<%BANNER%>

Breaking the Link Between Cue and Perception of Threat -- A Threat-Based Approach to Reducing Prejudice

Permanent Link: http://ufdc.ufl.edu/UFE0042199/00001

Material Information

Title: Breaking the Link Between Cue and Perception of Threat -- A Threat-Based Approach to Reducing Prejudice
Physical Description: 1 online resource (83 p.)
Language: english
Creator: Schember, Tatiana
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: A threat-based approach to prejudice was used to examine an intervention aimed at reducing prejudice towards people who are obese and who may appear diseased. From a threat-based perspective, a cue of threat leads to a perception of threat, which leads to emotions and cognitions associated with that threat, which leads to behavioral inclinations, which would remove or mitigate the threat. By weakening the link between cue and perception of threat, prejudice may be reduced, but the effectiveness of this strategy may depend on the type of threat cue (high diagnosticity threat cue versus low diagnosticity threat cue). Participants in this study interacted with a confederate who appeared to be either skin-diseased (high diagnosticity threat cue) or obese (low diagnosticity threat cue). In the intervention conditions, the link between cue and perception of threat was weakened by having the confederate wash his or her hands. Results showed that the effectiveness of the intervention at reducing threat perception, feelings of disgust, prejudice, and behavioral responses to threat perception and disgust (measures of hand-washing) depended on the type of threat cue. The results suggest that interventions which weaken the link between cue and perception of threat may be effective at reducing prejudice and deserve further investigation.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Tatiana Schember.
Thesis: Thesis (M.S.)--University of Florida, 2010.
Local: Adviser: Webster, Gregory Daniel.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042199:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042199/00001

Material Information

Title: Breaking the Link Between Cue and Perception of Threat -- A Threat-Based Approach to Reducing Prejudice
Physical Description: 1 online resource (83 p.)
Language: english
Creator: Schember, Tatiana
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: A threat-based approach to prejudice was used to examine an intervention aimed at reducing prejudice towards people who are obese and who may appear diseased. From a threat-based perspective, a cue of threat leads to a perception of threat, which leads to emotions and cognitions associated with that threat, which leads to behavioral inclinations, which would remove or mitigate the threat. By weakening the link between cue and perception of threat, prejudice may be reduced, but the effectiveness of this strategy may depend on the type of threat cue (high diagnosticity threat cue versus low diagnosticity threat cue). Participants in this study interacted with a confederate who appeared to be either skin-diseased (high diagnosticity threat cue) or obese (low diagnosticity threat cue). In the intervention conditions, the link between cue and perception of threat was weakened by having the confederate wash his or her hands. Results showed that the effectiveness of the intervention at reducing threat perception, feelings of disgust, prejudice, and behavioral responses to threat perception and disgust (measures of hand-washing) depended on the type of threat cue. The results suggest that interventions which weaken the link between cue and perception of threat may be effective at reducing prejudice and deserve further investigation.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Tatiana Schember.
Thesis: Thesis (M.S.)--University of Florida, 2010.
Local: Adviser: Webster, Gregory Daniel.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042199:00001


This item has the following downloads:


Full Text

PAGE 1

1 BREAKING THE LINK BETWEEN CUE AND PERCEPTION OF THREAT A THREAT BASED APPROACH TO REDUCING PREJUDICE By TATIANA OROZCO SCHEMBER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLME NT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Tatiana Orozco Schember

PAGE 3

3 To John, for all his love and encouragement

PAGE 4

4 ACKNOWLEDGMENTS First, I would like to thank Dr. Gregory Webster for i nviting me to the University of Florida to pursue my graduate studies. I could not ask for a better advisor. I would also like to thank the other members of my supervisory committee, Dr. Cathy Cottrell and Dr. Bonnie Moradi ; their help and advice has been immeasurable. I also owe a great debt of gratitude to all of the r esearch assistants who helped collect data for this project: Robert Boggs, Nick Ciancioso, Kim Dalve, Jessica Darby, Laura Howell, Grace Jeon, Kendall Johnson, Corey McGann, and Nick Ungson I simply could not have conducted this research without all of their hours of hard work. Special thanks go to my family and close friends, especially my mom, Robert, and Ellie for all their support and for believing in me Finally, I thank my husband, John, for all his help, love, understanding, and encouragement

