3, and 4) asked participants make judgments about two cases, a cheating case and an assault case. For the assault case, the student-defendant’s name was either Juan Garcia (stereotype condition) or John Garner (no stereotype condition). For the cheating case, the target’s name was either followed by “a well-known track-and-field athlete” (stereotype condition) or by no such phrase (no stereotype condition). Judgments of guilt were higher when the target fit the stereotype associated with a particular crime. Another social judgment task utilized in five studies, consisted of asking participants to allocate rewards to ingroup and outgroup targets (Burgess, 1993: study 1; Forgas & Fiedler, 1996; Son Hing et al., 2002). For example, in one study, participants were told to complete a set of allocation matrices for ingroup and outgroup members with the promise that at the end of the experiment, candy would be distributed to members of each group based on the reward allocations of all the participants (Forgas & Fiedler, 1996). In another study, Son Hing et al. (2002) used a behavioral measure of discrimination adapted from Haddock et al. (1993) in which participants were asked to fill out an anonymous ballot concerning financial cuts that various student organizations would receive from the Federation of Students. Participants were asked how the budget

PAGE 47

37 cuts should be distributed, and researchers were interested in the percentage of cuts that would be recommended for the Asian Students Association. The budget reduction exercise was expected to detect subtle discrimination against Asians. In five studies participants engaged in a group representation task (Dovidio et al., 1995; Dovidio et al.: study 2, 3, 4 and 5). For example, participants in one of these studies (Dovidio et al., 1995) first worked with a group on a problem-solving task, after which they were told that they would be working with another group on a different task. They saw a video of the other ostensible group and were asked to make judgments about the extent to which they felt the two groups would operate as different entities or as one group on a subsequent task (Dovidio et al., 1995). Finally, in two studies, an employment/hiring context was used in which participants made judgments about whether to hire a target for a position (Lambert et al., 1997: study 3; Stepper & Strack, 1993). In one study, participants were asked to role play as a job interviewer in which they received relevant information (e.g., vita, job history, etc) about each job candidate, as well as a photograph, after which they were expected to make a hiring decision on the basis of a person’s professional vitae. In a similar study, participants had to role play being a personnel manager whose task it was to select an eastern or western German candidate for a management position. The measures of interest in this case were not judgments, but actual hiring decisions. Dependent Variables As is apparent from the earlier discussion of judgment tasks, the types of dependent variables used in the studies generally comprised some form of judgment of a target. In thirty two studies, participants were asked to rate targets on a number of

PAGE 48

38 dimensions (Abele et al., 1998: study 1 and 2; Bless, Schwarz & Wieland, 1996: study 1 and 4; Burgess, 1993: study 1 and 2; Dovidio et al., 1995; Dovidio et al., 2000: study 2, 3, and 4; Forgas, 1992b; Forgas, 1992c: study 1 and 3; Forgas & Fiedler, 1996: study 3; Haddock et al., 1994; Jackson & Sullivan, 2001; Lambert et al., 1997: study 2 and 3; Mackie et al., 1989: study 1 and 2; Meiland, 1996; Moreno & Bodenhausen, 2001; Queller et al., 1996: study 1 and 2; Son Hing et al., 2002; Stepper & Strack, 1993; Stroessner et al., 1992: study 1 and 2; Stroessner & Mackie, 1992: study 1 and 2; Stroessner & Mackie, 1993: study 3). Nine studies also included recall measures (Bless, Schwarz & Wieland, 1996: study 1 and 4; Forgas, 1992c: study 2 and 3; Lambert et al., 1997: study 1; Mackie et al., 1989: study 1; Queller et al., 1996: study 1 and 2; Stroessner et al., 1992: study 2), and seven included measures of latencies or processing time (Abele et al., 1998: study 1 and 2; Dovidio et al., 2000: study 5; Forgas, 1992c: study 3; Forgas & Fiedler, 1996: study 3; Forgas & Moylan, 1991; Mackie et al., 1989: study 2). Seven studies included judgments of the extent to which targets were representative of certain groups (Dovidio et al., 1995; Dovidio et al., 2000: study 2, 3, and 4; Gold, 2002: study 2; Jackson et al., 2001; Park & Banaji, 2000: study 1), and six studies included stereotype judgments of targets (Esses & Zanna, 1995: study 1, 2, 3, and 4; Lambert et al., 1997: study 1; Schiff et al., 1992). Six studies required participants to make judgments of guilt of a target (Bodenhausen, Kramer & Susser, 1994: study 1, 2, 3, and 4; Bodenhausen, Sheppard & Kramer, 1994: study 1; Krauth-Gruber & Ric, 2000), four studies included measures of resource allocations (Burgess, 1993: study 1; Forgas & Fiedler, 1996: study 1, 2 and 3; ), two studies included signal detection analyses (Park & Banaji, 2000: study 2 and 3), two studies included measures of prejudice (Rodriguez Torres, 1998, Son Hing

PAGE 49

39 et al., 2002), and two assessed attributions for par ticipants’ own behaviors (Sheppard, 1996: study 1 and 2). In some cases, additiona l dependent variables were measured, but only the primary dependent variab les are included in Table 1. Goal of Study Although the focus of this review is to synthesize research examining the influence of mood on intergroup judgments, th e focus of the studies included in this review was often above and beyond the influe nce of mood on stereo typing,. In fact, the goals of the synthesized studies varied from comparing the differentia l impact of positive and negative moods, to exploring the unde rlying explanations for the differential influences of mood, to testing the predictions of specific models of affect and intergroup judgments. Given that most studies had inde pendent, multiple goals, a brief summary of each study’s goals is included in Table 1, but is not summarized here. Findings in the Reviewed Literature Influence of Mood on Heuristi c or Stereotypic Processing The majority of studies that have exam ined the relationship between affect and intergroup judgments have focused on the imp act of the valence of mood in order to determine when positive or negative mood will lead to an increased reliance on heuristic processing or stereotyping. Additional studies have examined the impact of mood arousal level (high or low). The results of these studi es and general results regarding valence and arousal level are discussed below.3 Two studies did not demonstrate mood effects, 3 An additional study examined the types of attributions that angry, grateful, happy and sad participants made for stereotype consis tent or stereotype inconsistent positive and negative behaviors (Jackson et al ., 2001). In general, grateful participants made stronger internal attributions for positive behaviors th an happy participants and angry participants made stronger internal attributions for nega tive behaviors than sad participants. The

PAGE 50

40 although this may have been the result of a problem with mood i nduction procedures (Sheppard, 1996: study 1 and 2). Influence of Positive Mood on Heuristic or Stereotypic Processing. Among those that examined the impact of valence, fourteen studies concluded that positive mood leads to greater reliance on heuristics or stereotypes (Abele et al ., 1998: study 1 and 2; Bless, Schwarz & Wieland, 1996: study 1 and 4; B odenhausen, Kramer & Susser., 1994: study 1 and 2; Forgas & Fiedler, 1996: study 1; Gold, 2002: study 2; Lambert et al., 1997: study 1 and 2; Park & Banaji, 2000: study 1, 2, and 3; Stepper & Strack, 1993; Stroessner & Mackie, 1992: study 1, 2, and 3). Ten additi onal studies demonstr ated that positive mood leads to increased heuris tic or stereotype-based pr ocessing, but only under certain conditions. For instance, in one study, positive mood led to increased reliance on stereotypes except when the participants were told that they would be held accountable for their judgments (Bodenhausen, Kramer & Susser, 1994: study 4). In another study, positive mood led to more intergroup discrimination when the personal relevance of the issue was low, but sadness led to more disc rimination when relevance was high (Forgas & Fiedler, study 2 and 3). Happy particip ants perceived greater homogeneity among group members, unless distinctive informati on was made salient (Queller et al., 1996: study 1 and 2). In a similar study, when group boundaries were distin ct and meaningful, positive affect increased intergroup bias. Howe ver, when a superordinate identity was made salient, the intergroup bi as diminished (Dovidio et al., 2000: study 3). In yet another study, happy participants demonstrat ed greater reliance on stereotypes except when stereotypes clearly contradicted specifi c information (Krauth-Gruber & Ric, 2000). results also suggest that it is important to distinguish between types of affect on dimensions other than valence when predicting effects on social judgments.

PAGE 51

41 In two studies, when participants expected to compete, positive mood led to less positive judgments of outgroup members, but when participants expected to cooperate with the group, positive mood led to more positive judgments of outgroup members (Dovidio et al., 1995: study 2 and 4). In another study, women in positive moods did form stronger stereotypes, but men failed to demonstrate any mood effects (Meiland, 1996). Several papers focused on the influence of positive affect on intergroup contexts from a different angle. Rather than look at the ways that positive mood can increase stereotyping and outgroup differentiation, these studies focused on ways in which positive mood can reduce intergroup bias by increasing cognitive complexity or reducing outgroup differentiation. Two studies found that positive mood led to more inclusive judgments of outgroup members (Burgess, 1993: study 2; Dovidio et al., 1995). Another study found that positive mood led to judgments of outgroup members as more similar to themselves, but not significantly differently from neutral mood, whereas sadness led to judgments that an outgroup member was more dissimilar (Burgess, 1993: study 1). Further, happiness reduced ingroup favoritism, whereas sadness increased ingroup favoritism (Burgess: study 1). Influence of Negative Mood on Heuristic or Stereotypic Processing. In eight studies, participants demonstrated increased reliance on stereotypes and higher levels of prejudice toward outgroup or members when in a negative mood (Esses & Zanna, 1995: study 1, 2, 3 and 4; Jackson & Sullivan, 2001; Moreno & Bodenhausen, 2001; Rodriguez Torres, 1998; Schiff et al., 1992; ). For example, one study found that negative mood led to negative evaluations of minority group members (i.e., homosexuals), relative to evaluations of majority group members (Jackson & Sullivan, 2001). In a similar study,

PAGE 52

42 females in a negative mood formed stronger stereotypes than females in a positive mood, although males failed to demonstrate any mood effects (Meiland, 1996). Seven studies in the review, however, reported more systematic processing when participants were in a sad relative to happy mood (Bless, Schwarz & Wieland, 1996: study 1 and 4; Lambert et al., 1997: study 1, 2 and 3; Park & Banaji, 2000: study 3) and relative to angry mood states (Bodenhausen, Sheppard & Kramer, 1994: study 1). Similarly, one study concluded that the induction of hypocrisy led to a reduction in prejudice among aversive racists (Son Hing et al., 2002). Therefore it seems clear that a negatively valenced mood alone is not sufficient to increase systematic processing. Nonetheless, the existing literature is silent with respect to the conditions under which each pattern of findings emerge. Mood Congruent Influence of Positive and Negative Mood on Judgments and Evaluations Twelve studies demonstrated a mood congruent bias in intergroup judgments (Abele et al., 1998: study 1 and 2; Esses & Zanna, 1995: study 1, 2, 3, and 4; Forgas, 1992b; Forgas, 1992c: study 1, 2, and 3; Forgas & Moylan, 1991). For instance, in one study, among participants who experienced high intensity affect, those in a positive mood expressed more favorable attitudes and stereotypes than participants in negative moods, who expressed more negative attitudes and stereotypes (Haddock et al., 1994). Influence of Mood Arousal on Heuristic or Stereotypic Processing In addition to valence, two studies examined the impact of the arousal level of moods. Among these studies, one study concluded that there was not a difference between high and low arousal moods in their reliance on stereotyping (Bodenhausen,

PAGE 53

43 Kramer & Susser, 1994: study 3), whereas another study concluded that high arousal moods lead to a greater reliance on heuristic processing than low arousal moods (Bodenhausen, Sheppard & Kramer, 1994: study 1). Theoretical Interpretation of the Review Findings Various models have been proposed to explain mood effects on intergroup judgments, including the mood congruency hypothesis, the reduced processing hypothesis, the physiological disruption hypothesis, the mood maintenance model, the hedonic contingency model, the mood as information model, the mood and general knowledge model, the positive mood and cognitive complexity hypothesis, the positive mood and generative processes hypothesis, the mood and information integration model, and the affect infusion model. Each model or hypothesis is described below and integrated into the description of the various ways that affect may influence intergroup judgments. In addition to the description of each model, the conclusions of the review regarding support for each model is also noted. Nine studies looked at the influence of mood on intergroup judgments, but did not address any of these models to explain the underlying processes of mood effects (Bodenhausen, Kramer & Susser, 1994: study 1; Gold, 2002: study 2; Jackson et al., 2001; Rodriguez Torrez, 1998; Sheppard, 1996: study 1 and 2; Son Hing et al., 2002; Stepper & Strack, 1993; Stroessner & Mackie, 1992: study 1). The remaining studies are discussed in the following sections as supporting or failing to support the various models. In general, this categorization is made based on the authors’ interpretation of their own results sections.

PAGE 54

44 Theories Explaining Reduced Processing or Increased Heuristic/Stereotypic Processing The reduced processing hypothesis proposes that certain mood states (e.g., happiness) lead to a reduction in capacity to process incoming information (Mackie & Worth, 1989). Most of the research related to mood and reduced processing has emphasized that happy people are more likely to engage in heuristic processing, whereas sad people are more likely to engage in systematic processing (Bodenhausen, Kramer & Susser, 1994; Mackie & Worth, 1989). For example, Bodenhausen, Kramer and Susser (1994) demonstrated that people in a positive mood assigned more guilt to someone who was a member of a negatively stereotyped group than to someone who was a member of a nonstereotyped group. Four studies in the review concluded that their results were consistent with the reduced processing hypothesis (Bodenhausen, Sheppard & Kramer, 1994: study 1; Park & Banaji, 2000: study 1; Queller et al., 1996: study 1 and 2). Cognitive explanations for reduced processing effects. The processing decrements may be linked to cognitive or motivational explanations’, and most studies have been unable to disentangle the two. The cognitive explanation contends that mood actually decreases an individual’s cognitive resources available for processing (Asuncion & Lam, 1995; Isen, 1987; Mackie & Worth, 1989; Worth & Mackie, 1987). However, recent research has challenged the cognitive explanation by demonstrating that under certain conditions, happy people are capable of processing information systematically and carefully (Bless et al., 1990; Schwarz, Bless & Bohner, 1991). In this review, three studies provided support for a cognitive explanation of reduced processing effects (Stroessner & Mackie, 1992: study 1 and 2; Stroessner & Mackie, 1993: study 3). For example, in one study, Stroessner and Mackie (1993: study 3) designed an experiment in

PAGE 55

45 which motivational and cognitive explanations of reduced processing would make different predictions by manipulating the amount of time that participants were given to look at information presented. After a mood induction, participants had either 3 or 7 seconds to look at behavioral information about the targets. If happy participants had difficulty processing the information, those who had 3 seconds to see each item should have been expected to do as poorly as participants in previous studies (Stroessner & Mackie, 1992: study 1 and 2). The critical condition, however, was when participants were given extra time. If processing decreases are the result of cognitive deficits, when participants in a positive mood have 3 seconds to perform the task, they should apply stereotypes to members of groups with either high or low variability, which in turn indicates a lack of systematic processing. Participants in a positive mood, however, should be able to compensate for the deficit by allocating extra time if they can. As a result, when they have 7 seconds for the task, they should apply stereotypes significantly more when the group was homogeneous than when it was heterogeneous. In contrast, if the observed decrease in cognitive processing for happy participants is the result of motivational deficiencies, participants in neither the 3 second condition nor the 7 second condition should differentiate between high and low variability groups. After all, having extra time is unlikely to motivate a participant to think about the task, The evidence uncovered by Stroessner and Mackie (1992) indicated that, when allocating extra time for the task, participants in positive and negative mood states were able to differentiate between the high and low variability groups as well as participants in a neutral mood state. Although these studies lended support to a cognitive constraint explanation of reduced processing In happy moods, five studies in the review suggest

PAGE 56

46 otherwise (Bodenhausen, Kramer & Susser, 1994: study 2; Lambert et al., 1997: study 1, 2 and 3; (Krauth-Gruber & Ric, 2000). For example, to test whether happy people who engage in heuristic processing are simply distracted by happy thoughts (cognitive distraction hypothesis), Bodenhausen, Kramer and Susser (1994) utilized a facial feedback mood induction in which participants would contract facial muscles into poses associated with the expression of an emotion. Bodenhausen, Kramer and Susser, (1994) reasoned that if the reason that happy people show greater reliance on stereotypes is the presence of distracting mental content, then people who are made happy by this content-free procedure should not demonstrate the same increased reliance on stereotypes. Contrary to this possibility, however, participants made happy by the facial feedback procedure stilled used stereotypes as a basis for their decisions. Another study that did not support the cognitive explanation of reduced processing hypothesis demonstrated that happy participants are capable of elaborative processing (Krauth-Gruber & Ric, 2000). In this study, following a mood induction (happy, sad and neutral mood), participants were asked to read a case description of an act of physical aggression for which a person had been charged. Participants were given detailed information about the act of aggression, category membership of the accused person (a skinhead or a young priest), and specific case information that varied depending on the experimental condition. The case information consisted of eight statements detailing evidence that varied in the proportion of incriminating information. Statements either implied guilt (e.g., the suspect was seen in a quarrel with the victim shortly before the aggression) or innocence (e.g., no objects belonging to the suspect was sound at the crime scene). There were 3 levels of case information: the innocence condition (6 items

PAGE 57

47 implying innocence and 2 implying guilt), the guilt condition (6 items implying guilt and 2 implying innocence), and the mixed condition (4 items implying innocence and 4 items implying guilt). In general, happy participants relied on stereotypes more than participants in both sad and neutral moods. Specifically, happy participants only demonstrated this reliance in the presence of ambiguous case information (the mixed condition with half innocent and half guilty items). In the innocence and guilt conditions, happy participants judged the skinhead and priest as equally innocent or guilty. The finding that happy participants can and did use specific case information (rather than category information) in the guilt and innocence conditions is inconsistent with the explanation that they utilize stereotypes because of reduced processing. Instead when the inconsistency made it necessary for happy participants to try harder, these partipants were motivated and able to compensate for deficits and engage in more deliberative processing in making judgments of guilt or innocence. A body of evidence that also supports the reduced capacity explanation has demonstrated a relationship between physiological arousal and increased stereotyping (Kim & Baron, 1988; Revelle & Loftus, 1992), suggesting that arousal and not valence are responsible for the effects I described. In this regard, Bodenhausen, Kramer and Susser (1994) reasoned that if higher levels of arousal are associated with stereotyping, then happy states that are low in arousal level (e.g., contentment) should not produce the same degree of stereotyping as happy states that are high in arousal level (e.g., excitement). In order to test this possibility, these researchers utilized a musical mood induction that would permit them to produce happy moods that varied in their energetic arousal levels (happy energetic music versus happy calm music). If the physiological

PAGE 58

48 disruption hypothesis has merit, then people in happy, high arousal mood states (happy energetic music condition) should demonstrate greater reliance on stereotypes than people in happy, low arousal mood states (happy calm music condition). However, although happy, energetic participants were significantly more excited than happy, calm participants, the two groups demonstrated equivalent levels of stereotypic bias. Thus, in our review, one study reported data that were inconsistent with the physiological disruption hypothesis (Bodenhausen, Kramer & Susser, 1994: study 3). Motivational explanations for reduced processing effects. One motivational explanation for reduced processing effects is that people in a positive mood may be motivated to maintain their positive mood state and avoid deliberative processing to do so (Isen, 1984, 1987). Another explanation is that participants in a positive mood interpret that they are in a safe environment that requires little systematic processing (Schwarz, 1990). An additional motivational explanation is that participants may simply prefer not to exert additional effort, particularly if there is a cue (e.g., a viable shortcut or heuristic cue) that justifies the reduction in effortful processing (Bodenhausen, Kramer & Susser, 1994). Two studies provided support for motivational explanations of reduced processing effects (Bodenhausen, Kramer & Susser, 1994: study 4; Dovidio et al., 1995: study 5; Meiland, 1996). For example, Bodenhausen, Kramer and Susser’s (1994) findings were suggestive of a motivational interpretation of reduced processing effects because they demonstrated that happy individuals process information more systematically when they are made accountable for their judgments. Similarly, Dovidio et al. (1995: study 5) demonstrated that participants in a positive mood were able to engage in more deliberative processing when they encountered expectancy-inconsistent information.

PAGE 59

49 Now, the mood maintenance model is ba sed on the assumption that people are motivated to maintain their pos itive mood states and to repa ir their negative mood states (Isen, 1987). In the context of intergroup judgments, this mo del would predict that that the main motivation for using stereotypes is to maintain good mood and repair bad mood. The hedonic contingency hypothesis is simila r to the mood mainte nance hypothesis in that people who are in a happy mood are expect ed to attend to information and tasks that will permit them to maintain their positive mood state (Wegener, Petty & Smith, 1995).4 Several studies were inconsis tent with both mood maintena nce explanations. Lambert et al. (1997) reasoned that their findings were not consistent with the hedonic contingency hypothesis because the hedonic contingency assumes no difference between sad and neutral mood participants in their use of st ereotypes. This was supported in one condition (pro-sorority attitudes) but not the other (antisorority attitudes). A dditionally, they argue that their results were also not consistent with mood maintenance model because happy and sad participants differed in their relia nce on stereotypes in one condition (antisorority attitudes) but not th e other (pro-sorority attitudes) . Thus, three studies did not support mood maintenance explanations or the hedonic contingency hypothesis (Lambert et al., 1997: study 1, 2 and 3). 4 However, in contrast to mood repair e xplanations, the hedonic contingency hypothesis assumes that people in sad moods and neut ral moods do not differ in their choice of processing strategies.

PAGE 60

50 Theories Explaining the Mood Congruent Effects on Evaluations and Other Judgments Mood priming. A substantial amount of research has supported the notion that moods influence a person’s cognitive processing in a way that is consistent with the valence of their affective state (Bower, 1991; Isen, 1993). These mechanisms facilitate mood-congruent judgments (Bower, 1981, 1991; Forgas, 1991a) —which imply a match between the affective content of a person’s affective state and his or her social judgments (Mayer, Gaschke, Braverman, & Evans, 1992). First, individuals are likely to recall information that is consistent with their mood, pay particular attention to external information consistent with their mood, and interpret all information in a way that is consistent with their mood (Bower, 1981; Isen, 1984). According to mood priming models (Bower, 1991), mood activates concepts in memory that are associated with the affective state. These cues in turn facilitate the recall of specific, mood-congruent information (Bower, 1991; Bower, Monteiro, & Gilligan, 1978; Clark & Isen, 1982; Forgas & Bower, 1988; Isen, 1984; Isen, Shalker & Karp, 1978) that may then be used in the interpretation of evidence that is relevant to the judgment, resulting in the production of a judgment that is biased in the direction of mood, such that positive mood leads to more positive judgments and negative mood leads to more negative judgments (Bower, 1991; Forgas & Moylan, 1987; Isen, 1987). Second, the valence of the affect one experiences may influence stereotyping by activating similarly valenced schemas or stereotype s. Some research has demonstrated the consistency of affective cuing such that positive affect induces positive cognitions, whereas negative affect stimulates negative cognitions (Clark, Milberg, & Erber, 1984; Isen, 1987; Isen & Levin, 1982). For example in one study, negative mood induced

PAGE 61

51 negative beliefs about the target group, strengthening prejudice toward the target (Rodriguez Torres, 1998). In four studies, Esses and Zanna (1995) demonstrated that participants in a negative mood attributed unfavorable stereotypes to certain ethnic minority groups and also interpreted certain attributes in a more unfavorable light. Thirteen studies in the review provided support for mood congruency in intergroup judgments (Abele et al., 1998: study 1 and 2; Esses & Zanna, 1995: study 1, 2, 3, and 4; Forgas, 1992b; Forgas, 1992c: study 1, 2, and 3; Forgas & Moylan, 1991; Haddock et al., 1994; Rodriguez Torres, 1998). For example, a study by Forgas and Moylan (1991) demonstrated that a positive mood led to more positive evaluations of Caucasian and Asian targets presented in pictures, relative to neutral mood. Nonetheless, five studies in the review provided data that were inconsistent with mood congruency and therefore priming type of mechanisms (Bodenhausen, Sheppard & Kramer, 1994; Lambert et al., 1997: study 3; Mackie et al., 1989: study 1 and 2). For example, in a study that examined the effects of anger and sadness on stereotyping, angry participants demonstrated an increased reliance on stereotypes, whereas sad participants engaged in systematic processing (Bodenhausen, Sheppard & Kramer, 1994). Thus angry participants made mood congruent judgments, but sad participants did not. Mood as information. Mood can also inform social judgments as people directly consult their mood, using a “how do I feel about it?” heuristic (Clore, Schwarz & Conway, 1994; Schwarz & Clore, 1988). According to the affect-as-information model, the impact of mood is a function of its perceived informational value. Thus, when mood is perceived to be relevant to a social judgment, it can serve as an important source of information (Schwarz, 1990; Schwarz & Clore, 1983; Schwarz & Clore, 1988). People

PAGE 62

52 use their mood when they fail to realize that the affective reactions are independent of the target being judged. Correspondingly, they are unaffected by mood when they attribute it to other sources (Schwarz & Clore, 1983) and also when they do not focus on their affective reactions at all (Albarracn & Kumkale, 2003). Like the mood priming hypothesis, the mood as information model assumes that people in positive moods should like other people more than people in a negative mood. In fact, Schwarz and Clore (1983, 1988) posit that mood congruency effects arise not because of memory-congruent priming, but because of people mistake (misattribute) their affect state for a reaction to the object they have to judge (Schwarz & Clore, 1983, 1988). As already discussed, thirteen studies in the review provided support for mood congruency in intergroup judgments (Abele et al., 1998: study 1 and 2; Esses & Zanna, 1995: study 1, 2, 3, and 4; Forgas, 1992b; Forgas, 1992c: study 1, 2, and 3; Forgas & Moylan, 1991; Haddock et al., 1994; Rodriguez Torres, 1998 ). Ten additional studies in the review were consistent with the moderators proposed by mood as information explanations (Bodenhausen, Kramer & Susser, 1994: study 4; Haddock et al., 1994; Jackson & Sullivan, 2001; Lambert et al., 1997: study 1, 2 and 3; Moreno & Bodenhausen, 2001; Park & Banaji, 2000: study 1, 2 and 3). For example, Moreno and Bodenhausen (2001) found that mood states can influence social judgments by serving as a source of information. Specifically, negative feelings about a minority group are only likely to influence judgments about a group member when the feelings can be misattributed to some aspect of the group, especially if there appears to be some legitimate basis for the misattribution. Thus, Moreno and Bodenhausen (2001) argue that their study both supported and extended mood as information theory by emphasizing the

PAGE 63

53 importance of admissibility concerns in the impact of affect on intergroup judgments. In a similar vein, Haddock et al. (1994) found that, among people who experience high intensity affect, those in a positive mood expressed more favorable attitudes and stereotypes than participants in negative mood. The authors suggested that mood as information may explain this effect such that people who experience high intensity affect may be more likely to make misattributions of the source of their mood. One study in the review reported data that was inconsistent with mood as information (Forgas, 1992b). In this study, following a mood induction, happy, sad, or neutral mood participants were asked to make judgments of typical and atypical targets. The author found a mood congruent bias that was significantly greater for atypical targets than for typical targets, as well as better recall of typical people when participants were in a positive mood and atypical people when participants were in a negative mood. Forgas (1992b) contends that mood as information may have some difficulty explaining the differential mood effects for typical and atypical targets. According to mood as information theory, mood should either be (incorrectly) interpreted as diagnostic of reactions to a target and influence judgments of that target, or be attributed to external causes and have no impact on judgments (Schwarz & Clore, 1988; Schwarz, 1990). Forgas (1992b) argues that the target-specific mood effects seem more consistent with multi-process models, rather than mood as information.

PAGE 64

54 Theories that Explain Increased Heuristic/Stereotypic Processing as a Function of the Valence of the Mood It is important to note that, according to the mood as information approach, mood may be utilized as a source of information alerting the person to the necessity of engaging in either heuristic or systematic thought (Schwarz, 1990; Schwarz & Clore, 1983, 1988). Depending on the goals of the person when they are making judgments, positive mood can serve as a signal that the environment is not problematic. Such a signal in turn reduces the motivation for more attentive, systematic processing. In contrast, sad moods alert a person of a possible problem in the environment, leading the person to engage in more systematic, elaborative thought to solve the problem. The mood and general knowledge model (MAGK) provides a functionalist account for the influence of mood on information processing and is an extension of the mood as information approach (Bless, Clore et al., 1996; Bless, Schwarz & Kemmelmeir, 1996; Bless, Schwarz & Wieland, 1996). Therefore, this model also proposes that mood states inform a person about the current state of the environment (Schwarz, 1990). A difference, however, is that the mood and general knowledge model posits that the simplified information processing demonstrated by people in a positive mood state is not due to reduced processing or reduced motivation, but to an increased confidence in their general knowledge structures (e.g., stereotypes, Krauth-Gruber & Ric, 2000). A positive mood (e.g., happiness) signals a safe environment in which reliance on general knowledge structures is adaptive and appropriate. In this case, the person feels confident relying on stereotypes, and is able to allocate cognitive resources to other tasks. In contrast, a negative mood (e.g., sadness) signals a problem in the environment. Therefore,

PAGE 65

55 confidence in general knowledge structures does not appear adaptive, leading sad people to process the situation more systematically. Another notable characteristic of the mood and general knowledge model is the absence of assumptions about mood effects on the amount processing capacity or processing motivation. Thus, rather than conceptualizing reliance on general knowledge structures as a consequence of reduced processing, this model considers reliance on general knowledge structures an antecedent of reduced processing (Bless, Schwarz & Wieland, 1996). Eight studies in the review provided support for the mood and general knowledge model in intergroup judgments (Bless, Schwarz & Wieland, 1996: study 1 and 4; Dovidio et al., 2000: study 4 and 5; Krauth-Gruber & Ric, 2000; Park & Banaji, study 1, 2, and 3). For example, Krauth-Gruber, and Ric (2000) found that when making judgments of a stereotypical and non-stereotypical target, happy participants only demonstrated an increased reliance on stereotypes in the presence of ambiguous case information. When case information was only slightly inconsistent, happy participants judged the two targets as equally innocent or guilty. In a similar study, participants in a sad mood were influenced by individuating information, but not category membership information, whereas participants in a happy mood were influenced by category information as long as the individuating information was not inconsistent with the category (Bless, Schwarz & Wieland, 1996). Theoretical Explanations for Other Effects of Mood Positive mood and cognitive complexity . Positive mood has also been demonstrated to increase cognitive complexity, such that people in positive moods are able to see both similarities and differences among items and are more responsive to

PAGE 66

56 category labels than participants in neutral or sad moods (Murray, Sujan, Hirt, & Sujan, 1990). The implications for intergroup judgments are that people in a positive mood should detect more similarities, as well as more variability when judging outgroup members. Four studies supported that positive mood increases cognitive complexity and leads to more inclusive categorizations of outgroup members (Burgess, 1993: study 2; Dovidio et al., 1995; Dovidio et al., 2000: study 2 and 3). Consistent with these ideas, the common in-group identity model proposes that intergroup bias can be reduced by influencing how group members categorize group boundaries and specifically that positive affect facilitates this process (Dovidio et al., 1995; Dovidio et al., 2000). The model proposes that the influence of positive affect is two-fold. First, positive mood increases cognitive complexity in categorization of groups (Murray et al, 1990). Second, positive mood promotes more inclusive representations of groups and reduced intergroup distinctions (Isen, 1987; Isen & Daubman, 1984; Isen, Niedenthal, & Cantor, 1992). However, the two relevant studies in our review did not provide support for the idea that positive mood increases cognitive complexity in processing outgroup members (Burgess, 1993: study 1; Stroessner & Mackie, 1992: study 2). For example, in one study, participants in a positive mood were asked to judge intragroup similarity and failed to perceive that some groups were more variable than others (Stroessner & Mackie, 1992: study 2). Therefore, instead of increasing cognitive flexibility, positive mood appeared to lead to reduced processing. Mood and information integration model. The mood and information integration approach proposes that mood has a dual impact due to its informational function and its

PAGE 67

57 enhancement of the weight of categorizing information (Abele et al., 1998). The model expands on the theoretical approaches of mood as information (Schwarz, 1990; Schwarz & Clore, 1983, 1988) and mood’s generative processes (Bless, Clore et al, 1996; Fiedler, 1988). It assumes that a person’s mood is always integrated into a judgment as a piece of information, although the degree of impact may be influenced by other variables, such as the amount or diagnosticity of additional information (Abele et al., 1998). If no other non-diagnostic information is present, judgments should be mood congruent, becoming more positive when positive mood is integrated and more negative when negative mood is integrated. If additional diagnostic information is present, however, then mood congruence should decrease as a result of the decreasing impact of that one piece of information (i.e., mood). Furthermore, the mood and information integration approach posits that positive mood leads to more elaboration of information and the elaboration likelihood associated with any piece of information is dependent on the variability of meaning that can be assigned to it (Abele et al., 1998). Because categorizing information such as stereotypes or group labels can be interpreted in varied ways relative to individuating information, more weight may be assigned to categorizing information when a person is in a positive mood (Abele et al., 1998). The mood and information integration theory makes several complex predictions regarding the impact of mood, category information and individuating information (Abele et al., 1998). First, exposure to a group label should generally weaken the influence of individuating information. Second, the less information there is to be integrated, the stronger mood congruency effects should be. Third, if mood impacts the weighting of categorical information, then the difference between evaluations

PAGE 68

58 of ingroup and outgroup members should be greater when people are in a positive mood. Fourth, if the weight of individuating information is not affected when people are in a positive mood, then the impact of individuating information should be the same for people in positive and neutral moods in the absence of group membership information. Thus, according to the model, positive mood should influence judgments of ingroup and outgroup members as a result of two processes: the integration of mood as one piece of information and the weighting of categorical information (which is also dependent on mood). In a minimal group paradigm, participants rated the likability of targets after receiving individuating information about them. In half of the cases, targets were also characterized as ingroup or outgroup members. In both studies, participants made greater ingroup-outrgoup differentiation when in a positive mood than a neutral mood. In the absence of a group label, the impact of individuating information was equivalent for participants in positive and neutral mood states, however, in the presence of a group level, the impact of individuating information diminished for those in a positive mood. Thus, two studies in the review provided support for the mood and information integration model (Abele et al., 1998: study 1 and 2). Positive mood and generative processes . Positive mood is also repeatedly associated with more creative elaboration of stimulus materials (Abele, 1992; Fiedler, 1988; Murray et al., 1990) and an increase in generative processes (Bless, Clore et al., 1996; Bless, Schwarz, & Wieland, 1996). These generative processes may in turn lead to an increase in stereotyping. For example, when considering a member of a stereotyped group, a person in a positive mood may think more about the implications of the

PAGE 69

59 stereotype and engage in more elaborative thinking about the stereotype, resulting in a stronger influence of the stereotype (Abele et al., 1998). Given that the relationship between positive mood and generative processes is a component of the mood and information integration model, it follows that the same two studies in the review have provided support for the influence of positive mood on generative processes (Abele et al., 1998). Affect infusion model . To clarify the influence of affect on social judgments, Forgas (1992a, 1995) recently incorporated the informational and processing effects of mood into a comprehensive multiprocess affect infusion model. Affect infusion is defined as the process by which affect influences and becomes part of the judgmental process. The model suggests that the degree to which affect influences judgments varies along a processing continuum and specifies alternative processing strategies that depend on (a) the kind of information search strategies used by judges (i.e., open vs. targeted search) and (b) the exhaustiveness of the information considered (i.e., full vs. partial search). The four processing strategies identified by the affect-infusion model are (a) direct access of a stored or pre-existing judgment, (b) motivated processing in service of a pre-existing goal or mood repair, (c) simplified or heuristic processing, and (d) systematic or substantive processing. According to the affect-infusion model, the choice of processing strategy (e.g., recruiting high or low infusion strategies) is determined by target, judge, and situational variables. Target variables that may influence processing strategy include familiarity, complexity, and typicality. The affect-infusion model argues that complex, unusual targets involve more extensive processing strategies, which lead to a greater influence of

PAGE 70

60 mood, especially when participants require more time to judge atypical targets. A target is judged to be familiar when the judge has previous exposure to the stimulus that gives the judge detailed information about the target (which would lead to a direct access strategy). Mood is not expected to influence the evaluation of familiar targets (unless personal relevance is high), but judgments of unfamiliar targets may be greatly influenced by affect. Features of the judge that may affect choice of processing strategy include personal relevance, motivational goals, and cognitive capacity of the judge. In the absence of a specific prior motivation, personally relevant judgments are more likely to be deeply processed. Recent evidence has demonstrated that the combination of high personal relevance and sad mood leads to a motivated mood repair processing strategy via increased stereotyping of an outgroup (Clark & Isen, 1982; Forgas, 1989; Forgas, 1991b; Forgas, 1991c; Forgas & Fiedler, 1996). For example, consider an Asian-American man in a sad mood, who must evaluate a Latino-American coworker in a domain that is important (or personally relevant) to him and may potentially affect his outcomes (e.g., an upcoming raise). Given that he is in a sad mood, the Asian-American man is likely to systematically consider all of the factors relevant to evaluating Latino-American. However, given the personal relevance of the outcome he may stereotype the coworker in an effort to make himself look better or feel better. The affect-infusion model also argues that when the target is not personally relevant, the judge is likely to utilize a direct access or heuristic processing strategy. When a judge is influenced by a prior motivation, affect infusion is limited. Sometimes, affect itself can have inherent motivational properties (e.g., mood repair in the case of

PAGE 71

61 sadness). As research has repeatedly demonstrated, the cognitive capacity of the judge influences choice of processing strategy in that when cognitive capacity is impaired, heuristic processing is more likely than systematic processing. Three studies in the review provided support for the affect infusion model (Forgas & Fiedler, 1996: study 1, 2, and 3). The affect-infusion model was recently applied in the context of intergroup discrimination to examine the effects of positive and negative mood on intergroup allocation decisions using the minimal group paradigm (Forgas & Fiedler, 1996). As predicted by the affect-infusion model, Forgas and Fiedler (1996) found that, when personal relevance was low, positive mood led to increased heuristic processing and greater intergroup discrimination. Moreover, when personal relevance was high, negative mood resulted in more systematic processing and yet, greater intergroup discrimination. One interesting prediction of the affect-infusion model is that systematic processing does not necessarily lead to reduced discrimination, as people may engage in a targeted, motivated strategy to repair one’s negative mood state. Furthermore, five studies in the review provided support for a multi-process model by Forgas, prior to the development of the affect infusion model (Forgas, 1992b; Forgas, 1992c: study 1, 2 and 3; Forgas & Moylan, 1991). In these studies, Forgas has proposed the need for a multi-process model that takes additional variables (e.g., target, judge, situation) into account when trying to explain the influence of mood and reasoned that the differential influence of mood based on target type (e.g., typical versus atypical) in these studies provides support for such a multi-process model.

PAGE 72

62 Limitations of Previous Studies Scarcity of Studies Examining Relevant Affect Of the fifty six studies examining the influence of affect on intergroup judgments, only four explored the influence of relevant affect (Moreno & Bodenhausen, 1996; Rodriguez Torres, 1998; Sheppard, 1996: study 2; Son Hing et al., 2002). One of those studies assessed a type of chronic relevant affect, specifically anti-gay affect (Moreno & Bodenhausen, 1996), whereas the other three studies induced episodic relevant affect (Sheppard, 1996: study 2; Son Hing et al., 2002). Rodriguez Torres (1998) relevant mood was induced by priming black and white faces with positive and negative words, in combination with a similarly valenced mood induction. Sheppard (1996: study 2) manipulated happiness associated with a gay confederate by having participants observe a positive interaction between a gay and non-gay confederate and manipulated anger toward a gay confederate by having participants observe a negative interaction. Son Hing et al. (2002) induced relevant affect (in the form of hypocrisy) by having participants take a public, pro-equality stance. Following the hypocrisy induction, participants completed a behavioral measure of discrimination against Asians. Additionally, in the three studies examining episodic relevant affect, the types of affect manipulated were anger, happiness and hypocrisy (Sheppard, 1996: study 2; Son Hing et al., 2002) or positive and negative affect (Rodriguez Torres, 1998). Additionally, given problems with the mood manipulations in Shepperd (1996), it is difficult to draw conclusions about the influence of relevant angry and happy moods. Thus, although these three studies take important first steps toward understanding the role that relevant affect plays in intergroup

PAGE 73

63 judgments, clearly more empirical work needs to examine the impact of both chronic and episodic relevant affect. Scarcity of Studies Examining Arousal Level Only two studies in the review examined the influence of moods with different levels of arousal (Bodenhausen, Kramer & Susser, 1994: study 3; Bodenhausen, Sheppard & Kramer, 1994: study 1). Bodenhausen, Kramer & Susser (1994: study 3) compared the influence of high and low arousal positive moods (happy-energetic versus happy-calm), and Bodenhausen, Sheppard & Kramer, 1994: study 1) compared high and low arousal negative moods (anger versus sadness). Ambiguity Regarding Mechanisms Unde rlying Influence of Mood on Intergroup Judgments Although the recent resurgence of research on affect has identified a number of mechanisms through which affect influences stereotyping and prejudice (e.g., mood as information, mood priming), a great deal of ambiguity remains regarding the conditions in which those mechanisms are influential. Fu ture research needs to integrate and clarify seemingly inconsistent findings in the lite rature, as well as linking the theoretical explanations that are available in the literature in a coherent manner.

PAGE 74

64 Table 2.1 Description of Studies in the Review Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influenceof mood in Intergroup Context Model/Theory Endorsement Abele, Gendolla & Petzold (1998: study 1 of 2), PSPB, English 2 (grp assn: overestim vs underestim) x 2 (order: mood induction in 1st sess vs 2nd sess) x 2 (mood: pos vs neut) x 2 (target's grp memb info: absent or present) x 2 (type of grp info: ingroup vs outgroup) x 2 (valence of the ind trait info: pos or neg) 50 psych students (32f, 18 m) video mood induction (successful), no affect manipulation for neutral positive vs neutral not studied irrelevant fictitious/ hypothetical group: outgroup member (overestimator vs. underestimator) judgment (minimal group), minimal group paradigm, impression formation task, traits likability ratings, response latencies for the judgments. Test whether mood enhanced ingroupoutgroup differentiation. Test the influence of group label on the impact of individuating info on mood. Test for mood congruency effects. Positive mood enhanced outgroup differentiation (inc in ingroup enhancement, outgroup devaluation). Impact of individuating info the same in positive and neutral mood (but smaller when group label was present). Mood congruency effects. Support for Mood and information integration model (Mood as info, Generative processes of pos mood), Mood congruency Abele, Gendolla & Petzold (1998: study 2 of 2), PSPB, English 2 (grp assn: overestim vs underestim) x 2 (order: mood induction in 1st sess vs 2nd sess) x 2 (mood: pos vs missing) x 2 (grp memb info: absent or present) x 2 (type of grp info: ingroup vs outgroup) x 2 (valence of the ind trait info: pos or neg) x 2 (trait homogen: homo or hetero) 40 students (28f, 12m) video mood induction, no video for neutral (successful) positive vs neutral not studied irrelevant fictitious/ hypothetical group: outgroup member (overestimator vs. underestimator) judgment (minimal group), minimal group paradigm, impression formation task, traits likability ratings, response latencies for the judgments. Further test the limits of the mood and information integration model by increasing the complexity of the individuating information. Stronger outgroup differentiation in positive mood. Impact of individuating info the same in positive and neutral mood (but smaller when group label was present). Only trend toward mood congruency effects (w/ inc in amt of info). Support for Mood and information integration model (Mood as info, Generative processes of pos mood), Mood congruency

PAGE 75

65 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Bless, Schwarz, & Wieland (1996: study 1 of 4), EJSP, English 3 (mood: happy vs neutral vs sad) x 2 (category valence: pos vs neg evaluated) x 2 (individuating info valence: positive vs negative) 142 students vivid event recall (successful) happy vs neutral vs sad not studied irrelevant general social group: Greenpeace rep (pos category) vs manager of lg chemical company (neg) category judgment (social group), impression formation task, tape recorded description target evaluation (career mindedness, consumer orientation, ecological responsibility, likeability), recall of info about target Test whether Ps in a happy mood use category info more, does it depend on whether individuating info is consistent or inconsistent w/ category? Pos mood most influenced by category label, but dependent on indiv info. Sad Ps made least use of category info, more of individ info; neutral influenced more by indiv info, but also influ by category. Impact of categorical info moderated by its relationship to indiv info. Support for Mood and General Knowledge Model (MAGK) (Mood as Information) Bless, Schwarz, & Wieland (1996: study 4 of 4), EJSP, English 2 (mood: happy vs sad mood) x 2 (category valence: pos vs neg) x 2 (category timing: beg vs end) x 2 indiv info valence: pos vs neg) 149 students vivid event recall (successful) happy vs sad not studied irrelevant general social group: Greenpeace rep (pos category) vs manager of lg chemical company (neg) category judgment (social group), impression formation task, tape recorded description target evaluation (career mindedness, consumer orientation, ecological responsibility, likeability), recall of info about target Test the post hoc explanation that the differential degree of inconsistency when positive category membership info was combined with negative individuating info underlied the happy mood findings. Happy Ps are more likely to rely on category info, encoding of individuating info is guided by category info, and only possible when category info precedes indiv info. Support for Mood and General Knowledge Model (MAGK) (Mood as Information)

PAGE 76

66 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Bodenhausen, Kramer & Susser (1994: study 1 of 4), JPSP, English 2 (affect: happy vs neutral) x 2 (stereotype: present vs absent) btw Ps, 2 diff scenarios and stereotypes 94 students (73f, 21m) vivid event recall (successful) happy vs neutral not studied irrelevant ethnic/racial group: hispanic student (S) versus no S info for assault case; general social group: track and field athlete (S) versus no S info for cheating case legal/ student court context likelihood of guilt (stereotype activation) Test differential influence of happy and neutral conditions. Greater reliance on stereotypes among happy Ps. In both scenarios, neutral moods treated targets the same, while pos mood judged stereotypic targets as more guilty. N/A Bodenhausen, Kramer & Susser (1994: study 2 of 4), JPSP, English 2 (affect: happy vs neutral) x 2 (stereotype: present vs absent) 51 students (37f, 14m) facial contractions (successful) happy vs neutral not studied irrelevant general social group: track and field athlete (S) versus no S info for cheating case legal/ student court context likelihood of guilt (stereotype activation) Test whether distracting happy thoughts account for increased stereotyping (using facial feedback). Pos mood Ps still relied on stereotypes more. Significantly greater perceptions of guilt among happy Ps w/ S target Does not support cognitive distraction hypothesis. Bodenhausen, Kramer & Susser (1994: study 3 of 4), JPSP, English 2 (affect: happy excited vs happy calm) x 2 (stereotype: present vs absent) 53 students (25f, 28m) musical mood induction (successful) happy high (excited ) versus low (calm) irrelevant general social group: track and field athlete (S) versus no S info for cheating case legal/ student court context likelihood of guilt (stereotype activation) Test arousal disruption hypothesis Similar guilt ratings with high and low arousal mood; did not support arousal disruption hypothesis. Does not support physiological disruption hypothesis. Bodenhausen, Kramer & Susser (1994: study 4 of 4), JPSP, English 2 (affect: happy vs neutral) x 2 (stereotype: present vs absent) x 2 accountability (high vs low) btw Ps factorial design 131 students (93f, 38m) vivid event recall (successful) happy vs neutral not studied irrelevant general social group: track and field athlete (S) versus no S info for cheating case legal/ student court context likelihood of guilt (stereotype activation) Test responsiveness to motivation (accountability) in overcoming stereotyping. Test effort conservation hypothesis. With low accountability pos mood Ps relied on stereotypes more, but happy accountable Ps showed reduced stereotyping. Supports idea that w/ motivation Ps can avoid stereotypic judgments. Supports effort conservation hypothesis. Support for Mood as Information.