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 10 Prejudice Reduction ................................ ................................ ................................ 11 Threat Based Prejudice and the Sociofunctional Approach ................................ .... 14 Prejudice Reduction Rev isited ................................ ................................ ................ 17 Current Research and Hypotheses ................................ ................................ ......... 18 Overview ................................ ................................ ................................ .......... 18 Predict ions ................................ ................................ ................................ ........ 20 2 METHOD ................................ ................................ ................................ ................ 24 Participants ................................ ................................ ................................ ............. 24 Design and Variables ................................ ................................ .............................. 24 Procedure ................................ ................................ ................................ ............... 27 Predictions ................................ ................................ ................................ .............. 31 3 RESULTS ................................ ................................ ................................ ............... 32 Preliminary Analyses ................................ ................................ .............................. 32 Self Report Measures ................................ ................................ ............................. 34 Behavioral Measures ................................ ................................ .............................. 40 4 DISCUSSION ................................ ................................ ................................ ......... 57 Summary of Findings ................................ ................................ .............................. 58 Implications ................................ ................................ ................................ ............. 62 Limitations ................................ ................................ ................................ ............... 65 APPENDIX A TASTE TESTING QUESTIONNAIRE ................................ ................................ ..... 68 B QUESTIONS OF SPECIFIC THREAT ................................ ................................ .... 70 C QUESTIONS OF SPECIFIC EMOTIONS IN RESPONSE TO THREATS .............. 71

PAGE 6

6 D FEELING THERMOMETER ................................ ................................ ................... 73 E SOCIAL DISTANCING SCALE ................................ ................................ ............... 74 F PERCEIVED VULNERABILITY TO DISE ASE SCALE ................................ ........... 75 G DEMOGRAPHICS ................................ ................................ ................................ .. 78 REFERENCES ................................ ................................ ................................ .............. 80 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 83

PAGE 7

7 LIST OF TABLES Table page 3 1 Correlations for Dependent Variables and Descriptive Statistics ........................ 44 3 1 Continued ................................ ................................ ................................ ........... 45 3 2 Contrast C oding for Multiple Regression Predictors ................................ ........... 46 3 3 Self Report Measures (without PVD as a Covariate) ................................ ......... 46 3 4 Natural Log Transforme d Perceived Threat ................................ ....................... 47 3 5 Perceived Threat ................................ ................................ ................................ 47 3 6 Disgust ................................ ................................ ................................ ............... 47 3 7 General Negative Affect ................................ ................................ ..................... 47 3 8 Social Distancing ................................ ................................ ................................ 47 3 9 Feeling Thermometer ................................ ................................ ......................... 48 3 10 Natural Log Transformed Seconds Scrubbing Hands ................................ ........ 48 3 11 Seconds Scrubbing Hands ................................ ................................ ................. 48 3 12 Pumps 2 ................................ ................................ ................................ ............. 48 3 13 Behavioral Measures (Time 1 covariate, mean centered) ................................ .. 49 3 14 Regression Perceived Threat (ln) ................................ ................................ .... 50 3 15 Regression Disgust ................................ ................................ ......................... 51 3 16 Regression General Negative Affect ................................ ............................... 52 3 17 Regression Fee ling Thermometer ................................ ................................ ... 53 3 18 Regression Social Distancing ................................ ................................ .......... 54 3 19 Regression Natural Log Transformed Seconds Scrubbing Hands .................. 55 3 20 Regression Pumps 2 ................................ ................................ ....................... 56

PAGE 8

8 LIST OF FIGURES Figure page 1 1 Threat Management Syste m ................................ ................................ .............. 23 1 2 The link between Cue and Perception of Threat ................................ ................. 23