PAGE 77

67 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Bodenhausen, Sheppard, & Kramer (1994: study 1 of 3), EJSP, English 3 (affect: angry vs sad vs neutral) x 2 (stereotype: present vs absent) btw Ps factorial design 135 psych students vivid event recall (successful) angry vs sad vs neutral high (angry) vs low (sad) vs neutral irrelevant ethnic/racial group: hispanic student (S) versus no S info for assault case; general social group: track and field athlete (S) versus no S info for cheating case legal/ student court context likelihood of guilt (stereotype activation) Test whether angry people show greater reliance on stereotypes than sad people. Test mood congruency and arousal effects. Angry Ps used stereotypes more than sad Ps. Important to consider other aspects of mood than valence. Does not support a mood congruency effect. Negative mood appeared to have its effects via its impact on processing strategies (reduced motivation or reduced capacity). Burgess (1993: study 1 of 2), Dissertation, English 3 (mood: positive vs neutral vs sad) 105 students (60f, 45m) video mood induction (successful) happy vs. sad vs. neutral (2) not studied irrelevant general social group: outgroup major judgment (social group), minimal group paradigm, impression formation task; reward allocation task point allocations, trait ratings Test whether positive mood enhances cognitive complexity. Examine the effects of positive mood on intergroup perceptions (trait ratings) and intergroup behavior (point allocations) Happy Ps were mostly likely to see an outgroup member as similar to themselves, but not signif diff from neutral Ps. For trait ratings, happy Ps were the least likely to show favoritism relative to sad and neutral. Extends prior research that suggests that happy moods can weaken the tendency to perceive ones group as superior. In terms of intergroup behavior, all Ps gave more points to ingroup members. Does not support that positive mood increases cognitive complexity.

PAGE 78

68 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Burgess (1993: study 2 of 2), Dissertation, English 2 (mood: positive vs neutral) 85 male students vivid event recall (unsuccessful) positive vs neutral not studied irrelevant general social group: class president or homosexual judgment (social group), impression formation task, scenarios ratings of target, attitude and behavioral ratings Test the effects of positive mood on stereotyping processes from an interpersonal perspective. Happy Ps saw an atypical homosexual as more similar to themselves, but not the atypical class president. Weak support for positive mood and cognitive complexity. Dovidio, Gaertner, Isen & Lowrance (1995: study 1 of 1), PSPB, English 2 (mood: positive vs neutral) x 3 (group size: two vs three vs four) x 2 (appearance (similar or dissimilar dress) 132 students (83 f, 49m) candy manipulation (successful) positive vs neutral not studied irrelevant fictitious/ hypothetical group: another group working on a similar task group problem solving task, group representati on task ratings of group representations, ratings of both groups on a series of dimensions Test the common ingroup identity model and test the influence of other factors (including affective factors) on social categorization. Participants in a positive mood formed more inclusive superordinate group representations. Common Ingroup Identity model supported. Pos mood leads to more cognitive flexibility and more inclusive categorizations. Dovidio, Gaertner, & Loux (2000: study 2 of 5), Chapter, English 2 (mood: postive vs neutral) x 2 (relationship: cooperative vs competitive) 85 students video mood induction (successful) positive vs neutral not studied irrelevant fictitious/ hypothetical group: outgroup member (overestimator vs. underestimator) minimal group paradigm; group representati on task judge representations of groups, evaluations of outgroup target Test the influence of positive affect on intergroup judgments and whether the results are consistent w/ the common ingroup identity model. When Ps expected to cooperate positive affect led to more positive attitudes toward outgroup and less positive attitudes when they expected to compete. Suggests that for groups w/ histories of conflict, pos mood may increase, rather than decrease bias. Pos mood leads to more cognitive flexibility and more inclusive categorizations.

PAGE 79

69 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Dovidio, Gaertner, & Loux (2000: study 3 of 5), Chapter, English. Also described in Dovidio, Gaertner, Isen, Rust & Guerra (1998) unpublished study. 2 (mood: positive vs neutral) x 2 (context: two groups vs dual identity) 180 students video mood induction or candy manipulation (successful) positive vs neutral not studied irrelevant general social group: liberals or conservatives (meaningful group) Watch a video of a competing group; group representation task. judge representations of groups, evaluations of outgroup target Test the common ingroup identity model and test the influence of other factors (including affective factors) on social categorization while using meaningful social groups. When group boundaries were distinct and meaningful, positive affect increased intergroup bias, but when a superordinate id was made salient, it was reduced. Pos mood leads to more cognitive flexibility and more inclusive categorizations. Dovidio, Gaertner, & Loux (2000: study 4 of 5), Chapter, English 2 (mood: positive vs neutral) x 3 (context: cooperative vs competitive vs independent) 87 students combo of videotaped mood induction and a modified Velten procedure (successful) positive vs neutral not studied irrelevant general social group: prolife or prochoice Watch a video of a competing, cooperative or independent group; group rep task. judge representations of groups, evaluations of outgroup target, memory measures Conceptual replication of previous studies w/ some differences: different mood manipulations, real social ids, included independent group, also included memory measures to assess processing. Those in pos moods showed more positive attitudes toward outgroup members in cooperation context than competition context. Results indicate that pos mood is not necess assoc w/ reductions in resources or motivation, but perhaps inc conf in knowledge structures. Support for Mood and General Knowledge (MAGK). Support for motivational explanations of reduced processing.

PAGE 80

70 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Dovidio, Gaertner, & Loux (2000: study 5 of 5), Chapter, English 2 (mood: positive vs neutral) x 2 (intergroup contact: independent outcomes vs competition) 93 students candy manipulation (successful) positive vs neutral not studied irrelevant general social group: advertisers or accountants Group rep task. Ps were either accountants or advertisers interacting w/ the other group. Ps were given info that was consistent w/ adv or account stereotype and told to comment. amount of time spent commenting on expectancy consistent and inconsistent info. Replicate previous studies on the effect of positive affect on evals of outgroup members. Clarify the operation of deliberative or heuristic processing produced by positive affect. Those in pos moods showed more positive attitudes toward outgroup members in cooperation context than competition context. When presented w/ expectancy inconsistent info, Ps in pos mood are able and willing to process more elaboratively. Support for Mood and General Knowledge (MAGK). Support for motivational explanations of reduced processing. Esses & Zanna (1995: study 1 of 4), JPSP, English 3 (affect: positive vs neutral vs negative) x 2 (gender: male vs female) x 6 (target ethnic groups: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish) 60 students (30f, 30m) combo of Velten mood induction and vivid event recall (successful) positive vs neutral vs negative not studied irrelevant ethnic/racial group: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish ethnic groups Stereotype judgment of ethnic groups (stereotype measure) stereotype score (valence and % of groups to which characteristics were attributed) Test the influence of positive, neutral and negative mood on expression of ethnic stereotypes. Ps in neg mood expressed neg stereotypes of Native Indian and Pakistani people; Ps in pos and neg mood expressed esp pos stereotypes of Eng Canadians (ingroup). Outline a downward comparison explanation for stereotyping. Consistent w/ mood-congruent memory explanation.

PAGE 81

71 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Esses & Zanna (1995: study 2 of 4), JPSP, English 3 (affect: positive vs neutral vs negative) x 6 (target ethnic groups: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish) 123 students (45 Eng Can, 33 Chinese, 45 other) music and imagery mood induction (successful) positive vs neutral vs negative not studied irrelevant ethnic/racial group: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish ethnic groups Stereotype judgment of ethnic groups (stereotype measure) stereotype score (valence and % of groups to which characteristics were attributed) Determine if study 1 mood affects would replicate with a non-cognitive mood induction (musical) and with a more diverse sample. Ps in neg mood expressed neg stereotypes of Native Indian, Pakistani and Arabic people; No own group enhancement. Downward comparison explanation seems less likely. Supports mood congruent memory explanation. Esses & Zanna (1995: study 3 of 4), JPSP, English 2 (affect: positive vs negative) x 2 (time of rating: standard or delayed) 6 (target ethnic groups: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish) 52 students music and imagery mood induction (successful) positive vs negative not studied irrelevant ethnic/racial group: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish ethnic groups Stereotype judgment of ethnic groups (stereotype measure) stereotype score (valence and % of groups to which characteristics were attributed) Determine whether negative mood influences the specific types of characteristics attributed to the affected target groups or the meaning (valence) placed on the characteristics. Ps in neg mood expressed neg stereotypes of Native Indian, Pakistani, Arabic and Chinese people for standard timing (no effects in delayed condition). Mood seemed to influence the meaning attached to characteristics, rather than type of characteristics used to describe groups. Support for Mood Priming

PAGE 82

72 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Esses & Zanna (1995: study 4 of 4), JPSP, English 3 (affect: positive vs neutral vs negative) x 6 (target ethnic groups: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish) 45 students music and imagery mood induction (partially successful) positive vs. negative * not studied irrelevant ethnic/racial group: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish ethnic groups Stereotype judgment of ethnic groups (stereotype measure) stereotype score (valence and % of groups to which characteristics were attributed) Test whether results replicate when mood could influence the meaning attached to characteristics but not the type of characteristics used (by inducing mood after generation of characteristics). Ps in neg mood expressed neg stereotypes of Native Indian, Pakistani and Arabic people; mood influenced the meaning attached to characteristics used to describe the groups (Mood seemed to have a stronger impact on interpretation of stereotypes, rather than type of stereotypes.) Doesn't seem to support downward comp explanation. Support for Mood Priming. Does not support mood as info (b/c process seems more unconscious/ automatic). Forgas (1992b: study 1 of 1), EJSP, English 3 (mood: happy vs control vs sad) x 2 (prototypical: high vs low) 66 students (37 f, 29m) video mood induction (successful) positive vs neutral vs negative not studied irrelevant general social group: typical or atypical group members (Christians, college types, trendies, engineers, radical feminists, intellectuals, lazy bludgers, sporty types) Judgment (social group), impression formation task ratings on 14 bipolar scales Test the hypothesis that positive and negative moods have asymmetric effects on judgments of typical vs atypical people (stronger positive mood effects for typical targets and negative mood effects for atypical targets). Demonstrated a mood-congruent bias in judgments that was greater for atypical targets. Negative mood had greater influence on judgments of atypical targets than pos mood. Pos mood Ps had better recall of typical targets, and neg mood Ps had better recall of atypical targets. Support for multi-process affect models (Forgas), Mood congruence and Mood-priming theories, but inconsistent with Mood as Information.

PAGE 83

73 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Forgas (1992c: study 1 of 3), JPSP, English 3(mood: happy vs control vs sad) x 3 (prototypical: high vs intermediate vs low) 72 students (39f, 33m) video mood induction (successful) positive vs neutral vs negative not studied irrelevant general social group: prototypical, intermediate or atypical group members (Christians, college types, trendies, engineers, radical feminists, intellectuals, lazy bludgers, sporty types) Judgment (social group), impression formation task ratings on 12 bipolar scales test the hypothesis that judgments of atypical persons should require more systematic, constructive processing w/ greater affective biases Demonstrated a mood congruent effect. Judgments of typical targets were more impacted by positive mood and those of atypical targets were more impacted by negative mood. multi-process model by Forgas, mood priming Forgas (1992c: study 2 of 3), JPSP, English 3 (mood: happy vs control vs sad) x 2 (prototypical: high vs low) 42 students (22f, 20m) video mood induction (successful) positive vs neutral vs negative not studied irrelevant general social group: typical or atypical group members (Christians, college types, trendies, engineers, radical feminists, intellectuals, lazy bludgers, sporty types) Judgment (social group), impression formation task cued recall task Test mood effects on memory. Expected greater mood effects on memory for atypical targets. Mood effects on memory were greater effect for atypical persons. Recall was better for typical targets in pos mood and for atypical targets in neg mood conditions. multi-process model by Forgas, mood priming

PAGE 84

74 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Forgas (1992c: study 3 of 3), JPSP, English 3 (mood: happy vs control vs sad) x 2 (prototypical: high vs low) 60 students (30f, 30m) false feedback procedure (successful) positive vs neutral vs negative not studied irrelevant general social group: typical or atypical group members (Christians, college types, trendies, engineers, radical feminists, intellectuals, lazy bludgers, sporty types) Judgment (social group), impression formation task, on computer ratings on bipolar scales, cued recall tasks, response latencies Replicate study 1 and 2, and use a computer controlled procedure to allow measurement of latencies. Mood congruent bias, Replicated previous findings and greater mood effects on processing and judgmental latencies for atypical targets. multi-process model by Forgas, mood priming Forgas & Fiedler (1996: study 1 of 3), JPSP, English 2 (mood: happy vs sad) x 2(matrix type: conflict, no conflict) (within) 58 students video mood induction (successful) positive vs negative not studied irrelevant fictitious/ hypothetical group: outgroup members (overestimators/ underestimators) minimal group paradigm; reward allocation task allocation judgments Preliminary investigation of mood on intergroup judgments using the minimal group paradigm. Happy Ps were more likely to discriminate more than sad Ps in both conflict and no conflict matrices. Affect Infusion Model by Forgas, reduced processing Forgas & Fiedler (1996: study 2 of 3), JPSP, English 3 (mood: positive vs control vs negative) x 2 (group membership: high vs low) x 2 (matrix type: conflict vs no conflict) 60 students (aged 18-22) false feedback procedure (successful) positive vs. neutral vs. negative not studied irrelevant fictitious/ hypothetical group: outgroup members (overestimators/ underestimators) minimal group paradigm; reward allocation task allocation judgments Test hypothesis that when group membership is personally relevant, sad Ps are more likely to engage in intergroup discrimination to repair mood. Also different mood induction procedure and in field setting, rather than lab. Pos mood more likely to discriminate when relevance was low, whereas sad mood Ps are more likely to do so when relevance is high. Motivational explanation (for inc stereotyping w/ neg mood). Affect Infusion Model by Forgas, reduced processing

PAGE 85

75 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Forgas & Fiedler (1996: study 3 of 3), JPSP, English 3 (mood: happy vs control vs sad) x 2 (group membership: high vs low) x 2 (meaningful reward allocation task: high vs low) x 2 (matrix type: conflict vs no conflict) 144 students (aged 18-22) vivid event recall (successful) happy vs. control vs. sad not studied irrelevant fictitious/ hypothetical group: outgroup members (overestimators/ underestimators) minimal group paradigm; reward allocation task ratings, allocation judgments, judgment latencies Replicate and extend studies 1 and 2. Test whether neg mood would lead to inc ident w/ a relevant ingroup and rejection of the outgroup. Also computerized and different mood induction. Pos mood more likely to discriminate when relevance was low, whereas sad mood Ps are more likely to do so when relevance is high.Reaction time evidence also supported model. Motivational explanation (for inc stereotyping w/ neg mood). Affect Infusion Model by Forgas, reduced processing Forgas & Moylan (1991: study 1 of 1), CE. English 3 (mood: happy vs control vs sad) X 2 (racial balance of dyad: same race vs mixed race) 198 students (107f, 91m) video mood induction (successful) happy vs. control vs. sad not studied irrelevant ethnic/racial group: Asians or Caucasions Judgment (social group), impression formation task ratings on 9 bipolar scales Test whether mood would bias stereotypes in a mood consistent direction and whether mood effects are stronger for mixed race dyads. Found mood congruent biases, and biases were significantly greater when target was part of a mixed race dyad. multiprocess framework, mood congruence Gold (2002: study 2 of 2), IJSA, English 3 (mood: happy vs unhappy vs neutral) x 2 (vignette: stereotypical vs nonstereotypical) 83 Ps from gay bars vivid event recall (successful) happy vs unhappy vs neutral not studied irrelevant general social group: gay men portrayed as stereotypical or nonstereotypical Stereotype judgment, made AIDSrelated judgments of targets likelihood estimates (0 to 100) (likelihood that the target was HIV positive) Test whether mood influenced AIDS-related judgments of gay targets Positive mood led to increased reliance on stereotypes. N/A

PAGE 86

76 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Haddock, Zanna, & Esses (1994: study 1 of 1), EJSP, English 3 (mood: positive vs neutral vs negative) x 2 (affect intensity: high vs low) 120 students (58f, 62m) music and imagery mood induction (partially successful) positive vs. neutral vs. negative not studied irrelevant ethnic/racial group: French Canadians, Pakistanis Stereotype judgment of ethnic groups (stereotype measure) attitudes, stereotypic beliefs, emotional associates, Affect Intensity Measure Test the effect of mood on favorability of intergroup attitudes and moderating effect of affect intensity. Among those who experience high intensity affect (AIM), Ps in positive mood expressed more favorable attitudes and stereotypes than Ps in negative moods. Support for Mood Priming and Mood as Information Jackson, Lewandowski, Fleury, & Chin (2001: study 1 of 1), JSP, English 2 (stereotype consistency) x 2 (valence of behavior) x 5 (affect: happy vs grateful vs sad vs angry vs neutral) 229 Anglo Am students (133f, 96m) vivid event recall (successful) happy vs sad vs angry vs grateful vs neutral not studied irrelevant ethnic/racial group: AfricanAmerican male/Caucasi an male Stereotype judgments. Read one of 4 statements and judged likelihood that target was an Afr Amer male or Cauc male and causal attributions for behavior. likelihood judgments, % of Afr Am and Cauc who fit the descriptions, causal attributions Test aspects of cognitive appraisal theories of emotion-that grateful and angry Ps would attribute causality for similarly valenced behaviors to the target more strongly than happy and sad Ps. Grateful Ps made stronger internal attributions than happy Ps for positive behaviors, angry Ps made stronger internal attributions than sad Ps for negative behaviors. N/A Jackson & Sullivan (2001: study 1 of 1), JSP, English 2 (targets sexual preference: homosexual vs heterosexual) x 2 (occupation: elementary ed vs fine arts) x 2 (social distance: home-state vs distant-state applicant) 131 Am male students affect is not induced. N/A not studied irrelevant general social group: homosexual or heterosexual application to a program in elementary ed or fine arts Judgment (social group), ratings of applicant, % of people in targets category applicant ratings, % of people in targets category tested the influence of cognitive and affective measures on evaluations of stereotyped group members. Negative affect impacted evaluations of homosexuals but not heterosexuals. Support for Mood as infor mation

PAGE 87

77 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Krauth-Gruber & Ric (2000: study 1 of 1), PSPB, English 3 (mood) x 2 (category membership) x 3 (case information) between Ss design 332 female psych students vivid event recall (successful) happy vs neutral vs sad not studied irrelevant general social group: skinheard or young priest legal/ student court context guilt judgments, measures of perceived violence, recall, mood evaluations. To test the predictions of the MAGK model against those of the reduced processing hypotheses, First, judgments of happy Ps appeared to be contingent on the degree of inconsistency between general knowledge structures and specific info (giving up stereotypes strategy). Second, when case info didn't blatantly contradict category info, happy people expressed stereotypes-due to a reevaluation of the specific info-more complex processing than it has generally been assumed. Mood-andgeneralknowledge model (MAGK), does not support reduced processing or mood congruency models. Lambert, Khan, Lickel, & Fricke (1997: study 1 of 3), JPSP, English 3 (mood: happy vs neutral vs sad) x 2 (Ps group attitudes: favorable vs unfavorable) x 2 (target group membership: member vs nonmember) 125 students (82f, 43m) vivid event recall (successful) happy vs neutral vs sad not studied irrelevant general social group: sorority group member or non sorority group member Judgment (social group), impression formation task: rate on 3 pos stereotype traits and 3 neg stereotype traits positive and negative stereotype composite, based on ratings, recall task. To understand the role of mood in stereotype correction. To examine four models of mood and info processing (cognitive capacity, mood as info, mood maintenance, hedonic contingency hypothesis). Replicated greater reliance on negative stereotypes by happy Ps relative to neutral Ps. Sad Ps show opposite pattern-overcorrecting for negative group attitudes, Does not support Reduced cognitive capacity, Does not support hedonic contingency. Does not support mood maintenance. Some support for Mood as Info.

PAGE 88

78 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Lambert, Khan, Lickel, & Fricke (1997: study 2 of 3), JPSP, English 2 (mood: sad vs neutral) x 2 (stereotype valence: favorable vs unfavorable) x 2 (behavior: favorable vs unfavorable) 118 students (61f, 57m) vivid event recall (successful) neutral vs sad not studied irrelevant general social group: conservatives vs liberals; sympathetic to homeless (favorable) vs irritated by homeless (unfavorable) Judgment (social group), impression task: personality ratings, typicality ratings personality ratings, free recall, typicality ratings, scoring on a sympathy composite. To understand the role of mood in stereotype correction. To examine four models of mood and info processing (cognitive capacity, mood as info, mood maintenance, hedonic contingency hypothesis). Conservatives vs. Liberals. Sadness led to stereotype correction for negative stereotypes but not for positive stereotypes. Propose an overcorrection model for sadness and stereotypes. Does not support Reduced cognitive capacity, Does not support hedonic contingency. Does not support mood maintenance. Some support for Mood as Info. Lambert, Khan, Lickel, & Fricke (1997: study 3 of 3), JPSP, English 2 (mood: sad vs neutral) x 2 (job type: attractiveness appropriate vs attractiveness inappropriate) 61 students (33f, 28m) vivid event recall (successful) neutral vs sad not studied irrelevant general social group: attractive women in their 20s hiring context, intentions to hire and inferences of personality traits intentions to hire, inferences of personality traits. To understand the role of mood in stereotype correction. To examine four models of mood and info processing (cognitive capacity, mood as info, mood maintenance, hedonic contingency hypothesis). Attractive women. Asymmetry in results was not necessarily due to stereotype valence, but whether the stereotype was perceived as an appropriate basis for judgment. Propose an overcorrection model for sadness and stereotypes. Does not support Reduced cognitive capacity, Does not support hedonic contingency. Does not support mood maintenance. Does not support Mood congruency. Support for Mood as Info.

PAGE 89

79 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Mackie, Hamilton, Schroth, Carlisle, Gersho, Meneses, Nedler, & Reichel (1989: study 1 of 2), JESP, English 3 (mood: positive vs negative vs neutral) x 2 (valence of stimulus info: positive vs negative) 71 students (50f, 21m) video mood induction (successful) happy vs sad vs neutral not studied irrelevant general social group: accountant, construction worker, lawyer, policeman Judgment (social group), illusory correlation task: read 32 statements about a male target. ratings of liking, frequency estimates, recall Test the impact of mood on expectancy based illusory correlations Mood incongruence increased illusory correlations. Did not support mood congruence (facilitation). Mackie, Hamilton, Schroth, Carlisle, Gersho, Meneses, Nedler, & Reichel (1989: study 2 of 2), JESP, English 3 (mood: positive vs negative vs neutral) x 2 (valence of stimulus info: positive vs negative) 93 students (55f, 38m) video mood induction (successful) happy vs sad vs neutral not studied irrelevant general social group: accountant, construction worker, lawyer, policeman Judgment (social group), illusory correlation task: read 32 statements about a male target, on a computer processing times, frequency estimates, trait ratings Examine the interpretation of study 1 (that the mood congruency results were due to increased difficulty of processing). To investigate the amount of time spent processing sentences w/ stereotypic and unrelated attributes that were congruent and incongruent . Mood incongruency interferes w/ processing and increases illusory correlations. Ps took longer to read sentences containing stereotype confirming info when the valence of the descriptive attributes were incongruent with Ps mood states. Did not support mood congruence (facilitation).

PAGE 90

80 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Meiland (1996: study 1 of 1), Dissertation, English 3 (mood: positive vs neutral vs negative) x 2 (gender of Ps: male vs female) x 2 (gender of target: male vs female) x 2 (need for cognition: high vs low). 206 students (125f, 81m) combo of reading mood inducing stories and facial contractions (successful) happy vs neutral vs sad not studied irrelevant fictitious/ hypothetical group: member of hypothetical group A or group B Judgment (hypo group), illusory correlation task: read 36 statements (positive and negative) about group member ratings of similarity, trait ratings, frequency estimation task Test the effects of mood, gender, and need for cognition on distinctiveness based stereotyping. Females in a negative mood formed stronger stereotypes than females in a positive mood, whereas males failed to reveal any mood effects. Ps formed more negative stereotypes about same gender targets than opposite gender stereotypes. Some support for motivational explanations of neg mood effects (when judging male targets). Moreno & Bodenhausen (2001: study 1 of 1), GPIR, English 2 (anti-gay affect: low vs high) x 2 (superficial essay quality: low vs high) 242 psych students affect is not induced. N/A not studied relevant general social group: progay rights essayist (not explicitly described as gay) Judgment (social group), evaluation of the essayist traits of the essayist Understand the influence of relevant affect on reactions to social groups. When the Ps was high in anti-gay (negative integral) affect and bad writer condition, significantly more derogation of essay position and essay writer Supports and extends mood as information. Park, & Banaji (2000: study 1 of 3), JPSP, English 2 (mood: positive vs neutral) x 2 (target race: Afr Am vs Euro Am) x 2 (task: criminal judgment vs politician judgment) mixed design w race and task manip w/in Ps 58 students (30f, 28m) (39 euro am, 4 afr am, 15 asian am) video mood induction (successful) happy vs neutral not studied irrelevant ethnic/racial group: AfricanAmerican names or European American names Stereotype judgment, names labeled as either criminal or politician judgment as criminal or politician Test a procedure to examine reliance on stereotypes in happy mood states Positive mood increased stereotype use (afr am as criminal, euro am as politician). Consistent w/ mood as info, mood and general knowledge, and cognitive capacity models.

PAGE 91

81 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Park, & Banaji (2000: study 2 of 3), JPSP, English 2 (mood: positive vs neutral) x 2 (target race: Afr Am vs Euro Am) x 2 (task: basketball player vs nonbasketball player) 68 students (39f, 29m) (41 euro am, 8 afr am, 19 asian am) video mood induction (successful) happy vs neutral not studied irrelevant ethnic/racial group: AfricanAmerican names or EuropeanAmerican names Stereotype judgment, names labeled as either basketball player or not signal detection analyses: hit rates, false alarm rates, results for d' and results for log Conceptual replication of Exp 1, but also explore underlying mechanism: sensitivity stereotyping vs criterion stereotyping Pos mood increases stereotyping; the effect is located both in reduced sensitivity in separating signal from noise when judging S members (sensitivity) and also lowered criterion for judging S members (criterion). Afr am names more likely to be misidentifed as bball player, false alarm rates noticeably higher in positive mood. Sensitivity stereotyping. Support for Mood as information, Mood and general knowledge Park, & Banaji (2000: study 3 of 3), JPSP, English 3 (mood: positive vs neutral vs negative) x 2 (target race: Afr Am vs Euro Am) x 2 (task: basketball player vs nonbasketball player) 90 students (45f, 45m) (59 euro am, 9 afr am, 22 asian am) video mood induction (successful) happy vs neutral vs sad not studied irrelevant ethnic/racial group: AfricanAmerican names or EuropeanAmerican names Stereotype judgment, names labeled as either basketball player or not signal detection analyses: hit rates, false alarm rates, results for d' and results for log Replicate study 3 and explore effects of negative mood Pos mood increases reliance on stereotyping, neg mood dampens such reliance. Neg mood tempered criterion stereotyping and did not disrupt sensitivity. Sensitivity stereotyping. Support for Mood as information, Mood and general knowledge

PAGE 92

82 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Queller, Mackie, & Stroessner (1996: study 1 of 2), JESP, English 2 (mood: neutral or positive) x 2 (info presentation: concentrated or dispersed) between Ps design 82 students (37f, 45m) read mood inducing newspaper stories (successful) positive vs neutral not studied irrelevant general social group: surfers Judgment (social group), impression formation task (8 surfers), judge similarity and typicality recall about target, homogeneity of the surfers, similarity of statements to the stereotype (consistent, inconsistent, or irrelevant), typicality of each surfer Test several hypotheses related to the idea that pos mood increases perceptions of similarities rather than differences (regarding similar results w/ behavioral consistency and member typicality w/in a social group, boundary conditions, and whether the effects were due to an overestimation of typicality or an underestimation of atypicality). Happy Ps tend to perceive similarities rather than differences. Happy Ps perceived the group as more homogenous except when distinctive info was made salient. Evidence of subtyping. reduced processing Queller, Mackie, & Stroessner (1996: study 2 of 2), JESP, English 2 (mood: neutral or positive) x 2 (group impression task:read or sort) between Ps design 90 students read mood inducing stories (successful) positive vs neutral not studied irrelevant general social group: members of Big Brother organization Judgment (social group), impression formation task while reading, or while sorting into subgroups homogeneity of targets, similarity on trait dimensions, recall about target. Again, test hypotheses related to idea that positive mood increases perceptions of similarity. Prevent subtyping by presenting all Ps w/ the same info about a mod heterogeneous group. Different processing goals for the read group than sort group. Happy Ps tend to perceive similarities rather than differences. Happy Ps perceived the group as more homogenous except when distinctive info was made salient. Thus under some conditions can reduce negative effects of pos mood. reduced processing

PAGE 93

83 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Rodriguez Torres, R. (1998: study 1 of 1), RPSA, Spanish 2 (majority of RPSA faces during priming: black vs white) x 2 (valence of adjectives: positive vs negative) x 3 (valence of music: positive vs negative vs neutral) 38 psych students Combo of priming asian, african and caucasian faces with pos and neg adjectives and musical mood induction. (successful) positive vs negative not studied irrelevant ethnic/racial group: Black and White faces Judgment (social group), recognition task Rosenthal's index of prejudice Test the hypothesis that negative affect induced negative beliefs about target group, thus strengthening prejudice. Wanted to use a measure that wasn’t sensitive to control attempts, hence the use of priming measures. Negative affect induced negative beliefs about the target group, thus strengthening prejudice. N/A Schiff, Esses, & Lamon (1992: study 1 of 1), CE, English 2 (facial contractions: right vs left) groups 42 students facial contractions (successful) positive vs. negative not studied irrelevant ethnic/racial group: Chinese, Native Indian, English Canadian, Arabic, Pakistani, and Jewish ethnic groups Stereotype judgment of ethnic groups (stereotype measure) stereotype score (valence and % of groups to which characteristics were attributed) Test whether unilateral facial contractions of the left and right sides of the face have the same impact on stereotyping as positive and negative mood induced in other ways. Ps who did facial contractions on left side (neg mood inducer) were more likely to express neg stereotypes than Ps who contracted right side (pos mood inducer), Facial contractions produced same effects as mood w/o cognitive involvement. Support for mood priming

PAGE 94

84 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Sheppard (1996: study 1 of 2), Dissertation, English 4 (mood: happiness vs sadness vs anger vs control) x 2 (prejudice level: low vs high) x 2 (target: gay vs nongay) 380 students vivid event recall (unsuccessful) happy vs sad vs angry vs control not studied irrelevant general social group: gay men Judgment (social group), read scenarios about traffic jam due to gay rally and reaching cab at same time as gay person who offers to share the cab. amendedshould-would measure, feelings index, attribution measure Test how mood and prejudice interact to influence the expression of prejudice. Regardless of mood, low Prej Ps tended to respond more negatively to nongay targets, whereas high prej reacted the same to both targets. N/A Sheppard (1996: study 2 of 2), Dissertation, English 2 (prejudice level: low vs high) x 3 mood (anger vs happiness vs neutral) x 2 (confederate: gay vs non-gay) x 2 (scenario target: gay vs non-gay. 166 students witnessing an interaction between a gay and non-gay confederate (partially successful) happy vs angry vs neutral not studied relevant general social group: gay men Judgment (social group), read scenarios about traffic jam due to gay rally and reaching cab at same time as gay person who offers to share the cab. amendedshould-would measure, feelings index, attribution measure Replicate study 1 and examine the influence of integral (relevant) affect. Replicated study 1Regardless of mood, low Prej Ps tended to respond more negatively to nongay targets, whereas high prej reacted the same to both targets.. All Ps reported feeling that their endorsed behaviors corresponded to their standards w/ gay targets, rather than nongay targets. N/A

PAGE 95

85 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Son Hing, Li, & Zanna (2002: study 1 of 1), JESP, English 2 (control vs hypocrisy) x 2 (low vs high implicit prejudice) 49 psych students hypocrisy induction (take pro-equality stance, then describe past discrimination s against Asians) (successful) hypocris y vs control not studied relevant ethnic/racial group: Asians behavioral discrimination measure: allocate reduction of resources for Asians External Motivation Scale, Asian Modern Racism Scale, Implicit Prejudice, Ratings of Essay, Meas of Negative feelings. Test whether a hypocrisy induction would reduce prejudiced responses among aversive racists. Aversive racists responded to a hypocrisy induction w/ guilt and discomfort and a reduction in prejudice (relative to control). N/A Stepper & Strack (1993: study 1 of 1), ZS, German 2 (background of resume: west german vs east german) x 2 (disposition: negative vs positive) 35 students false feedback procedure (successful) positive vs negative not studied irrelevant ethnic/racial group: East German or West German candidate Hiring context, Personnel selection of graduate engineer attitudes, attitude confidence, judgments of initiative, self confidence and professional competence of applicant Test whether the application of stereotypes depended on mood, based on evidence from Devine, Bodenhausen and Wyer among others. Pos mood Ps applied stereotype and devalued eastern candidate to simplify judgment. Bad mood Ps made positive evaluations of the eastern candidate b/c he had good qualifications, despite professional handicaps (systematic processing). N/A Stroessner, Hamilton, & Mackie (1992: study 1 of 2), JPSP, English 3 (mood: positive vs neutral vs negative) 2 (group: A vs B) 84 students (54f, 30m) video mood induction (successful) positive vs negative vs neutral not studied irrelevant fictitious/ hypothetical group: members of hypothetical groups Judgment (hypo group), impression formation task; illusory correlation task liking ratings, estimations of desirable and undesirable behaviors, recall of behaviors Test the influence of affect on distinctiveness based illusory correlations and test cognitive and motivational explanations. Only Ps in a neutral mood formed illusory correlations. Equivalent levels of inaccuracy for negative and positive mood, which contradicts idea that neg mood leads to deliberative processing. Supports cognitive explanation for influence of affect. Did not support motivational explanation.

PAGE 96

86 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Stroessner, Hamilton, & Mackie (1992: study 2 of 2), JPSP, English 3 (mood: positive vs neutral vs negative) 2 (group: A vs B) 60 students (33f, 27m) video mood induction (successful) positive vs negative vs neutral not studied irrelevant fictitious/ hypothetical group: members of hypothetical groups Judgment (hypo group), impression formation task; illusory correlation task liking ratings, estimations of desirable and undesirable behaviors, recall of behaviors, latencies More directly test the influence of mood on the encoding processes underlying the illusory correlation effect. Again, only Ps in neutral moods formed illusory correlations and differentially attended to the minority group's infrequent behaviors. Equivalent levels of inaccuracy for negative and positive mood. Supports cognitive explanation for influence of affect. Did not support motivational explanation. Stroessner & Mackie (1992: study 1 of 2), PSPB, English 3 (mood: happy vs sad vs neutral) x 2 (stimulus material variability: high vs low) 106 students (68f, 38m) video mood induction (successful) happy vs neutral vs sad not studied irrelevant fictitious/ hypothetical group: members of hypothetical groups Judgment (hypo group), Read set of statements (high/low variability) about a group member, then make judgments about group member. judgments of general intragroup similarity, similarity in terms of sociability and intelligence, group sociability and intelligence Test influence of happy, sad and neutral mood on perceptions of the variability in social groups Both Ps in positive and negative mood were less aware of the variability in the heterogenous groups, although Ps in a neutral mood perceived differential variability for the groups. N/A

PAGE 97

87 Table 2.1. Continued. Short citation Design Ps Affect Manipulation Type of affect Arousal Level Relevance Target Social Judgment Task DVs Goal Influence of mood in Intergroup Context Model/Theory Endorsement Stroessner & Mackie (1992: study 2 of 2), PSPB, English 3 (mood: happy vs sad vs neutral) x 2 (stimulus material variability: high vs low) 58 female students video mood induction (successful) happy vs neutral vs sad not studied irrelevant general social group: sorority group members Judgment (social group), read set of statements (high or low variability) about a member of a sorority, then make judgments about that person. judgments of general intragroup similarity and group traits (sociability, intelligence, promiscuity, and irresponsibility ), free recall of all behaviors. Eliminate hypothesis that heightened flexibility inflated similarity judgments of the HV set when Ps in a pos mood were asked about similarities-so in exp 2, they were asked about differences. Also eliminate hypothesis that mood may have influenced judgments only because Ps read about a meaningless, unfamiliar, uninvolving group. Both Ps in positive and negative mood were less aware of the variability in the heterogenous groups, although Ps in a neutral mood perceived differential variability for the groups. However those in positive mood showed other indices of reduced processing (lower recall and reliance on stereotypes). Does not support positive mood and cognitive complexity. Stroessner & Mackie (1993: study 3 of 3), Chapter, English 3(mood: happy vs control vs sad) x 2 (exposure time: 3 seconds vs 7 seconds) students video mood induction (manipulation check not reported) happy vs neutral vs sad not studied irrelevant fictitious/ hypothetical group: members of hypothetical groups Judgment (hypo group), read set of statements about group member, then made judgments. judgments of similarity Test motivational (motive to maintain good mood) and cognitive (reduced processing capacity) explanations. Both Ps in positive and negative moods underestimated variability in limited time conditions, but detected variability more accurately when given sufficient time (at levels comparable to neutral mood conditions). Support for reduced processing hypothesis for positive and negative mood. Did not support motivational explanation.

PAGE 98

88 Table 2.2 Description of Video Mood Inductions Short citation Mood Video Description Abele, Gendolla & Petzold (1998: study 1 of 2) positive 8 minute humorous clip neutral novideo Abele, Gendolla & Petzold (1998: study 2 of 2) positive 8 minute humorous clip neutral novideo Burgess (1993: study 1 of 2) happy 910 minute clip of tv comedy show, Cheers sad 9-10 minute clip of dramatic movie, Steel Magnolias neutral 9-10 minute clip about an ma n with no short-term or long-term memory, Clive WearingA M an Without a Memory Dovidio, Gaertner, & Loux (2000: study 2 of 5) positive comedy clip neutral clip of a demons tration of wine corking Dovidio, Gaertner, & Loux (2000: study 3 of 5) positive comedy clip neutral clip of a demons tration of wine corking Dovidio, Gaertner, & Loux (2000: study 4 of 5) positive comedy clip neutral clip of a demons tration of wine corking Forgas (1992b: study 1 of 1) positive 10 minute clip from popular comedy show neutral 10 minute clip from a program on architecture negative 10 minute clip from a film dealing with death from cancer Forgas (1992c: study 1 of 3) positive 10 minute clip from popular comedy show neutral 10 minute clip from a program on architecture negative 10 minute clip from a film dealing with death from cancer

PAGE 99

89 Table 2.2. Continued. Short citation Mood Video Description Forgas (1992c: study 2 of 3) positive 10 minute clip from a different popular comedy show neutral 10 minute clip from a lecture on art negative 10 minute clip fr om a film depicting a severe sports injury Forgas & Fiedler (1996: study 1 of 3) positive humorous clip negative depressiveclip Forgas & Moylan (1991: study 1 of 1) happy 10 minute clip from popular comedy show neutral 10 minute clip from a program on architecture sad 10 minute clip from a film dealing with death from cancer Mackie, Hamilton, Schroth, Carlisle, Gersho, Meneses, Nedler, & Reichel (1989: study 1 of 2) happy clip from Saturday Night Live neutral clip of a demons tration of wine corking sad clip from interviews of children with cancer Mackie, Hamilton, Schroth, Carlisle, Gersho, Meneses, Nedler, & Reichel (1989: study 2 of 2) happy clip from Saturday Night Live neutral clip of a demons tration of wine corking sad clip from interviews of children with cancer Park, & Banaji (2000: study 1 of 3) happy clip from tv show, Late Night with David Letterman neutral clip from tv show, Wild, Wild, World of Animals Park, & Banaji (2000: study 2 of 3) happy clip from tv show, Late Night with David Letterman neutral clip from tv show, Wild, Wild, World of Animals Park, & Banaji (2000: study 3 of 3) happy clip from tv show, Candid Camera neutral clip from tv show, Wild, Wild, World of Animals sad clip from movie, Terms of Endearment

PAGE 100

90 Table 2.2. Continued. Short citation Mood Video Description Stroessner, Hamilton, & Mackie (1 992: study 1 of 2) positive 5 minute clip of a stand up comedian neutral 5 minute clip from tv show, National Geographic of a volcano site negative 5 minute clip discussing ch ild abuse that resulted in death Stroessner, Hamilton, & Mackie (1 992: study 2 of 2) positive 5 minute clip of a stand up comedian neutral 5 minute clip from tv show, National Geographic of a volcano site negative 5 minute clip discussing ch ild abuse that resulted in death Stroessner & Mackie (1992: study 1 of 2) happy 4 minute clip from a tv comedy neutral 4 minute clip in wine corking sad 4 minute clip from interviews of children with cancer Stroessner & Mackie (1992: study 2 of 2) happy 5 minute clip from tv show, Simpsons neutral 5 minute clip on histor y of Yellowstone National Park sad 5 minute clip from evening news Stroessner & Mackie (1993: study 3 of 3) happy 4 minute clip from a tv comedy neutral 4 minute clip in wine corking sad 4 minute clip from interviews of children with cancer

PAGE 101

91 Table 2.3 Description of Musical Mood Inductions Short citation Success of Manipulation Type of affect Type of affect Bodenhausen, Kramer & Susser (1994: study 3 of 4) successful happyexcited Borodin's Prince Igor happycalm Delius's A Village Romeo and Juliet ("The Walk Through Paradise Garden") neutral "Solar Winds" by Hykes Esses & Zanna (1995: study 2 of 4) successful positive 4 excerpts (e.g., Handel's Arrival of the Queen of Sheba) neutral 4 excerpts (e.g., Chopin's Nocturne #17) negative 4 excerpts (e.g., Tchaikovsky's Pathetique) Esses & Zanna (1995: study 3 of 4) successful positive 4 excerpts (e.g., Handel's Arrival of the Queen of Sheba) negative 4 excerpts (e.g., Tchaikovsky's Pathetique) Esses & Zanna (1995: study 4 of 4) partially successful positive 4 excerpts (e.g., Handel's Arrival of the Queen of Sheba) negative 4 excerpts (e.g., Tchaikovsky's Pathetique) Haddock, Zanna, & Esses (1994: study 1 of 1) partially successful positive 4 excerpts (e.g., Handel's Arrival of the Queen of Sheba) neutral 4 excerpts (e.g., Chopin's Nocturne #17) negative 4 excerpts (e.g., Tchaikovsky's Pathetique) Rodriguez Torres, R. (1998: study 1 of 1) successful positive Cheek to Cheek by I. Berlin negative El Cisne by M. Ravel

PAGE 102

CHAPTER 3 OVERVIEW Overview of the Present Study The present study examines the way in which affect valence, relevance, and arousal (through the effects on cognitive ability) jointly determine the influences of affect on intergroup judgments. Arousal governs the retrieval of detailed knowledge about the target one is judging, the retrieval of global structures such as stereotypes, and the emergence of affect identification and discounting that take part in the use of mood as information. Although arousal level is not directly examined, the effects of arousal are approximated by examining the influence of different levels of cognitive ability (by manipulating cognitive distraction). Valence influences the positivity of the material people retrieve from memory and the inferences people make when they use mood as information. Relevance influences the outcome of mood as information because people are likely to consider relevant information, but to discount irrelevant affect. The design is a 2 (mood induction: happy or angry) x 3 (distraction level: low, moderate, or high) x 2 (affect relevance: relevant or irrelevant) design with relevance manipulated within participants. Prior to the start of the experiment, participants will be randomly assigned to receive a happy or angry mood manipulation, and a low, moderate, or high distraction manipulation. 92

PAGE 103

93 The present experiment consisted of two sessions. The first session consisted of the mood induction. Participants viewed a photograph of an Asian-American or African-American man and were told to write a story about this character and a situation associated with this person that made them happy or angry, depending on the condition. All participants received one affect manipulation that served as both relevant and irrelevant affect given that they judged both a relevant and irrelevant target. For example, if mood was induced by writing a happy story about the Asian-American character, this mood was relevant to subsequent judgments of an Asian-American applicant, but irrelevant to judgments of an African-American applicant. After writing the story, participants responded to a number of questions about the task, including the mood manipulation check. After completing the first session, participants received detailed instructions about the second session in which they role-played being an employee of a roommate matching company called “Roommate Finders.” They were told that we were interested in comparing the effectiveness of traditional roommate matching procedures (e.g. personality testing) with an employee’s impressions based on applications. In addition, they were asked to repeatedly recite (and later produce) a number (the distraction manipulation) during other tasks, ostensibly to simulate the multi-tasking nature of office work. Numbers were 2, 4, or 6 digits long, depending on whether participants were assigned to the low, moderate or high distraction condition. Next, they reviewed the application of the first potential client and wrote a memo describing the applicant for a co-worker.

PAGE 104

94 Finally, the person completed a series of measures on the computer. They were told the first task was a measure of verbal aptitude (a word completion task), which was actually a measure of stereotype activation. After completing this measure, participants responded to a series of questions about their impressions of the applicant, which included the measures of stereotype application, endorsement of positive and negative non-stereotypic traits (priming effects) and attitudes toward the applicant. After completing judgments for the first applicant, participants repeated the entire exercise for a second potential applicant. The dependent variables were attitudes, stereotype activation scores, and stereotype application scores (positive and negative), endorsement of positive non-stereotypic traits (positive priming effects) and endorsement of negative non-stereotypic traits (negative priming effects). Each dependent variable was calculated for both Asian-American and African-American targets. For example, the first stereotype activation score measured the activation of stereotypes of Asian Americans, whereas the second stereotype activation score measured the activation of stereotypes of African Americans. Attitudes toward the Asian-American and African-American applicants were derived by summing four items that assessed overall attitudes toward each applicant. Stereotype activation scores for stereotypes of Asian Americans were calculated by summing the number of correct stereotypic word completions associated with Asian Americans. Stereotype activation scores for stereotypes of African Americans were calculated by summing the number of correct stereotypic word completions associated with African Americans.

PAGE 105

95 Positive stereotype application scores for African Americans were calculated by summing responses to the five positive stereotypic traits for African Americans on stereotypic traits for African Americans (e.g., musical). Conversely, positive stereotype application scores for Asian Americans were calculated by summing responses to the five positive stereotypic traits for Asian Americans on stereotypic traits for Asian Americans (e.g., responsible). Negative stereotype scores were calculated in the same fashion for dimensions that have undesirable connotations. Table 3.5 summarizes the various ways that mood was expected to influence stereotyping, but the mechanisms are addressed specifically in the following hypotheses. Review of the Hypotheses Hypothesis 1. Consistent with Gilbert and Hixon (1991), I predicted that high distraction would prevent stereotype activation. Thus, participants in high distraction conditions are expected to have lower scores on the measure of stereotype activation than participants in the moderate or low distraction conditions. The expected pattern of results can be seen in Table 3.1. Hypothesis 2 . As predicted by Gilbert and Hixon (1991), participants in high and low distraction conditions were expected to have lower stereotype application scores than participants in the moderate distraction conditions. As seen in Table 3.2, this prediction is in contrast with the predictions by Devine (1989). Hypothesis 3. Priming was expected to impact judgments of targets when participants are in low distraction conditions when ability is high and recall is easier for both relevant and irrelevant mood. Positive priming effects were assessed by summing the number of positive traits (e.g., considerate) that were endorsed for each applicant and

PAGE 106

96 negative priming effects were assessed by summing the number of negative traits (e.g., messy) that were endorsed for each applicant. The expected pattern for priming effects is in Table 3.3. Hypothesis 4. I predicted a curvilinear impact of distraction on the influence of irrelevant mood on attitudes (mood as information effects). Participants in low distraction conditions were not expected to use irrelevant mood, however their judgments may be biased by mood as a result of priming. For moderate and high distraction conditions, I expected a decrease in the influence of irrelevant mood from moderate to high levels of distraction. These effects should be apparent in people’s attitudes about the target being judged (Table 3.4). Hypothesis 5. Further, I predicted a linear impact of distraction on the influence of relevant mood on attitudes (mood as information effects). Initially, I predicted that at lower levels of distraction (in low and moderate distraction conditions), relevant mood would influence judgments. Under moderate levels of distraction, the influence of mood on judgments should be moderate. Finally, high distraction conditions should reduce the influence of mood altogether. Again, these effects should be apparent in people’s attitudes about the target being judged (Table 3.4).

PAGE 107

97 Table 3.1 Predictions for Stereotype Activation (Hypothesis 1) Completions for African-American Stereotypic Traits Completions for Asian-American Stereotypic Traits Distraction Condition African-American Character Asian-American Character African-American Character Asian-American Character Low +++ + + +++ Moderate ++ + + ++ High + + + +

PAGE 108

98 Table 3.2 Predictions for Stereotype Application (Hypothesis 2) Devine (1989) Gilbert & Hixon (1991) Low Distraction + + Moderate Distraction ++ ++ High Distraction +++ +

PAGE 109

99 Table 3.3 Predictions for Priming Effects (Hypothesis 3) Endorsement of Positive Traits Endorsement of Negative Traits Happy Angry Happy Angry Low Distraction +++ + --Moderate Distraction + + High Distraction + +

PAGE 110

100 Table 3.4 Predictions for Effects of Mood as Information on Attitudes (Hypothesis 4 and 5) Attitudes toward Candidate Irrelevant Mood Relevant Mood Low Distraction + +++ Moderate Distraction +++ ++ High Distraction + +

PAGE 111

101 Table 3.5 Summary of Mechanisms for Influence of Mood on Stereotyping Irrelevant Mood Relevant Mood Low Distraction Moderate Distraction High Distraction Low Distraction Moderate Distraction High Distraction Priming Mood as Info, Stereotyping No influence of mood Mood as Info, Priming Mood as Info, Stereotyping No influence of mood

PAGE 112

CHAPTER 4 METHOD Participants Seventy nine male (N = 31) and female (N = 48) participants completed the experiment in partial fulfillment of a course requirement. Participants were recruited from the General Psychology subject pool. Participants received partial credit toward fulfilling the experimental participation requirement for the General Psychology course. Design The design was a 2 (mood induction: happy or angry) x 3 (distraction level: low, moderate, or high) x 2 (affect relevance: relevant or irrelevant) design with relevance manipulated within participants. Materials Induction of affect. Participants were randomly assigned into one of two affective states (happy or angry) at one of two levels of relevance (relevant or irrelevant affect). In order to manipulate participants’ affective states, I used a modified a procedure of the induction developed by Schwarz and Clore (1983). Participants were told to write a story (which could be in the form of a letter to a friend) about a character (for specific instructions, see Appendix A and B). They were given a photograph of the character for the story, who was either an Asian-American man or an African-American man. 102

PAGE 113

103 Specifically, participants were told that for the purpose of their story, they should imagine that the character is a neighbor, and they should write about an imagined interaction with the character that could make them very happy or very angry. They were also instructed to write a story that maintains the tone from beginning to end and without resolving the affect (e.g., in the case of angry mood). As suggested by this description, there were four types of stories, depending on the affective tone of the story (happy or angry) and the ethnicity of the character of the story (Asian-American or African-American). Using this procedure, I expected to link affective reactions to one specific ethnicity, so that the affect would become relevant later when participants made a judgment of a different Asian-American or African-American target. For example, when participants were in the happy relevant mood condition, they were given a picture of an African-American male to be the main “character” for the happy story they wrote. Then, during a subsequent social judgment task (ostensibly part of an unrelated study), the participant was asked to make judgments about an African-American male. Thus, the happy mood that was induced by the story was relevant to the target being judged in the later social judgment task. The same mood, however, was presumably “irrelevant” for making a judgment of an Asian target immediately after writing the letter. Photos for affect induction. As part of the affect induction, participants were shown a photograph of an Asian man or a Black man (see Appendix C and D). Both photographs displayed the face and upper body of the subject. In each photograph, the subject was dressed casually in a blue shirt. Both men were clean cut and of average attractiveness. The expressions on the faces of both men were neutral to slightly positive.

PAGE 114

104 Participants were told to write a story based on the photograph that they viewed, wherein the main character in the story was the person portrayed in the photograph. Social judgment task. The second experiment was presented as a study on impression formation. For this part of the experiment, the participant was told that he or she was to role-play that he or she was an employee at Roommate Matchers, a company that offers roommate matching services (see Appendix E). The participant read the application of someone who was previously matched using typical roommate matching techniques. Then they made judgments about this applicant on a number of dimensions. The application that participants read either indicated that the person was African-American or Asian-American (see Appendix F and G). Aside from the ethnicity and name of the prospective roommate, and the student organizations listed (i.e., Asian Student Union vs. Black Student Union), the applications were identical in every other way. After looking the application over, the participant was also instructed to make some brief notes about the applicant for a “co-worker” (see Appendix H). Distraction. Participants were assigned to a low, moderate or high distraction condition. Distraction was manipulated through the use of a number that participants will be asked to remember (see Appendix I, J, K, L, M, and N) while they read the applications and engaged in the subsequent judgment task. They were told that they should remember this number in an effort to simulate the multi-tasking environment of an office. Participants in the low distraction condition were asked to remember a 2 digit number (32 for the first judgment task and 57 for the second judgment task). Participants in the moderate distraction condition were asked to recall a 4 digit number (3275 for the first judgment task and 5742 for the second judgment task). Finally, participants in the

PAGE 115

105 high distraction condition were asked to recall a 6 digit number (327549 for the first judgment task and 574218 for the second judgment task). Participants were asked to report the numbers at a later time during the experiment. Probe for suspicion. Participants were asked to describe what they think the researchers were interested in learning, in addition to any specific hypotheses they thought that researchers were testing (see Appendix O). Dependent measures Manipulation checks. After the first phase of the experiment (the story/mood induction), participants reported the extent to which they felt happy or angry after writing the story (see Appendix P). Responses to all items were made along a scale from 1 (not at all) to 9 (extremely). The participants also responded to a series of questions about the writing task, including the difficulty of writing the story, finding adequate words, and structuring sentences, as well as whether writing the story was interesting and fun. Stereotype activation. To measure stereotype activation, a modification of Gilbert and Hixon’s (1991) word fragment completion task was used. The word completion task required that the participant complete twenty-three words (See Appendix Q). Five of the word completions were stereotypically associated with Asian-Americans (e.g., shy, rice), whereas five of the word completions were stereotypically associated with African Americans (e.g., lazy, funny). The target words from the original stereotype activation measure Gilbert and Hixon (1991) only included words that are stereotypically associated with Asian Americans, but the modification used here also included words that are stereotypically associated with African Americans. The remaining word completions

PAGE 116

106 were fillers (e.g., talk, box). Stereotype activation was assessed by calculating the correct number of stereotypic completions. Stereotype Application. For each ethnic group, stereotype application (see Appendix R) was assessed by asking participants to rate the extent to which the stereotypic traits apply to the roommate applicant along scales ranging from 1 to 9 (e.g., not at all aggressive vs. extremely aggressive; not at all athletic vs. extremely athletic). Application of positive stereotypes for African Americans was calculated by combining responses to the five positive stereotypic traits for African Americans (e.g., fun-loving, athletic) and application of negative stereotypes for African Americans was calculated by combining responses to the five negative stereotypic traits for African Americans (e.g., aggressive, unintelligent). Similarly, application of positive stereotypes for Asian Americans was calculated by combining responses to the five positive stereotypic traits for Asian Americans (e.g., hard-working, intelligent) and application of negative stereotypes for Asian Americans was calculated by combining responses to the five negative stereotypic traits for Asian Americans (e.g., cold, arrogant). Stereotypic traits for African Americans and Asian Americans can be seen in Appendix S. Priming Effects . Participants were asked to rate the extent to which a series of traits apply to the roommate applicant along scales ranging from 1 to 9 (e.g., not at all incompetent vs. extremely incompetent; not at all flexible vs. extremely flexible). A score of six or higher indicated endorsement of a trait. Positive priming effects were assessed by calculating the number of non-stereotypic positive traits (out of the twelve most positive non-stereotypic traits) that were endorsed for each applicant (e.g., fashionable, sincere). Similarly, negative priming effects were assessed by calculating the number of

PAGE 117

107 non-stereotypic negative traits (out of the twelve most negative non-stereotypic traits) that were endorsed for each applicant (e.g., stingy, immoral). The twelve most positive and most negative non-stereotypic traits can be seen in Appendix T. Attitudes. Attitudes (see Appendix U) were assessed by asking participants to rate the roommate applicant along scales ranging from 1 to 9 (not at all likable vs. extremely likable; not at all pleasant vs. extremely pleasant; a bad match for a roommate vs. a good match for a roommate). Scale items were averaged and used as a summary index of attitude. Procedure Participants were run through the experiment one at a time. An experimenter greeted the participant when he or she arrived at the laboratory, and he or she was told that he or she was going to participate in two unrelated experiments combined in one session for practical reasons (i.e., to save time and participants), for which they would receive credit (see Appendix V). To facilitate this deception, the first part of the experiment was in one folder and was on pen-and-paper, whereas the second part of the experiment was computerized. Then he or she completed an informed consent form that briefly described the first experiment. In the first phase of the experiment, the participant viewed one of the photographs (the first phase of the mood induction) depending on the condition to which he or she was assigned. After viewing the appropriate photograph, the participant received instructions for the second phase of the mood induction, the story they wrote describing the photograph. Participants were told to write the story on a specific form (see Appendix W). After completing the story, the participant completed the mood manipulation checks.

PAGE 118

108 After the end of the first experiment, participants were given instructions for the second experiment. The cover story was that they were expected to role play being an employee of a roommate matching company because I was interested in comparing the effectiveness of traditional roommate matching procedures (e.g., extensive expensive personality testing, etc.) with a trained employee’s impressions of potential roommates. Participants were given a folder that included lengthy written instructions for the social judgment task and a separate page that had a number written on it (the distraction manipulation). The instructions explained that they should begin reciting the number immediately to simulate the busy, multi-tasking nature of an office environment and that they will be asked to produce the number at a later time. Additionally, the folder included the prospective roommate’s application for matching. Relevance was a within participants variable, thus all participants judged both a relevant and irrelevant target. If they previously wrote a story about their African-American neighbor, then the affect induced was relevant to judgments of the African-American candidate and irrelevant to the Asian-American candidate. Participants were told to read the application. First, the participant was asked to write a brief “memo” describing the prospective roommate to a fellow coworker. Second, based on the limited information presented in the application and following a brief reminder of the instructions for the social judgment task (see brief set of instructions in Appendix X), the participant was asked to respond to a series of questions about the applicant on the computerized Personality Questionnaire. These questions included a measure of stereotype activation, judgments of the

PAGE 119

109 prospective roommate, including attitudes, and the Modern Racism scale (McConahay, 1986, see Appendix Y). The stereotype activation measure consisted of a word completion task, presented on the pretense that the activity tested verbal aptitude, which is important for a roommate selection job. Words related to the Africanand Asian-American stereotypes (e.g., lazy, polite) were included to examine the facility with which participants complete these words. Then, the participant made a series of judgments of the prospective roommate, including the likelihood that the prospective roommate possessed various traits (stereotypic and non-stereotypic), and general attitudes toward the prospective roommate.

PAGE 120

CHAPTER 5 RESULTS Manipulation Checks To assess whether the affect manipulations were successful, I conducted an independent samples t-test on items assessing whether writing the story made the participant feel happy or angry. Results revealed that participants in the happy condition indicated that, on a scale from 0 to 9, they felt significantly happier (M = 5.05, SD =1.76) than did participants in the angry condition (M = 3.74, SD = 1.93), t (76) = 3.13, p < .01. Additionally, participants in the angry condition indicated that, on the same scale, they felt significantly angrier (M = 3.08, SD =2.55) than did participants in the happy condition (M = .67, SD = 1.55), t (75) = -5.03, p < .001. Thus, the affect manipulations were successful. Order Effects The order in which participants rated the Asian-American and African-American applicants had no effect on attitudes, endorsement of various non-stereotypic positive and negative traits for each applicant, application of any Asian-American stereotypes, or application of positive African-American stereotypes. Thus, all results presented are collapsed across order for these dependent variables. However, there was a significant order effect for African-American stereotype application, t (77) = 2.21, p < .05. When participants evaluated the African-American applicant after evaluating the Asian-American applicant, there was a significantly higher application of the negative African110

PAGE 121

111 American stereotype (M = 26.08, SD =4.93) than when the African-American applicant was rated first (M = 23.10, SD =6.89). Descriptive Statistics Reliability statistics, means, and standard deviations for the stereotype application, endorsement of various positive and negative personality traits, and attitude measures are presented in Table 5.1. Each of the measures was assessed twice (once for each candidate), except for Modern Racism (McConahay, 1986), which appeared at the end of the study. With the exception of the measure of application of African-American stereotypes, all measures demonstrated adequate reliability. Although the measures of application of positive and negative African-American stereotypes had rather modest internal consistency, the measures were used in analyses based on our predictions and the fact that reliability is rarely assessed for these measures. Hypothesis Tests Hypothesis 1: Stereotype Activation Consistent with Gilbert and Hixon (1991), participants in high distraction conditions should have lower scores on the measures of activation of Asian-American and African-American stereotypes than participants in the moderate or low distraction conditions. To test this hypothesis, the scores for stereotype activation were entered into two 3 x 2 x 2 MANOVAs with manipulated distraction level (low, moderate, or high), the ethnicity of the character in affect manipulation (Asian-American or African-American), and manipulated affect (happy or angry) as between subjects factors. The means for the activation of African-American and Asian-American stereotypes appear in Tables 5.2 and 5.3, respectively.

PAGE 122

112 Results revealed no significant main effect of distraction on activation of the African-American stereotype, F (2, 77) = 2.35, p = .64 or for activation of the Asian-American stereotype, F (2,77) = 0.45, p = .10. Thus, contrary to predictions, participants at low, moderate and high levels of distraction did not differ in their levels of stereotype activation. The pattern of means for Hypothesis 1 appears in Table 5.4. Also unexpectedly, results revealed a significant main effect of affect on activation of the African-American stereotype, F (1, 78) = 9.42, p < .01. Participants in angry moods had higher African-American stereotype activation scores (M = 2.80, SD =0.97) than participants in happy moods (M = 2.15, SD =0.78). Results, however, indicated there was no significant main effect of affect on activation of the Asian-American stereotype, F (1, 78) = 1.48, p = .23. Results revealed no significant main effect of the ethnicity of the character on activation of the African-American stereotype, F (1,78) = 3.36, p = .07 or for activation of the Asian-American stereotype, F (1,78) = 1.95, p = .17. Finally, there were no significant interactions among affect, distraction, or relevance for the activation of the Asian-American or African-American stereotypes. All Fs < 1. Further, to examine whether different levels of prejudice influence stereotype activation might have clouded the study results, the analyses were repeated, controlling for prejudice (i.e., modern racism). Mean scores of stereotype activation were entered into two 3 x 2 x 2 MANOVAs with manipulated distraction level (low, moderate, or high) , the ethnicity of the character in affect manipulation (Asian-American or African-American), manipulated affect (happy or angry) as between subjects factors and with

PAGE 123

113 modern racism as a covariate. Results revealed no significant main effect of distraction for activation of the African-American stereotype, F (2,77) = 0.18, p = .84 or for activation of the Asian-American stereotype, F (2,77) = 2.15, p = .13. Thus, controlling for levels of prejudice, participants at low (M = 2.39, SD =1.56), moderate (M = 2.50, SD =1.50), and high (M = 2.54, SD =1.57) levels of distraction had similar levels of activation of the African-American stereotype. Further, controlling for individual differences in prejudice, participants at low, moderate, and high levels of distraction also had similar levels of activation of the Asian-American stereotype. However, controlling for prejudice, results still revealed a significant main effect of affect on activation of the African-American stereotype, F (1,78) = 7.66, p < .01. Participants in angry moods (M = 2.76, SD =1.27) had higher scores of activation of the African-American stereotype than participants in happy moods (M = 2.20, SD =1.27). Also like before, results revealed no significant main effect of affect on activation of the Asian-American stereotype, F (1, 78) = 1.82, p = .18. Controlling for prejudice, results also revealed a significant main effect of the ethnicity of the character on activation of the African-American stereotype, F (1, 78) = 4.10, p < .05. Thus, participants who viewed a photograph of an African-American character had higher scores of activation of the African-American stereotype (M = 2.68, SD =1.25) than those who viewed a photograph of an Asian-American character (M = 2.27, SD =1.27). Results revealed no significant main effect of the ethnicity of the character on Asian-American stereotype activation, F (1,78) = 2.15, p = .15.

PAGE 124

114 Finally, controlling for prejudice, there were no significant interactions among affect, distraction or relevance for activation of either the Asian-American or African-American stereotypes. Hypothesis 2: Stereotype Application Consistent with Gilbert and Hixon (1991), participants in high and low distraction conditions should have lower stereotype scores than participants in moderate distraction conditions. To test this hypothesis, mean scores of application of positive and negative African-American and Asian-American stereotypes were entered into four 3 x 2 x 2 MANOVAs with manipulated distraction level (low, moderate, or high) , relevance (relevant or irrelevant affect), and manipulated affect (happy or angry) as between subjects factors. Means for the application of African-American and Asian-American stereotypes appear in Tables 5.5 and 5.6, respectively. Results revealed no significant main effect of distraction on application of a positive African-American stereotype, F (2,77) =0.49, p = .61 or on application of a negative African-American stereotype, F (2,77) = 0.41, p = .67. Similarly, results revealed no significant main effect of distraction on the application of a positive Asian-American stereotype, F (2,77) = 0.82, p = .44 or on the application of a negative Asian-American stereotype, F (2,77) = 0.11, p = .89. Thus, contrary to predictions, participants at different levels of distraction did not differ in their levels of stereotype application. The pattern of means for Hypothesis 2 appears in Table 5.7. Results revealed no significant main effect of affect on the application of a positive African-American stereotype, F (1,78) = 0.24, p = .63, or on the application of a negative African-American stereotype, F (1,78) = 1.00, p = .32. Further, results revealed

PAGE 125

115 no significant main effect of affect on the application of a positive Asian-American stereotype, F (1,78) = 0.13, p = .72 or on the application of a negative Asian-American stereotype, F (1,78) = 2.21, p = .14. However, results revealed a significant main effect of relevance on the application of a negative African-American stereotype, F (1,78) = 4.30, p < .05. Participants who viewed a (relevant) photograph of an African-American character during the affect manipulation had higher scores of application of a negative African-American stereotype (M = 26.05, SD =6.57) than those who viewed a(n) (irrelevant) photograph of an Asian-American character (M = 23.20, SD =5.37). However, results indicated no significant main effect of relevance on the application of a positive African-American stereotype, F (1,78) = 0.50, p = .48. Additionally, results revealed no significant main effect of relevance on the application of a positive Asian-American stereotype, F (1,78) = 1.48, p = .23, or on the application of a negative Asian-American stereotype, F (1,78) = 0.01, p = .94. Finally, there were no significant interactions involving affect, distraction, or relevance for any of the four stereotype application measures. All Fs. < 2.50. As with stereotype activation, the analyses were repeated, controlling for prejudice (i.e., modern racism). Again, mean scores of stereotype application were entered into two 3 x 2 x 2 MANOVAs with manipulated distraction level (low, moderate, or high), the ethnicity of the character in the affect manipulation (Asian-American or African-American), and manipulated affect (happy or angry) as between subjects factors, and with modern racism as a covariate. These analyses produced similar results to the analyses without the covariates. The only significant main effect, controlling for

PAGE 126

116 prejudice was a significant main effect of relevance on the application of a negative African-American stereotype, F (1,78) = 5.73, p < .05. Thus, participants who viewed a (relevant) photograph of an African-American character had higher scores of application of a negative African-American stereotype (M = 26.22, SD =8.59) than those who viewed an (irrelevant) photograph of an Asian-American character (M = 22.95, SD =8.51). Further, controlling for prejudice, there were no significant interactions involving affect, distraction or relevance for the application of positive or negative Asian-American stereotypes or the application of positive or negative African-American stereotypes. Hypothesis 3: Priming Effects When participants are in low distraction conditions (and ability is high) and thus recall is easy, both negative and positive moods may prime the recall of valence-consistent materials. Thus, positive mood should lead to greater endorsement of non-stereotypic positive traits about the person one is judging, whereas negative mood should lead to greater endorsement of non-stereotypic negative traits about the person one is judging. To test this hypothesis, the mean number of endorsed positive and negative traits were entered into two 3 x 2 x 2 MANOVAs with manipulated distraction level (low, moderate, or high), relevance (relevant or irrelevant affect), and manipulated affect (happy or angry) as between subjects factors. Means for the trait-endorsement measures for African-American and Asian-American candidates appear in Tables 5.8 and 5.9, respectively. Results revealed no significant main effect of distraction on positive trait endorsement for the African-American candidate, F (2,77) =0.42, p = .66, or on negative trait endorsement for the African-American candidate, F (2,77) = 0.58, p = .57. Similarly,

PAGE 127

117 results revealed no significant main effect of distraction on positive trait endorsement for the Asian-American candidate, F (2,77) = 0.40, p = .67, or on negative trait endorsement for the Asian-American candidate, F (2,77) = 0.45, p = .64. Thus, contrary to predictions, participants at different levels of distraction did not differ in their levels of trait endorsement. The pattern of means for Hypothesis 3 appears in Tables 5.10 and 5.11. Results revealed no significant main effect of affect on positive trait endorsement for the African-American candidate, F (1,78) = 0.42, p = .52, or on negative trait endorsement for the African-American candidate, F (1,78) = 0.97, p = .33. Further, results revealed no significant main effect of affect on positive trait endorsement for the Asian-American candidate, F (1,78) = 0.24, p = .63, or on negative trait endorsement for the Asian-American candidate, F (1,78) = 2.92, p = .09. However, results revealed a significant main effect of relevance for negative trait endorsement for the African-American candidate, F (1,78) = 6.87, p < .05. Participants who viewed a (relevant) photograph of an African-American character during the affect manipulation endorsed more negative characteristics for the African-American candidate (M = 6.03, SD =5.10) than those who viewed an (irrelevant) photograph of an Asian-American character (M = 3.43, SD =3.25). In contrast, results indicated no significant main effect of relevance for positive trait endorsement for the African-American candidate, F (1,78) = 0.01, p = .93. Furthermore, results revealed no significant main effect of relevance on positive trait endorsement for the Asian-American candidate, F (1,78) = 0.46, p = .50 or negative trait endorsement for the Asian-American candidate, F (1,78) = 0.53, p = .47.

PAGE 128

118 Finally, there were no significant interactions involving affect, distraction, or relevance for any of the four trait endorsement measures. All Fs < 2.04. As with the previous dependent variables, to partial out variance coming from individual differences, the analyses were repeated, controlling for prejudice (i.e., modern racism). Again, the mean scores of trait endorsement measures were entered into two 3 x 2 x 2 MANOVAs with manipulated distraction level (low, moderate, or high), the ethnicity of the character in the affect manipulation (Asian-American or African-American), and manipulated affect (happy or angry) as between subjects factors, and with modern racism as a covariate. These analyses replicated the results from the analyses without the covariate. Controlling for individual differences in prejudice, the only significant main effect was an effect of relevance on negative trait endorsement for the African-American candidate, F (1,78) = 9.37, p < .01. Thus, participants who viewed a (relevant) photograph of an African-American character during the affect manipulation endorsed more negative characteristics for the African-American candidate (M = 6.13, SD =5.97) than those who viewed an (irrelevant) photograph of an Asian-American character (M = 3.23, SD =5.92). Further, controlling for prejudice, there were no significant interactions among affect, distraction or relevance for positive or negative trait endorsement for either the Asian-American or the African-American candidate. Hypotheses 4 and 5: Attitudes Hypotheses 4 and 5 described different patterns of influence of affect as a function of distraction and affect relevance. Participants in low and high distraction conditions should not use irrelevant mood as information, although their judgments may be biased by mood as a result of priming. In contrast, participants in moderate distraction

PAGE 129

119 conditions should be more influenced by irrelevant mood than participants in either low or high distraction situations. These effects should demonstrate a curvilinear impact of distraction on the influence of irrelevant mood on attitudes. With respect to relevant affect, one should observe a linear impact of distraction on the influence of relevant mood on attitudes. At low and moderate levels of distraction, relevant mood should increasingly influence attitude judgments, whereas under high distraction conditions, the influence on attitudes should be reduced altogether. To test these hypotheses, mean scores of attitudes were entered into two 3 x 2 x 2 MANOVAs with manipulated distraction level (low, moderate, or high), relevance (relevant or irrelevant affect), and manipulated affect (happy or angry) as between subjects factors.. Means for attitudes toward the African-American and Asian-American candidates appear in Tables 5.12 and 5.13, respectively. Results revealed no significant main effect of distraction on attitudes toward the African-American candidate, F (2,77) =0.80, p = .46 or attitudes toward the Asian-American candidate, F (2,77) = 0.69, p = .51. Similarly, results revealed no significant main effect of affect on attitudes toward the African-American candidate, F (1,78) = 0.06, p = .80 or attitudes toward the Asian-American candidate, F (1,78) = 0.16, p = .69, or of relevance on attitudes toward the African-American candidate, F (1,78) = 0.64, p = .43 or attitudes toward the Asian-American candidate, F (1,78) = 0.39, p = .54. Most importantly based on this thesis’ hypotheses, there were no significant interactions involving affect, distraction, or relevance for attitudes towards the Asian-American candidate or attitudes toward the African-American candidate. All Fs < 2.00. The pattern of means for Hypothesis 4 and 5 appears in Table 5.14.

PAGE 130

120 As with the previous dependent variables, the analyses were repeated, controlling for prejudice (i.e., modern racism). Again, mean scores of attitudes were entered into two 3 x 2 x 2 MANOVAs with manipulated distraction level (low, moderate, or high), the ethnicity of the character in affect manipulation (Asian-American or African-American), and manipulated affect (happy or angry) as between subjects factors and with modern racism as a covariate. Like the previous analyses, these produced no significant main or interactive effects of the manipulations on attitudes toward the Asian-American or the African-American candidate.

PAGE 131

121 Table 5.1 Descriptive Statistics for Measures _______________________________________________________________________ Cronbach’s Standard alpha Mean deviation Scale Asian-American Positive Stereotype Application (1st) .79 28.71 5.50 Asian-American Positive Stereotype Application (2nd) .88 29.37 6.58 African-American Positive Stereotype Application (1st) .55 24.08 4.74 African-American Positive Stereotype Application (2nd) .62 23.04 4.89 Asian-American Negative Stereotype Application (1st) .73 26.14 5.14 Asian-American Negative Stereotype Application (2nd) .68 26.67 5.19 African-American Negative Stereotype Application (1st) .81 21.58 5.85 African-American Negative Stereotype Application (2nd) .85 22.81 6.92 Endorsement of Positive Traits for Candidate (1st) .88 9.89 5.63 Endorsement of Positive Traits for Candidate (2nd) .83 10.48 5.45 Endorsement of Negative Traits for Candidate (1st) .86 4.27 3.90 Endorsement of Negative Traits for Candidate (2nd) .86 3.95 3.75 Attitudes toward Candidate (1st) .90 25.14 5.50 Attitudes toward Candidate (2nd) .84 24.89 4.26 Modern Racism .80 12.91 4.78 ________________________________________________________________________

PAGE 132

122 Table 5.2 Means for Activation of African-American Stereotypes Distraction Affect African-American Asian-American Condition Condition Character Character M SD M SD Low Happy 2.50 (0.84) 1.88 (0.64) Angry 2.83 (0.41) 2.17 (1.33) Moderate Happy 2.33 (0.82) 2.00 (1.10) Angry 2.88 (1.25) 2.88 (0.64) High Happy 2.29 (0.76) 2.00 (0.63) Angry 3.17 (1.17) 2.83 (0.93) ________________________________________________________________________

PAGE 133

123 Table 5.3 Means for Activation of Asian-American Stereotypes Distraction Affect African-American Asian-American Condition Condition Character Character M SD M SD Low Happy 1.67 (1.21) 1.38 (1.19) Angry 1.50 (1.23) 1.50 (1.05) Moderate Happy 2.83 (0.41) 2.00 (1.27) Angry 1.50 (0.76) 1.88 (1.13) High Happy 1.86 (0.90) 1.33 (0.82) Angry 1.83 (0.98) 1.17 (0.98) ________________________________________________________________________

PAGE 134

124 Table 5.4 Means for Stereotype Activation (Hypothesis 1) Completions for AfricanCompletions for AsianAmerican Stereotypic Traits American Stereotypic Traits ________________________________________________________________________ AfricanAsianAfricanAsianAmerican American American American Character Character Character Character M SD M SD M SD M SD Low 2.67 (0.65) 2.00 (0.96) 1.58 (1.17) 1.43 (1.09) Distraction Moderate 2.64 (1.08) 2.50 (0.94) 2.07 (0.92) 1.93 (1.14) Distraction High 2.69 (1.03) 2.42 (0.79) 1.85 (0.90) 1.55 (1.06) Distraction ________________________________________________________________________

PAGE 135

125 Table 5.5 Means for Application of African-American Stereotypes Positive Stereotypes Negative Stereotypes ________________________________________________________________________________________________ Distraction Affect Relevant Irrelevant Relevant Irrelevant Condition Condition Affect Affect Affect Affect M SD M SD M SD M SD Low Happy 25.00 (2.76) 27.38 (7.75) 27.33 (3.39) 23.50 (9.10) Angry 26.83 (8.89) 28.50 (4.81) 27.17 (6.37) 23.83 (2.79) Moderate Happy 29.00 (6.84) 25.33 (3.93) 23.00 (7.07) 22.33 (6.38) Angry 27.88 (7.22) 27.25 (5.06) 26.13 (7.42) 24.25 (3.85) High Happy 29.29 (5.56) 27.67 (3.08) 25.00 (6.16) 22.17 (4.62) Angry 30.33 (5.57) 26.67 (1.63) 27.83 (9.15) 22.67 (3.62) ________________________________________________________________________________________________

PAGE 136

126 Table 5.6 Means for Application of Asian-American Stereotypes Positive Stereotypes Negative Stereotypes ________________________________________________________________________________________________ Distraction Affect Relevant Irrelevant Relevant Irrelevant Condition Condition Affect Affect Affect Affect M SD M SD M SD M SD Low Happy 28.63 (7.96) 30.83 (6.01) 22.75 (8.84) 25.17 (4.12) Angry 28.00 (4.56) 28.50 (7.50) 25.17 (1.94) 22.00 (2.90) Moderate Happy 27.83 (4.12) 33.17 (6.68) 21.00 (3.63) 23.50 (8.98) Angry 31.25 (6.56) 32.63 (5.07) 25.75 (5.39) 26.50 (4.50) High Happy 29.33 (2.42) 33.29 (8.26) 22.33 (4.23) 22.86 (5.24) Angry 31.00 (5.02) 28.50 (9.10) 25.67 (1.21) 23.17 (3.97) ________________________________________________________________________________________________

PAGE 137

127 Table 5.7 Means for Stereotype Appl ication (Hypothesis 2) ________________________________________________________________________ Completions for AfricanCompletions for AsianAmerican Stereotypic Traits American Stereotypic Traits ________________________________________________________________________ Positive Negative Positive Negative AfricanAfricanAsianAsianAmerican American American American Stereotypes Stereotype s Stereotypes Stereotypes M SD M SD M SD M SD Low 26.96 (6.35) 25.31 (6.21) 28.96 (6.45) 23.69 (5.45) Distraction Moderate 27.39 (5.77) 24.11 (6.09) 31.32 (5.76) 24.46 (5.90) Distraction High 28.52 (4.33) 24.44 (6.26) 30.64 (6.66) 23.48 (3.98) Distraction

PAGE 138

128 Table 5.8 Means for Endorsement of Nonstereotypic Traits for th e African-American Candi date (Priming Effects) ________________________________________________________________________________________________ Endorsement of Positive Traits Endorsement of Negative Traits ________________________________________________________________________________________________ Distraction Affect Relevant Irrelevant Relevant Irrelevant Condition Condition Affect Affect Affect Affect M SD M SD M SD M SD Low Happy 5.50 (3.02) 5.50 (4.04) 3.50 (2.43) 2.25 (2.96) Angry 6.50 (2.17) 4.33 (1.75) 4.17 (2.86) 2.33 (1.51) Moderate Happy 5.67 (2.81) 5.00 (3.35) 1.67 (1.63) 2.17 (1.84) Angry 5.13 (4.22) 8.00 (2.73) 3.38 (4.14) 2.25 (1.28) High Happy 6.29 (4.72) 5.17 (4.07) 3.29 (2.87) 0.83 (1.60) Angry 5.33 (3.01) 5.83 (4.36) 3.33 (3.08) 1.67 (1.37)

PAGE 139

129 Table 5.9 Means for Endorsement of Nonstereotypic Traits for the Asian-Am erican Candidate (Priming Effects) Endorsement of Positive Traits Endorsement of Negative Traits ________________________________________________________________________________________________ Distraction Affect Relevant Irrelevant Relevant Irrelevant Condition Condition Affect Affect Affect Affect M SD M SD M SD M SD Low Happy 5.00 (4.00) 5.50 (4.09) 1.38 (2.13) 1.17 (0.98) Angry 4.17 (1.94) 7.00 (2.76) 2.67 (1.37) 1.50 (1.05) Moderate Happy 5.83 (4.12) 4.83 (3.37) 1.67 (1.21) 2.67 (2.94) Angry 6.88 (2.80) 5.00 (3.67) 2.25 (1.17) 1.63 (1.60) High Happy 5.83 (3.55) 5.57 (4.72) 0.50 (0.55) 1.57 (1.72) Angry 5.17 (4.36) 6.33 (2.88) 2.50 (1.05) 2.50 (2.17)

PAGE 140

130 Table 5.10 Means for Priming Effects for Afri can-American Candidate (Hypothesis 3) ________________________________________________________________________ Positive Trait Endorsement Negative Trait Endorsement ________________________________________________________________________ Happy Angry Happy Angry M SD M SD M SD M SD Low 5.50 (3.50) 5.42 (2.19) 2.79 (2.72) 3.25 (2.38) Distraction Moderate 5.33 (2.96) 6.56 (3.74) 1.92 (1.68) 2.81 (3.02) Distraction High 5.77 (4.29) 5.58 (3.58) 2.15 (2.61) 2.50 (2.43) Distraction ________________________________________________________________________

PAGE 141

131 Table 5.11 Means for Priming Effects for As ian-American Candidate (Hypothesis 3) ________________________________________________________________________ Positive Trait Endorsement Negative Trait Endorsement ________________________________________________________________________ Happy Angry Happy Angry M SD M SD M SD M SD Low 5.21 (3.89) 5.58 (2.71) 1.29 (1.68) 2.08 (1.31) Distraction Moderate 5.33 (3.63) 5.94 (3.30) 2.17 (2.21) 1.94 (1.39) Distraction High 5.69 (4.05) 5.75 (3.57) 1.08 (1.38) 2.50 (1.62) Distraction ________________________________________________________________________

PAGE 142

132 Table 5.12 Means for Attitudes toward African-American Candidate ________________________________________________________________________ Distraction Affect African-American Asian-American Condition Condition Character Character ________________________________________________________________________ M SD M SD Low Happy 23.00 (3.23) 26.38 (7.05) Angry 23.83 (0.98) 22.83 (3.13) Moderate Happy 25.00 (3.69) 25.50 (3.73) Angry 21.88 (7.75) 26.50 (3.67) High Happy 25.43 (7.07) 24.67 (5.05) Angry 27.17 (5.00) 26.00 (4.38) ________________________________________________________________________

PAGE 143

133 Table 5.13 Means for Attitudes toward Asian-American Candidate Distraction Affect African-American Asian-American Condition Condition Character Character ________________________________________________________________________ M SD M SD Low Happy 25.17 (4.54) 25.13 (8.58) Angry 25.50 (3.27) 22.67 (3.33) Moderate Happy 24.00 (3.35) 26.67 (3.50) Angry 23.00 (4.54) 26.13 (3.72) High Happy 25.86 (5.21) 26.00 (5.76) Angry 25.83 (4.79) 27.00 (5.18) ________________________________________________________________________

PAGE 144

134 Table 5.14 Means for Effects of Mood as Information on Attitudes (Hypothesis 4 and 5) ________________________________________________________________________ Attitudes toward AfricanAttitudes toward AsianAmerican Candidate American Candidate ________________________________________________________________________ AfricanAsianAfricanAsianAmerican American American American Character Character Character Character (Relevant) (Irrelevant) (Relevant) (Irrelevant) M SD M SD M SD M SD Low 23.42 (2.31) 24.86 (5.82) 24.07 (6.74) 25.33 (3.77) Distraction Moderate 23.21 (6.34) 26.07 (3.58) 26.36 (3.50) 23.43 (3.96) Distraction High 26.23 (6.02) 25.33 (4.56) 26.50 (5.25) 25.85 (4.81) Distraction ________________________________________________________________________

PAGE 145

CHAPTER 6 DISCUSSION Study Rationale and Summary of Results The first purpose of this dissertation was to examine the influence of cognitive ability on stereotype activation and application. Consistent with Gilbert and Hixon (1991), I predicted that high distraction will prevent stereotype activation, (in contrast with Devine’s [1989] prediction that activation occurs automatically when a person merely encounters a member of a stereotyped group). Thus participants in high distraction conditions were expected to have lower scores on the measure of stereotype activation than participants in the moderate or low distraction conditions. Further, as predicted by Gilbert and Hixon (1991), participants in high and low distraction conditions were expected to have lower scores on measures of stereotype application than participants in the moderate distraction conditions. A second purpose of this research was to examine the influence of mood on mood congruent recall and interpretation (priming effects). I hypothesized that priming would impact judgments of targets when participants were in low distraction conditions (when ability is high) and thus facilitate recall of mood-congruent material, and that this effect should be apparent for both relevant and irrelevant moods. Thus, under conditions of low distraction, a positive mood should lead to positive priming or endorsement of positive traits, whereas negative mood would lead to negative priming or endorsement of negative traits. 135

PAGE 146

136 A final purpose of this dissertation was to test two mechanisms (i.e., mood as information and mood priming) through which mood may impact the process of intergroup judgments. Specifically, consistent with the model of affect identification and discounting (Albarracn & Kumkale, 2003), I predicted a curvilinear impact of ability on the influence of irrelevant mood on attitudes. Participants in low distraction conditions were not expected to use irrelevant mood as information because ability should have been sufficiently high to allow them to discount their mood. For moderate and high distraction conditions, I expected a decrease in the influence of irrelevant mood due to increased inability to identify the experienced mood. Further, I predicted a linear impact of ability on the influence of relevant mood on attitudes. I predicted that lower levels of distraction would allow people to realize that the mood is relevant to their judgment (in low and moderate distraction conditions), and thus relevant mood should influence judgments. Higher levels of distraction, however, should reduce the impact of mood by interfering with their ability to judge the mood as relevant, and thus the influence of mood on judgments should decrease in a linear way. Findings about Stereotype Activation In stark contrast to the first hypothesis of this thesis, low, moderate and high levels of distraction yielded similar levels of activation of both the Asian-American and the African-American stereotypes. Despite the lack of support for the first hypothesis, some significant effects emerged for activation of African-American stereotypes. First, participants in angry moods had greater levels of activation of the African-American stereotype than those in happy moods, and this effect remained significant even after controlling for individual differences in prejudice.

PAGE 147

137 Another interesting effect that emerged after controlling for individual differences in prejudice was that participants who viewed a photograph of an African-American person had higher levels of activation of the African-American stereotype than those who viewed a photograph of an Asian-American character. This effect, which was only marginally significant without controlling for prejudice, appears to render some support for Devine’s (1989) assumption that merely encountering a member of a stereotyped group can activate stereotypes for that group. This effect, however, appears to be the case only for African Americans. Findings for Stereotype Application This thesis results also failed to support the second hypothesis because participants who experienced low, moderate and high levels of distraction yielded similar levels of stereotype application for both the Asian-American and the African-American stereotypes. Despite the lack of support for the second hypothesis, one significant effect emerged for the application of negative African-American stereotypes. First, like was the case for stereotype activation, participants who viewed a (relevant) photograph of an African-American person had higher application of a negative African-American stereotype than those who viewed an (irrelevant) photograph of an Asian-American person. This effect, which remained significant when controlling for prejudice, suggests that, even when participants reflected on a positive encounter with an African-American neighbor, this reflection led to greater endorsement of negative stereotypes about African Americans.

PAGE 148

138 Findings for Endorsement of Various, Non-Stereotype-Specific Traits The third hypothesis was also not supported. Thus, positive mood did not lead to positive priming or endorsement of positive traits and negative mood did not lead to negative priming or endorsement of negative traits under conditions of low distraction. Despite the lack of support for this hypothesis, one significant effect emerged for the endorsement of negative traits about the African-American candidate. Participants who saw a (relevant) photograph of an African-American person endorsed more negative traits for the African-American candidate than those who viewed an (irrelevant) photo of an Asian-American person. This effect remained significant when controlling for prejudice. Once again, independent of positive or negative mood, the relevant-mood induction led to greater endorsement of negative traits for the African-American candidate. Thus, simply reflecting on an encounter with an African-American person (whether positive or negative) led participants to endorse more negative traits when evaluating an African-American candidate later. Findings for Attitudes Results also failed to support the fourth hypothesis that the attitudes of participants in low and high distraction conditions would be less influenced by irrelevant mood relative to those in moderate distraction conditions (demonstrating a curvilinear relationship). Additionally, results failed to support the fifth hypothesis that relevant mood would increasingly influence the attitudes of participants in low, moderate and high levels of distraction (demonstrating a linear relationship).

PAGE 149

139 Explanations for the Study’s Null Findings and Unexpected Results Several possible explanations exist for the failure to support the primary hypotheses. First, the present experiment may not have been able to detect the influence of distraction on stereotype activation and stereotype application simply because the manipulation of distraction may not have been effective. In the absence of a manipulation check for distraction, it is difficult to determine whether or not participants who repeatedly recited two, four or six digit numbers actually experienced significantly different low, moderate and high levels of distraction. The fact that similar rehearsal tasks have successfully been used in other studies (e.g., Gilbert & Hixon, 1991; Gilbert & Osborne, 1989) to manipulate cognitive ability suggests that the manipulation may have been appropriate. However, this suggestion would need to be verified with a manipulation check. One notable difference between the rehearsal task used in this study is that I attempted to manipulate three levels of distraction (low, moderate, and high) using two, four or six digit numbers rather than manipulating two levels of distraction (distracted or not distracted) using an eight digit number or no rehearsal task at all. Second, although the measure of stereotype activation has been successfully used in previous studies (e.g., Gilbert & Hixon, 1991), the present experiment may not have revealed differences for stereotype activation because it did not include a sufficiently sensitive measure of stereotype activation. An implicit measure of stereotype activation may have been more appropriate if one wished to conduct further research in this area (see e.g., Greenwald, Nosek, & Banaji, 2003; Nosek, Greenwald & Banaji, 2005). Third, the experiment may have failed to replicate previous patterns of stereotype application because of the context of the judgments requested in the study. Because

PAGE 150

140 participants believed that they were evaluating two candidates for potential selection as roommates, the context may have emphasized the need to be impartial in their evaluations of the applicants. Nonetheless, given that (through the impact of relevance) the African-American neighbor in the mood manipulation subsequently influenced the activation and application of the African-American stereotype as well as endorsement of more negative traits, bias nevertheless entered into the picture. Another interesting pattern that was not predicted was the significant order effect for African-American stereotype application. When participants evaluated the African-American applicant after evaluating the Asian-American applicant, they endorsed significantly more negative stereotypes about African Americans than when the African-American applicant was evaluated first. The fact that ratings of the African-American candidate became more negative following the evaluation of an Asian-American candidate, whereas the evaluation of the Asian-American candidate was not influenced by order, suggests a negativity bias for African Americans wherein African Americans are hurt by the comparison, although Asian Americans are neither hurt nor helped. Directions for Future Research The results of the current study raise a number of questions for future research. First, future research should address some of the limitations of the current study. As mentioned previously, it is possible that the distraction manipulation was not effective in creating different levels of distraction. Thus, it may be beneficial to implement a different manipulation of distraction. One possibility is that the numbers used to create the low, moderate and high levels of distraction did not differ enough to create significantly different levels of distraction. If an eight digit number is used in a typical rehearsal task

PAGE 151

141 (Gilbert & Hixon, 1991; Pendry & Macrae, 1999) to create a basic level of distraction, perhaps an eight digit number should be used to represent moderate levels of distraction and a six and ten digit number should be used to create low and high levels of distraction. Another possibility would be to have participants memorize word lists of different lengths which they would be asked to recall at a later point in the experiment (see Drolet & Luce, 2004; Shiv & Huber, 2000; Ward & Mann, 2000). In addition, it is possible that the timing of the distraction manipulation should be modified. Perhaps the distraction manipulation should be implemented during the affect manipulation when stereotype activation presumably occurs. In this study, the distraction manipulation began during the second phase of the experiment (during the stereotype activation measure) to maintain the perception that participants were engaged in two distinct experiments. However, to truly test whether cognitive distraction impedes stereotype activation, it may be necessary to initiate the distraction manipulation when participants are initially exposed to the relevant mood manipulation. Moreover, it is possible that a different measure of stereotype activation may have been appropriate. For example, using a computerized implicit measure of stereotype activation (Greenwald, Nosek, & Banaji, 2003; Nosek, Greenwald & Banaji, 2005) may be more successful in detecting subtle differences in stereotype activation than the word completion task utilized in the current study. Further, the failure to detect stereotype activation may have resulted directly from the affect manipulation. As part of the affect manipulation, participants were given a photograph of an African-American or an Asian-American character and told to write a story about a happy or angry interaction with this character. The photograph was

PAGE 152

142 expected to activate stereotypes which w ould then be detected by the stereotype activation measure. However, it is possible that as particip ants wrote the story about the character, they applied stereotypes within th e context of the story. This may explain why the measure did not subsequently detect diffe rential activation of stereotypes, because participants had already had th e opportunity to actively apply stereotypes as part of the writing task.1 Thus, it may be necessary to employ an affect manipulation that manages to activate stereotypes, but doesn’t afford pa rticipants the opportuni ty to actively apply stereotypes. For example, participants coul d view film clips in which a members of the target ethnic groups perform monologues th at induce the intended mood states (e.g., stand up comedy, an anger-inducing speech, a sad story). Another modification of the current st udy that may prove fruitful would be changing the social judgment context utiliz ed in the study. Rather than the roommate selection task that was used in the current study, perhaps it would be advantageous to use a general impression formation task such as thos e that have been successfully used in past studies (Bless, Schwarz & Weiland, 1996; Bu rgess, 1993; Forgas, 1992b; Forgas, 1992c; Forgas & Moylan, 1991; Jackson & Sullivan, 20 01; Lambert et al., 1997; Mackie et al., 1989; Moreno & Bodenhausen, 2001; Quelle r et al., 1996; Rodriguez Torres, 1998; Sheppard, 1996; Stroessner & Mackie, 1992). Some of these may reduce the fairness norms that might have been prevalent in our study. Second, the current study unexpectedly de monstrated that when an AfricanAmerican applicant was evaluated after rather than before an Asian-American applicant, 1 Two exceptions to this statement include th e main effect of affect on activation of African American stereotypes and the main ef fect of relevance on activation of African American stereotypes (when controlling fo r individual differences in prejudice).

PAGE 153

143 participants endorsed significantly more negative stereotypes of African Americans. Future research may investigate this negative bias for African Americans wherein African Americans are negatively impacted by the comparison with an Asian American, although evaluations of Asian Americans are unaffected. One possible interpretation is that despite a general awareness of stereotypes of Asian Americans, beliefs in the model minority myth neutralize any negative impact of those stereotypes on evaluations of Asian Americans. Similarly but much more unfortunately, the relative absence of positive stereotypes about African Americans may neutralize any positive impact of those stereotypes on African Americans. In an investigation of the model minority myth, Wong, Lai, Nagasawa, and Lin (1998) found that European Americans, African Americans, Latino/a Americans, and Native Americans all reported that they evaluated Asian Americans as superior to their own ethnic groups on a number of traits. In the current study, the comparison with an Asian American may make the absence of such positive global evaluations for African Americans more salient, leading to increased stereotype application for African Americans. In a related study, Dunn and Spellman (2003) demonstrated that stereotypes about a target member of a stereotyped group could be inhibited by having participants rehearse alternative aspects (e.g., individuating information) of the target member’s identity. Further, when participants practiced stereotypic information about the target member, they demonstrated inhibited memory for traits associated with another stereotypic aspect of the person’s identity. Thus, it is possible that in the case of Asian Americans, positive stereotypic aspects (model minority beliefs) may have inhibited the application of negative stereotypic beliefs.

PAGE 154

144 Third, this thesis showed that viewing a photo of an African American led to greater activation of African-American stereotypes, greater application of negative African-American stereotypes, and greater endorsement of non-stereotypic negative traits. This effect is particularly shocking because viewing the photo of an African American produced negative reactions about other African Americans, even when participants were asked to imagine a positive experience with the African American they viewed in the first place (positive relevant affect conditions). Although this finding is consistent with previous research that positive mood is generally associated with an increased reliance on stereotypes (Bodenhausen et al., 2001; Mackie & Worth, 1989), it is interesting that this effect was only demonstrated for African Americans. Future research could further explore the differential roles of relevant affect for different ethnic groups, including African Americans and Asian Americans. Conclusion My study primarily provided support for the role of relevant affect in intergroup judgments. When controlling for individual differences in prejudice, participants who viewed a (relevant) photograph of an African-American person as part of the affect manipulation had higher levels activation of African-American stereotypes than those who viewed an (irrelevant) photograph of an Asian-American person as part of the affect manipulation. With and without controlling for individual differences in prejudice, participants who saw the relevant photo of an African-American person in relation to the affect manipulation also demonstrated greater application of negative African-American stereotypes and greater endorsement of non-stereotypic negative traits for the African-American candidate. Interestingly, the role of relevance seemed to be primarily negative

PAGE 155

145 for African Americans. When participants viewed a relevant photo of an African American, it led to greater activation of African-American stereotypes, greater application of negative African-American stereotypes and greater endorsement of non-stereotypic negative traits. Further, participants who evaluated an African-American applicant after evaluating an Asian-American applicant endorsed significantly more negative stereotypes about African Americans than participants who evaluated the African-American applicant before evaluating the Asian-American applicant. Future research may investigate this negativity bias for African Americans wherein African Americans are negatively impacted by the comparison with an Asian American, although Asian Americans are neither positively nor negatively impacted. Such research may have important implications for understanding the effects of person impressions in real-life contexts.

PAGE 156

APPENDIX A INSTRUCTIONS FOR ANGRY AFFECT INDUCTION “A Picture is Worth a Thousand Words” In the first experiment today, we are interested in the way in which different people process narrative information. The particular condition to which you have been assigned deals with how people generate stories, such as the ones they imagine as fiction writers write or the ones they might generate to explain a particular situation. People in this condition will be writing about different things to make sure that we sample a wide variety of situations about which people are likely to write. To assist you in the task of writing, you will be given a photograph of the character in your story and of the type of situation about which you will write. Situation: This person is your neighbor, and on this particular day, this person has done something to make you very angry. Please take 5 minutes to write about an imagined event that could have led you to be angry or frustrated and to describe your feelings. In doing so, try to imagine how you would feel in that situation and to convey these feelings as you would if you were telling about this episode in a letter to a friend . Please note that the story should end at the time in which you feel angry and frustrated without going into further resolution of the event. Do not explain how the situation was resolved. 146

PAGE 157

APPENDIX B INSTRUCTIONS FOR HAPPY AFFECT INDUCTION “A Picture is Worth a Thousand Words” In the first experiment today, we are interested in the way in which different people process narrative information. The particular condition to which you have been assigned deals with how people generate stories, such as the ones they imagine as fiction writers write or the ones they might generate to explain a particular situation. People in this condition will be writing about different things to make sure that we sample a wide variety of situations about which people are likely to write. To assist you in the task of writing, you will be given a photograph of the character in your story and of the type of situation about which you will write. Situation: (This person is your neighbor, and on this particular day, this person has done something to make you very happy. Please take 5 minutes to write about an imagined event that could have led you to be happy or excited and to describe your feelings. In doing so, try to imagine how you would feel in that situation and to convey these feelings as you would if you were telling about this episode in a letter to a friend .) Please note that the story should end at the time in which you feel happy and excited without going into further resolution of the event. 147

PAGE 158

APPENDIX C PHOTOGRAPH OF ASIAN-AMERICAN MALE CHARACTER/NEIGHBOR 148

PAGE 159

APPENDIX D PHOTOGRAPH OF AFRICAN-AMERICAN MALE CHARACTER/NEIGHBOR 149

PAGE 160

150 APPENDIX E SOCIAL JUDGMENT TASK INSTRUCTIONS Instructions: You are an employee at Roommate Matchers, a company that offers roommate-matching services. Generally, when companies offer such se rvices, they utilize various means for determining whether a person would be a high-quality potential roommate, such as questionnaires, personality inventories, etc.. We are interested in assessing whether the questionnaires typically used by roommate-matching services lead to more satisfaction in roommate selection than other methods of roommate matching. The typical procedure is to have evaluators form a general impression of the applicant based on the information in the application. Because many impressions about a person are likely to be shared by many, the evaluator can often generate impressions of an applicant that are similar to the impressions a potential roommate might form of this applicant. For instance, the evaluator prediction that the applicant might be “friendly” may match a similar judgment by a person looking for a roo mmate. As part of this st udy, we will compare your assessment of an individual based on reading the application with the real ratings of the roommate with whom the applicant is matched. In this study, you will read the application of someone who was previously matched using typical roommate matching techniques. As you read it, feel free to jot down some notes as you would if you were to share information about the applicant with a coworker. Once you read the file, we will give you a questionnaire to measure your reactions to this informati on and to find out your im pression of the candidate based on standard personality inventories. There is another issue to mention about your task today. Roommate-matching services using such procedures can require a great deal of time and resources on the part of their empl oyees. However, in a real office environment, employees are often required to multi-task. An employee may have to answer the phone, type memos, evaluate applications, and answer questions from their colleagues or boss, all at the same time. To simulate this environment and thus obta in accurate assessments of the selection procedures, we are going to ask you to remember a number while you are completing the other tasks for the experiment. Please recite this number repeatedly while you are evaluating applicants. It is important that you accurately recall this number, as it is an aspect of your job that you will be evaluated on. At the end of the questionnaire, you will be asked to report this number. You will find your number in the front of your folder. Plea se begin reciting this number now. Here is the folder with a potential roommate’s applica tion. When you are finished reading the application, please let me know and I will give you a questionnaire about your impressions about the applicant you evaluated.

PAGE 161

APPENDIX F ASIAN-AMERICAN MALE’S APPLICATION

PAGE 162

OFFICE USE ONLY Rep: DATE/TIME: Amt. Paid: ___CASH ___Check #________ Client#: 55448 WELCOME TO ROOMMATE FINDERS Name: Hae Ryong Kim Date: August 6, 2001 . Address: 1802 SW 69 th Ave City: Gainesville State: FL Zip: 32611 . Phone: (352) 331-7889 Call Hrs: any Other Phone: none Call Hrs: N/A . mail: hrkim@ufl.edu Fax: none . Desired Move-in/out Date: immediately Occupation/Major: Undeclared . How did you find out about Roommate Finders? Flyer Housing Office Apartment Community Alligator Yellow Pages Website Friend Gainesville Sun Chamber of Commerce Drive By Alligator Other _____________ __________________________________________________________________________________________ SITUATION DESIRED Roommate for my residence Rent Per Person $450 . Roommate with a residence OR Roommate to look together to get a place Rent Range N/A . Which Section of Town? No preference NW SW SE NE Within 1 Mile of UF __________________________________________________________________________________________ DESCRIPTION OF MYSELF Gender: Male Female Age: 20 Marital Status: Single . Ethnicity: European Am (White) African Am Asian Am Latino Am (Hispanic) Other Student Type: Non-student Freshman Sophomore Junior Senior Grad student Smoking: Yes No Pets: Yes No Own Car: Yes No Student Organizations: Gatorwell, Asian Student Union __________________________________________________________________________________________ ACCEPTABLE ROOMMATE Gender: Male Female Either Age: over 18 Marital Status: Single . Ethnicity: European Am (White) African Am Asian Am Latino Am (Hispanic) No preference Student Type: Non-student Freshman Sophomore Junior Senior Grad student No preference Smoking: Yes No No pref Pets: Yes No No pref Own Car: Yes No No pref Interests/ Hobbies: No preference 152 __________________________________________________________________________________________

PAGE 163

APPENDIX G AFRICAN-AMERICAN MALE’S APPLICATION

PAGE 164

154 154 OFFICE USE ONLY Rep: DATE/TIME: Amt. Paid: ___CASH ___Check #________ Client#: 55448 WELCOME TO ROOMMATE FINDERS Name: Tyrone Williams Date: August 6, 2001 . Address: 1802 SW 69 th Ave City: Gainesville State: FL Zip: 32611 . Phone: (352) 331-7889 Call Hrs: any Other Phone: none Call Hrs: N/A . mail: twilliams@ufl.edu Fax: none . Desired Move-in/out Date: immediately Occupation/Major: Undeclared . How did you find out about Roommate Finders? Flyer Housing Office Apartment Community Alligator Yellow Pages Website Friend Gainesville Sun Chamber of Commerce Drive By Alligator Other _____________ __________________________________________________________________________________________ SITUATION DESIRED Roommate for my residence Rent Per Person $350 . Roommate with a residence OR Roommate to look together to get a place Rent Range N/A . Which Section of Town? No preference NW SW SE NE Within 1 Mile of UF __________________________________________________________________________________________ DESCRIPTION OF MYSELF Gender: Male Female Age: 20 Marital Status: Single . Ethnicity: European Am (White) African Am Asian Am Latino Am (Hispanic) Other Student Type: Non-student Freshman Sophomore Junior Senior Grad student Smoking: Yes No Pets: Yes No Own Car: Yes No Student Organizations: Gatorwell, Black Student Union __________________________________________________________________________________________ ACCEPTABLE ROOMMATE Gender: Male Female Either Age: over 18 Marital Status: Single . Ethnicity: European Am (White) African Am Asian Am Latino Am (Hispanic) No preference Student Type: Non-student Freshman Sophomore Junior Senior Grad student No preference Smoking: Yes No No pref Pets: Yes No No pref Own Car: Yes No No pref Interests/ Hobbies: No preference __________________________________________________________________________________________

PAGE 165

APPENDIX H MEMO FOR NOTES ABOUT EACH APPLICANT Roommate Finders MEMO Assume that you must describe the applicant to a co-worker. Make some brief notes about the applicant to share with your co-worker at Roommate Finders. ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 155

PAGE 166

APPENDIX I LOW DISTRACTION MANIPULATION (APPLICANT 1) Please recite this number repeatedly while you are evaluating the application. It is important that you accurately recall this number, as it is an aspect of your job that you will be evaluated on. At the end of the questionnaire, you will be asked to report this number. 32 After you look at this number, please return it to the experimenter and begin reciting it. 156

PAGE 167

APPENDIX J MODERATE DISTRACTION MANIPULATION (APPLICANT 1) Please recite this number repeatedly while you are evaluating the application. It is important that you accurately recall this number, as it is an aspect of your job that you will be evaluated on. At the end of the questionnaire, you will be asked to report this number. 3275 After you look at this number, please return it to the experimenter and begin reciting it. 157

PAGE 168

APPENDIX K HIGH DISTRACTION MANIPULATION (APPLICANT 1) Please recite this number repeatedly while you are evaluating the application. It is important that you accurately recall this number, as it is an aspect of your job that you will be evaluated on. At the end of the questionnaire, you will be asked to report this number. 327549 After you look at this number, please return it to the experimenter and begin reciting it. 158

PAGE 169

APPENDIX L LOW DISTRACTION MANIPULATION (APPLICANT 2) Please recite this number repeatedly while you are evaluating the application. It is important that you accurately recall this number, as it is an aspect of your job that you will be evaluated on. At the end of the questionnaire, you will be asked to report this number. 57 After you look at this number, please return it to the experimenter and begin reciting it. 159

PAGE 170

APPENDIX M MODERATE DISTRACTION MANIPULATION (APPLICANT 2) Please recite this number repeatedly while you are evaluating the application. It is important that you accurately recall this number, as it is an aspect of your job that you will be evaluated on. At the end of the questionnaire, you will be asked to report this number. 5742 After you look at this number, please return it to the experimenter and begin reciting it. 160

PAGE 171

APPENDIX N HIGH DISTRACTION MANIPULATION (APPLICANT 2) Please recite this number repeatedly while you are evaluating the application. It is important that you accurately recall this number, as it is an aspect of your job that you will be evaluated on. At the end of the questionnaire, you will be asked to report this number. 574218 After you look at this number, please return it to the experimenter and begin reciting it 161

PAGE 172

APPENDIX O PROBE FOR SUSPICION Now we’d like you to guess what the experiment was testing. What do you think the main hypothesis was? Please list any ideas or guesses in the space provided below. ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 162

PAGE 173

APPENDIX P RATINGS OF THE WRITING TASK Often times the time of processing and production of written information depends on how one feels with respect to the task at the time one is involved in the task. Therefore, we would like to ask you a few questions about the writing task. These questions are important to understand the story that you wrote and to generate an analysis of the writing properties in your story. Please answer the following questions about your picture and story. 1. Writing the story was: 0 1 2 3 4 5 6 7 8 9 Not at all Difficult Extremely Difficult --------------------------------------------------------------------------------------------------------------------2. Finding adequate words was 0 1 2 3 4 5 6 7 8 9 Not at all Difficult Extremely Difficult --------------------------------------------------------------------------------------------------------------------3. Structuring the sentences of my story was 0 1 2 3 4 5 6 7 8 9 Not at all Difficult Extremely Difficult --------------------------------------------------------------------------------------------------------------------4. Writing the story was: 0 1 2 3 4 5 6 7 8 9 Not at all Interesting Extremely Interesting --------------------------------------------------------------------------------------------------------------------5. Writing the story was: 0 1 2 3 4 5 6 7 8 9 Not at All Fun Extremely Fun --------------------------------------------------------------------------------------------------------------------6. Writing the story made me feel: 0 1 2 3 4 5 6 7 8 9 Not at All Happy Extremely Happy --------------------------------------------------------------------------------------------------------------------7. 0 1 2 3 4 5 6 7 8 9 Not at All Angry Extremely Angry --------------------------------------------------------------------------------------------------------------------8. 0 1 2 3 4 5 6 7 8 9 Not at All Sad Extremely Sad You have completed the first experiment. Please indicate to the experimenter that you are ready to begin Experiment 2. #______________ 163

PAGE 174

APPENDIX Q STEREOTYPE ACTIVATION MEASURE You are an employee at Roommate Matchers, a company that offers roommate matching services. As part of your responsibilities you will be required to complete a number of verbal tasks. To test your verbal aptitude, we would like you to complete the following word completion task. Some of the words may have more than one possible completion. In this case, use the first completion that comes to mind. Please complete the following words: (Asian-American Words) S_Y RI_E POLI_E M_TH S_ART (African-American words) _LACK LA_Y _UNNY _OOR LO_D (Filler Words) TAL_ BE_T _EACH S_OP DR_P S_IRT P_ONE _OX STOR_ PLA_ _OSE PA_T T_P 164

PAGE 175

APPENDIX R LIKELIHOOD JUDGMENTS OF APPLICANTS (ALL TRAITS) The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Hardworking Extremely Hardworking -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Unclean Extremely Unclean -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Intelligent Extremely Intelligent -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Arrogant Extremely Arrogant -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Aggressive Extremely Aggressive -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Modest Extremely Modest -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Promiscuous Extremely Promiscuous -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Fun-Loving Extremely Fun-Loving -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Cold Extremely Cold -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Athletic Extremely Athletic -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Ambitious Extremely Ambitious -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Uncommunicative Extremely Uncommunicative 165

PAGE 176

166 -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Rigid Extremely Rigid -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Sincere Extremely Sincere -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Materialistic Extremely Materialistic -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Even-Tempered Extremely Even-Tempered -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Loud Extremely Loud -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Responsible Extremely Responsible -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Messy Extremely Messy -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Musical Extremely Musical -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Funny Extremely Funny -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Flexible Extremely Flexible -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Family-Oriented Extremely Family-Oriented -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Introverted Extremely Introverted -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Fashionable Extremely Fashionable --------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Spiritual Extremely Spiritual -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Submissive Extremely Submissive -------------------------------------------------------------------------------------------------------------------------------------------

PAGE 177

167 The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Short-tempered Extremely Short-tempered -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Outgoing Extremely Outgoing -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Mean Extremely Mean -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Extroverted Extremely Extroverted -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Neat Extremely Neat -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Incompetent Extremely Incompetent -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Feminine Extremely Feminine -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Flashy Extremely Flashy -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Conservative Extremely Conservative -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Lazy Extremely Lazy -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Quiet Extremely Quiet -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Moral Extremely Moral -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Competent Extremely Competent -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Unintelligent Extremely Unintelligent -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Immoral Extremely Immoral -------------------------------------------------------------------------------------------------------------------------------------------

PAGE 178

168 The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Irresponsible Extremely Irresponsible -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Helpful Extremely Helpful -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Masculine Extremely Masculine -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Religious Extremely Religious -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Judgmental Extremely Judgmental -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Stingy Extremely Stingy -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Easy-going Extremely Easy-going -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Poor Extremely Poor -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Generous Extremely Generous --------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Violent Extremely Violent -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Elitist Extremely Elitist -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Shy Extremely Shy -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Unaffectionate Extremely Unaffectionate -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Trustworthy Extremely Trustworthy -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Calculating Extremely Calculating -------------------------------------------------------------------------------------------------------------------------------------------

PAGE 179

169 The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Wasteful Extremely Wasteful -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Considerate Extremely Considerate ------------------------------------------------------------------------------------------------------------------------------------------

PAGE 180

APPENDIX S STEREOTYPE APPLICATION TRAITS Positive Stereotypic Traits (African-American) Fun-loving Athletic Musical Funny Fashionable Negative Stereotypic Traits (African-American) Aggressive Unintelligent Loud Lazy Unclean Positive Stereotypic Traits (Asian-American) Hardworking Intelligent Modest Responsible Ambitious Negative Stereotypic Traits (Asian-American) Cold Arrogant Rigid Stingy Uncommunicative 170

PAGE 181

APPENDIX T POSITIVE AND NEGATIVE NON-STEREOTYPIC TRAITS (PRIMING EFFECTS) Positive Traits Negative Traits Family-Oriented Unclean Fun-loving Unintelligent Sincere Lazy Outgoing Arrogant Fashionable Cold Generous Incompetent Helpful Poor Trustworthy Mean Considerate Irresponsible Easy-going Immoral Flexible Messy Moral Promiscuous 171

PAGE 182

APPENDIX U ATTITUDES TOWARD APPLICANTS The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Likable Extremely Likable -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 A Bad Choice A Good Choice For a Roommate For a Roommate -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 Not at All Pleasant Extremely Pleasant -------------------------------------------------------------------------------------------------------------------------------------------The applicant is likely to be: 1 2 3 4 5 6 7 8 9 10 A Bad Match A Good Match For a Roommate For a Roommate -------------------------------------------------------------------------------------------------------------------------------------------172

PAGE 183

APPENDIX V INSTRUCTIONS FOR BOTH STUDIES Introduction Today, you will be participating in two, unrelated experiments. Please do not begin each experiment until I tell you to do so. The materials for the first study can be found in the folder to your right. The second experiment is on the computer. 173

PAGE 184

APPENDIX W FORM TO RECORD STORY/AFFECT INDUCTION STORY #__________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 174

PAGE 185

APPENDIX X SHORT INSTRUCTIONS FOR SOCIAL JUDGMENT TASK Reminder of Instructions: You are an employee at Roommate Matchers, a company that offers roommate matching services. Generally, when companies offer such services, they utilize various means for determining whether a person would be a high-quality potential roommate, such as questionnaires, personality inventories, etc. The following questions are the ones used by the particular company that provided the materials and information. Please fill out these questions using the scale provided after each statement. 175

PAGE 186

APPENDIX Y MODERN RACISM SCALE Over the past few years, the government and news media have shown more respect to blacks than they deserve. 0 1 2 3 4 Disagree Disagree Neither Agree Agree Agree Strongly Somewhat Nor Disagree Somewhat Strongly It is easy to understand the anger of black people in America. (reverse coded) 0 1 2 3 4 Disagree Disagree Neither Agree Agree Agree Strongly Somewhat Nor Disagree Somewhat Strongly Discrimination against blacks is no longer a problem in the United States. 0 1 2 3 4 Disagree Disagree Neither Agree Agree Agree Strongly Somewhat Nor Disagree Somewhat Strongly Over the past few years, blacks have gotten more economically than they deserve. 0 1 2 3 4 Disagree Disagree Neither Agree Agree Agree Strongly Somewhat Nor Disagree Somewhat Strongly Blacks are getting too demanding in their push for equal rights. 0 1 2 3 4 Disagree Disagree Neither Agree Agree Agree Strongly Somewhat Nor Disagree Somewhat Strongly Blacks should not push themselves where they are not wanted. 0 1 2 3 4 Disagree Disagree Neither Agree Agree Agree Strongly Somewhat Nor Disagree Somewhat Strongly 176

PAGE 187

REFERENCES Abele, A. (1992). Positive and negative mood influences on creativity: Evidence for asymmetrical effects. Polish Psychological Bulletin, 23, 203-221. Abele, A., Gendolla, G.H.E., & Petzold, P. (1998). Positive mood and in-groupoutgroup differentiation in a minimal group settling. Personality and Social Psychology Bulletin, 24, 1343-1357. Adelman, P.K., & Zajonc, R.B. (1989). Facial efference and the experience of emotion. Annual Review of Psychology, 40, 249-280. Albarracn, D., & Kumkale, G.T. (2003). Affect as information in persuasion: A model of affect identification and discounting. Journal of Personality and Social Psychology , 84, 453-469. Allport, G.W. (1954). The nature of prejudice. Cambridge, MA: Addison-Wesley. Asuncion, A.G., & Lam, W. F. (1995). Affect and impression formation: Influence of mood on person memory. Journal of Experimental Social Psychology, 31, 437-464. Bettelheim, B., & Janowitz, M. (1950). Dynamics of prejudice. Oxford, England: Harper. Blaney, P.H. (1986). Affect and memory: A review. Psychological Bulletin, 99, 229-246. Bless, H., Bohner, G., Schwarz, N., & Strack, N. (1990). Mood and persuasion: A cognitive response analysis. Personality and Social Psychology Bulletin, 16, 331-345. Bless, H., Clore, G.L., Schwarz, N., Golisano, V., Rabe, C., & Wlk, M. (1996). Mood and use of scripts: Does a happy mood really lead to mindlessness? Journal of Personality and Social Psychology, 71, 665-679. Bless, H., Schwarz, N., & Kemmelmeier, M. (1996). Mood and stereotyping: Affective states and the use of general knowledge structures. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 7, pp. 63-93). Chichester, UK: Wiley. 177

PAGE 188

178 Bless, H., Schwarz, N., & Wieland, R. (1996). Mood and the impact of category membership and individuating information. European Journal of Social Psychology, 26, 935-959. Blessum, K.A., Lord, C.G, & & Sia, T.L. (1998). Cognitive load and positive mood reduce typicality effects in attitude-behavior consistency. Personality & Social Psychology Bulletin, 24, 496-504. Bodenhausen, G.V. (1993). Emotions, arousal, and stereotype judgments: A heuristic model of affect and stereotyping. In D.M. Mackie & D.L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 13-37). San Diego, CA: Academic Press, Inc. Bodenhausen, G.V. & Kramer, G.P. (1990). Affective states trigger stereotypic judgments. Paper presented at the annual convention of the American Psychological Society, Dallas. Bodenhausen, G.V., Kramer, G.P. & Susser, K. (1994). Happiness and stereotypic thinking in social judgment. Journal of Personality and Social Psychology, 66, 621-632. Bodenhausen, G.V., Macrae, C.N. & Milne, A.B. (1998). Disregarding social stereotypes: Implications for memory, judgment, and behavior. In J.M. Golding and C.M. Macleod (Eds.) Intentional forgetting: Interdisciplinary approaches (pp. 349-368). Mahwah, NJ: Lawrence Erlbaum Associates, Inc., Publishers. Bodenhausen, G.V., Mussweiler, T., Gabriel, S., & Moreno, K.N. (2001). Affective influences on stereotyping and intergroup relations. In J.P. Forgas (Ed.), Handbook of Affect and Social Cognition (pp. 319-343). Mahwah, NJ: Lawrence Erlbaum Associates, Inc., Publishers. Bodenhausen, G.V., Sheppard, L.A., & Kramer, G.P. (1994). Negative affect and social judgment: The differential impact of anger and sadness. European Journal of Social Psychology, 24, 45-62. Bower, G.H. (1981). Mood and memory. American Psychologist, 36, 129-148. Bower, G.H. (1991). Mood congruity of social judgments. In J.P. Forgas (Ed.), Emotion and social judgments (pp.31-54). Oxford: Pergamon Press. Bower, G.H., Monteiro, K.P., & Gilligan, S.G. (1978). Emotional mood as a context for learning and recall. Journal of Verbal Learning and Verbal Behavior, 17, 573-585.

PAGE 189

179 Burgess, M.L. (1993). The effects of mood on stereotyping processes and intergroup behavior: A tendency to redefine one’s outgroup. Unpublished doctoral dissertation. University of Iowa. Clark, M.S., & Isen, A.M. (1982). Towards understanding the relationship between feeling states and social behavior. In A.H. Hastorf & A.M. Isen (Eds.), Cognitive social psychology (pp. 73-108). New York: Elsevier-North Holland. Clark, M.S., Milberg, S., & Erber, R. (1984). Effects of arousal on judgments of others’ emotions. Journal of Personality and Social Psychology, 46, 551-560. Clore, G.L., Schwarz, N., & Conway, M. (1994). Affective causes and consequences of social information processing. In R.S. Wyer & T.K. Srull (Eds.), Handbook of social cognition (2 nd ed.; pp. 323-419). Hillsdale, NJ: Erlbaum. Devine, P.G. (1989). Stereotypes and prejudice: Their automatic and controlled components. Journal of Personality and Social Psychology, 56, 5-18. Dollard, J., Miller, N.E. , Doob, L.W. , Mowrer, O. H. , Sears, R.R. (1939). Definitions. In J. Dollard, N.E. Miller, & L.W. Doob (Eds.), Frustration and aggression (pp. 1-26). New Haven, Conn: Yale University Press. Dovidio, J.F., Gaertner, S.L., Isen, A.M., & Lowrance, R. (1995). Group representations and intergroup bias: Positive affect, similarity, and group size. Personality & Social Psychology Bulletin, 21, 856-865. Dovidio, J.F., Gaertner, S.L., & Loux, S. (2000). Subjective experiences and intergroup relations: The role of positive affect. In H. Bless & J.P. Forgas (Eds.), The message within: The role of subjective experience in social cognition and behavior (pp.340-371). Philadelphia, PA: Psychology Press. Drolet, A., & Luce, M.F. (2004). The rationalizing effects of cognitive load on emotionbased trade-off avoidance. Journal of Consumer Research, 31, 63-77. Duclos, S.E., Laird, J.D., Schneider, E., Sexter, M., Stern, L., & Van Lighten, O. (1989). Emotion-specific effects of facial expressions and postures on emotional experience. Journal of Personality & Social Psychology, 57, 100-108. Dunn, E.W., & Spellman, B.A. (2003). Forgetting by remembering: Stereotype inhibition through rehearsal of alternative aspects of identity. Journal of Experimental Social Psychology, 39, 420-433. Eagly, A.H., & Chaiken, S. (1993). The psychology of attitudes. Orlando, FL: Harcourt Brace Jovanovich College Publishers.

PAGE 190

180 Ekman, P., Levenson, R.W., & Friesen, W.V. (1983). Autonomic nervous system activity distinguishes among emotions. Science, 221, 1208-1210. Esses,V.M., & Zanna, M.P. (1995). Mood and the expression of ethnic stereotypes. Journal of Personality & Social Psychology, 69, 1052-1068. Fiedler, K. (1991). On the task, the measures, and the mood in research on affect and social cognition. In J.P. Forgas (Ed.), Affect, cognition, and social behavior (pp. 100-119). Toronto, Canada: Hogrefe. Forgas, J.P. (1989). Mood effects on decision-making strategies. Australian Journal of Psychology, 41, 197-214. Forgas, J.P. (1991a). Affect and social judgments: An introductory review. In J.P. Forgas (Ed.), Emotion and social judgments (pp. 3-30). Oxford: Pergamon Press. Forgas, J.P. (1991b). Affective influences on partner choice: The role of mood in social decisions. Journal of Personality and Social Psychology, 61, 208-220. Forgas, J.P. (1991c). Mood effects on partner choice: Role of affect in social decisions. Journal of Personality and Social Psychology, 61, 708-720. Forgas, J.P. (1992a). Affect in social judgments and decisions: A multi-process model. In M. Zanna (Ed.), Advances in experimental social psychology (pp. 227-275). New York: Academic Press. Forgas, J.P. (1992b). Mood and the perception of unusual people: Affective asymmetry in memory and social judgments. European Journal of Social Psychology, 22, 531-547. Forgas, J.P. (1992c). On mood and peculiar people: Affect and person typicality in impression formation. Journal of Personality and Social Psychology, 62, 863-875. Forgas, J.P. (1995). Mood and judgment: The affect infusion model (AIM). Psychological Bulletin, 116, 39-66. Forgas, J.P., & Bower, G.H. (1988). Affect in social and personal judgments. In K. Fiedler & J. Forgas (Eds.), Affect, cognition and social behavior (pp. 183-208). Toronto, Ontario, Canada: Hogrefe. Forgas, J.P., & Fiedler, K. (1996). Us and them: Mood effects on intergroup discrimination. Journal of Personality and Social Psychology, 70, 28-40. Forgas, J.P. & Moylan, S.J. (1991). Affective influences on stereotype judgments. Cognition and Emotion, 5, 379-395.

PAGE 191

181 Gilbert, D.T., & Hixon, J.G. (1991). The trouble of thinking: Activation and application of stereotypic beliefs. Journal of Personality and Social Psychology, 60, 509-517. Gilbert, D.T., & Osborne, R.E. (1989). Thinking backward: Some curable and incurable consequences of cognitive busyness. Journal of Personality and Social Psychology, 57, 940-949. Gold, R.S. (2002). The effects of mood states on the AIDS-related judgments of gay men. International Journal of STD & AIDS, 13, 475-481. Greenwald, A.G., Nosek, B.A., & Banaji, M.R. (2003). Understanding and using the Implicit Association Test: I. An improved scoring algorithm. Journal of Personality and Social Psychology, 85, 197-216. Haddock, G., Zanna, M.P., & Esses, V.M. (1994). Mood and the expression of intergroup attitudes: The moderating role of affect intensity. European Journal of Social Psychology, 24, 189-205. Hovland, C.I., & Sears, R.R. (1940). Minor studies of aggression: VI. Correlation of lynchings with economic indices. Journal of Psychology: Interdisciplinary & Applied , 9, 301-310. Isbell, L., & Wyer, R.S. (1999). Correcting for mood-induced bias in the evaluation of political candidates: The roles of intrinsic and extrinsic motivation. Personality & Social Psychology Bulletin , 25, 237-249. Isen, A.M. (1984). Towards understanding the role of affect in cognition. In R.S. Wyer & T.K. Srull (Eds.), Handbook of social cognition (Vol. 3, pp. 179-236). Hillsdale, NJ: Erlbaum. Isen, A.M. (1987). Positive affect, cognitive processes and social behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol 20, pp. 203-253). Orlando, FL: Academic Press. Isen, A.M., & Daubman, K.A. (1984). The influence of affect on categorization. Journal of Personality & Social Psychology, 47, 1206-1217. Isen, A.M. & Levin, P.F. (1972). The effect of feeling good on helping: Cookies and kindness. Journal of Personality and Social Psychology, 21, 384-388. Isen, A.M., Niedenthal, P.M., & Cantor, N. (1992). An influence of positive affect on social categorization. Motivation and Emotion, 16, 65-78.

PAGE 192

182 Isen, A.M., Shalker, T., Clark, M., & Karp, L. (1978). Affect, accessibility of material in memory and behavior: A cognitive loop? Journal of Personality and Social Psychology, 36, 1-12. Izard, C.E. (1972). Patterns of emotions: A new analysis of anxiety and depression. New York: Academic Press. Jackson, L. A., Lewandowski, D.A., Fleury, R.E., & Chin, P.P. (2001). Effects of affect, stereotype consistency, and valence of behavior on causal attributions. The Journal of Social Psychology, 141, 31-48. Jackson, L.A., & Sullivan, L.A. (2001). Cognition and affect in evaluations of stereotyped group members. The Journal of Social Psychology, 129, 659-672. Jepson, C. & Chaiken, S. (1990). Chronic issue-specific fear inhibits systematic processing of persuasive communications. Journal of Social Behavior and Personality, 5, 61-84. Kim, H.S., & Baron, R.S. (1988). Exercise and illusory correlation: Does arousal heighten stereotypic processing? Journal of Experimental Social Psychology, 24, 366-380. Krauth-Gruber, S., & Ric, F. (2000). Affect and stereotypic thinking: A test of the mood-and-general-knowledge model. Personality and Social Psychology Bulletin, 26, 1587-1597. Lambert, A.J., Khan, S.R., Lickel, B.A., & Fricke, K. (1997). Mood and the correction of positive versus negative stereotypes. Journal of Personality and Social Psychology, 72, 1002-1016. Mackie, D.M., Hamilton, D.L., Schroth, H.A., Carlisle, C.J., Gersho, B.F., Meneses, L.M., Nedler, B.F., & Reichel, L.D. (1989). The effects of induced mood on expectancy-based illusory correlations. Journal of Experimental Social Psychology, 25, 524-544. Mackie, D.M. & Worth, L.T. (1989). Processing deficits and the mediation of positive affect in persuasion. Journal of Personality and Social Psychology, 57, 27-40. Mayer, J.D. , Gaschke, Y.N. , Braverman, D.L., & Evans, T.W. (1992). Mood-congruent judgment is a general effect. Journal of Personality & Social Psychology , 63, 119-132. McConahay, J.B. (1986). Modern racism, ambivalence, and the Modern Racism Scale. In J.F. Dovidio & S.L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 91-126). New York: Academic Press.

PAGE 193

183 Meiland, P.A. (1996). The effects of mood, gender, and need for cognition on distinctiveness-based stereotyping. Unpublished doctoral dissertation. Columbia University. Moreno, K.N., & Bodenhausen, G.V. (2001). Intergroup affect and social judgment: Feelings as inadmissible information. Group Processes and Intergroup Relations, 4, 21-29. Moskowitz, G.B., Salomon, A.R., & Taylor, C.M. (2000). Preconsciously controlling stereotyping : Implicitly activated egalitarian goals prevent the activation of stereotypes. Social Cognition, 18, 151-177. Murray, N., Sujan, H., Hirt, E.R., & Sujan, M. (1990). The influence of mood on categorization: A cognitive flexibility interpretation. Journal of Personality & Social Psychology, 59, 411-425. Nosek, B.A., Greenwald, A.G., & Banaji, M.R. (2003). Understanding and using the Implicit Association Test: II. Method variables and construct validity. Personality and Social Psychology Bulletin, 31, 166-180. Ottati, V.C., & Isbell, L.M. (1996). Effects on mood during exposure to target information on subsequently reported judgments: An on-line model of misattribution and correction. Journal of Personality & Social Psychology , 71, 39-53. Park, J. & Banaji, M.R. (2000). Mood and heuristics: The influence of happy and sad states on sensitivity and bias in stereotyping. Journal of Personality & Social Psychology, 78, 1005-1023. Pendry, L.F., & Macrae, C.N. (1999). Cognitive load and person memory: The role of perceived group variability. European Journal of Social Psychology, 29, 925-942. Petty, R.E., Schumann, D.W., Richman, S.A., & Strathman, A.J. (1993). Positive mood and persuasion: Different roles for affect under highand low-elaboration conditions. Journal of Personality & Social Psychology , 64, 5-20. Polivy, J., & Doyle, C. (1980). Laboratory induction of mood states through the reading of self-referent mood statements: Affective changes or demand characteristics? Journal of Abnormal Psychology , 89, 286-290. Queller, S., Mackie, D.M., & Stroessner, S.J. (1996). Ameliorating some negative effects of positive mood: Encouraging happy people to perceive intragroup variability. Journal of Experimental Social Psychology, 32, 361-386.

PAGE 194

184 Revelle, W. & Loftus, D.A. (1992). The implications of arousal effects for the study of affect and memory. In S.A. Christianson (Ed.), Handbook of emotion and memory (pp. 113-149). Hillsdale, NJ: Erlbaum. Rodriguez Torres, R. (1998). The effect of affective induction on racial prejudice. Revista de Psicologia Social, 13, 279-289. Schiff, B.B., Esses, V.M., & Lamon, M. (1992). Unilateral facial contractions produce mood effects on social cognitive judgments. Cognition and Emotion, 6, 357-368. Schwarz, N. (1990). Feelings as information: Informational and motivational functions of affective states. In E.T. Higgins & R.M. Sorrentino (Eds.), Handbook of motivation and cognition (Vol. 2, pp. 527-561). New York: Guilford Press. Schwarz, N. Bless, H., & Bohner, G. (1991). Mood and persuasion: Affective states influence the processing of persuasive communications. In M.P. Zanna (Ed.), Advances in experimental social psychology (Vol. 24, pp. 161-199). San Diego, CA: Academic Press. Schwarz, N., & Clore, G.L. (1983). Mood, misattribution, and judgments of well-being: Informative and directive functions of affective states. Journal of Personality and Social Psychology, 45, 513-523. Schwarz, N., & Clore, G.L. (1988) How do I feel about it? The informative function of affective states. In K. Fiedler & J.P. Forgas (Eds.), Affect, cognition, and social behavior (pp.44-62). Toronto, Canada: Hogrefe. Sheppard, L.A. (1996). The effects of incidental and integral affect upon the expression of prejudice. Unpublished doctoral dissertation. Michigan State University. Singer, J.A., & Salovey, P. (1988). Mood and memory: Evaluating the network theory of affect. Clinical Psychology Review, 8, 211-251. Shiv, B., & Huber, J. (2000). The impact of anticipating satisfaction on consumer choice. Journal of Consumer Research, 27, 202-216. Singer, J.A., & Salovey, P. (1988). Mood and memory: Evaluating the Network Theory of Affect. Clinical Psychology Review , 8, 211-251. Son Hing, L.S., Li, W., & Zanna, M.P. (2002). Inducing hypocrisy to reduce prejudicial responses among aversive racists. Journal of Experimental and Social Psychology, 38, 71-78. Stepper, S. & Strack, F. (1993). The influence of mood on stereotypical evaluations of East and West Germans. Zeitschrift fr Sozialpsychologie, 24, 218-225.

PAGE 195

185 Strack, F., Schwarz, N., & Gschneidinger, E. (1985). Happiness and reminiscing: The role of time perspective, mood, and mode of thinking. Journal of Personality & Social Psychology, 49, 1460-1469. Stroessner, S.J., Hamilton, D.L., & Mackie, D.M. (1992). Affect and stereotyping: The effect of induced mood on distinctiveness-based illusory correlations. Journal of Personality & Social Psychology, 62, 564-576. Stroessner, S.J., & Mackie, D.M. (1992). The impact of induced affect on the perception of variability in social groups. Personality and Social Psychology Bulletin, 18, 546-554. Stroessner, S.J., & Mackie, D.M. (1993). Affect and perceived group variability: Implications for stereotyping and prejudice. In D.M. Mackie & D.L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 63-86). San Diego, CA: Academic Press, Inc. Velten, E. (1968). A laboratory task for induction of mood states. Behaviour Research and Therapy, 6, 473-482. Ward, A., & Mann, T. (2000). Don’t mind if I do: Disinhibited eating under cognitive load. Journal of Personality & Social Psychology, 78, 753-763. Wegener, D.T., Petty, R.E., & Smith, S.M. (1995). Positive mood can increase or decrease message scrutiny: The hedonic contingency of mood and message processing. Journal of Personality and Social Psychology, 69, 5-15. Wong, P., Lai, C.F., Nagasawa, R., & Lin, T. (1998). Asian Americans as a model minority: Self perceptions by other racial groups. Sociological Perspectives, 41, 95-118. Worth, L.T. & Mackie, D.M. (1987). Cognitive mediation of positive affect in persuasion. Social cognition, 5, 76-94. Wyer, R.S., & Srull, T.K. (1989). Memory and cognition in its social context. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

PAGE 196

BIOGRAPHICAL SKETCH Penny S. McNatt entered the University of Florida as a freshman in the fall of 1988. She received a B.A. in English with a minor in secondary education. After graduating in 1992, she decided to pursue an M.Ed. in special education from the University of Florida and began teaching high school English in a varying exceptionalities classroom. After teaching for several years, Penny decided to pursue graduate studies in social psychology at the University of Florida. Penny was granted the degree of Doctor of Philosophy from the University of Florida in May, 2005. Her dissertation focused on the influence of relevant affect on stereotype activation and application. 186