PAGE 9

9 Abstract of Thesis Presented to the Graduate School of the University of Florida i n Partial Fulfillment of the Requirements for the Degree of Master of Science BREAKING THE LINK BETWEEN CUE AND PERCEPTION OF THREAT A THREAT BASED APPROACH TO REDUCING PREJUDICE By Tatiana Orozco Schember December 2010 Chair: Gregory D. Webster Major : Psychology A threat based approach to prejudice was used to examine an intervention aimed at reducing prejudice towards people who are obese and who may appear diseased From a threat based perspective, a cue of threat leads to a perception of threat, w hich leads to emotions and cognitions associated with that threat, which leads to behavioral inclinations, which would remove or mitigate the threat By weakening the link between cue and perception of threat, prejudice may be reduced, but the effectivenes s of this strategy may depend on the type of threat cue (high diagnosticity threat cue versus low diagnosticity threat cue) Participants in this study interacted with a confederate who appeared to be either skin diseased (high diagnosticity threat cue) or obese (low diagnosticity threat cue) In the intervention conditions, the link between cue and perception of threat was weakened by having the confederate wash his or her hands Results showed that the effectiveness of the intervention at reducing threat perception, feelings of disgust, prejudice, and behavioral responses to threat perception and disgust (measures of hand washing) depended on the type of threat cue The results suggest that interventions which weaken the link between cue and perception of threat may be effective at reducing prejudice and deserve further investigation

PAGE 10

10 CHAPTER 1 INTRODUCTION Imagine you are walking through a zoo You come up to a gorilla habitat and notice a huge gorilla who is particularly angry, beating his chest, showi ng his teeth, and making all sorts of noise You start to feel uncomfortable and so continue exploring the zoo Later on, you decide to sit down at a picnic table for lunch You are about to take a bite out of your sandwich when a leaf from an adjacent tre e lands on the table, right in front of your lunch You go to brush aside the leaf when you realize that there, sitting on the leaf, is a big, slimy slug Thankfully, you noticed he was there before touching it, but you still feel a little queasy about you r close encounter If I asked you to think about your reactions to these two creatures, how would they compare? It would probably be safe to say that your feelings towards both of these animals were negative and unpleasant, but besides sharing a negative valence, your reactions to these two creatures were quite different from one another The gorilla likely evoked fear, an adaptation that prepares one for flight in the face of potential physical harm (Neuberg & Cottrell, 2006); the slug, on the other hand likely evoked disgust, an adaptation that predisposes one to avoid transmitters of contagious disease, such as rotting meat or sick people (Faulkner, Schaller, Park, & Duncan, 2004; Neuberg & Cottrell, 2006) Classifying these reactions as the same is cl early inappropriate If it is inappropriate to classify responses solely on valence when considering interactions with zoo residents, with whom most of us have little contact, how much more inappropriate is it to classify responses solely on valence when considering our interactions with other people, with whom we have endless contact? Although it may be common to view prejudice as affect on a negative positive continuum, this view masks

PAGE 11

11 the complexity of the different emotional and behavioral reactions t o others (Cottrell & Neuberg, 2005) Furthermore, a valence approach to prejudice also seems to ignore prejudice's underlying process; that is, it does not take into account the various reactions that occur upon encountering a target of prejudice (this wil l be discussed further in the section Threat Based Prejudice and the Sociofunctional Approach ) By failing to recognize the complexity of emotional reactions elicited by various others, and failing to recognize the underlying processes of prejudice, resear chers may be hindering efforts to reduce prejudice I will attempt to provide evidence in support of a threat based approach to prejudice reduction in the proposed study First in this proposal is a brief and non exhaustive review of perhaps the most pop ular type of prejudice reduction strategy in the literature, the contact strategy Following this review will be a description of the threat based approach to prejudice with a particular focus on perceived health threats This approach will then be applied to discuss a prejudice reduction strategy An experiment will then be proposed which applies this strategy to the reduction of health threat prejudices, specifically, prejudice against obese and skin diseased individuals Finally, predictions and planned analyses for this experiment will be outlined. Prejudice Reduction Perhaps the most popular prejudice reduction theory, and seemingly the most intuitive, is that to reduce prejudice, one only needs to increase contact between people who hold prejudiced be liefs towards one another A pervasive belief among the general population is that people are prejudiced because they fear the unknown (Nelson, 2006), or because they simply "don't know any better" (Kenrick, Neuberg, & Cialdini, 2007) An obvious prejudice reduction strategy, then, would be to dispel this fear of the unknown

PAGE 12

12 and teach people about others by increasing contact between groups Unfortunately, simply putting different groups in contact with each other in itself is not enough to reduce prejudice (Kenrick, Neuberg, & Cialdini, 2007; Nelson, 2006) For example, a study of the effects of desegregation of public schools found no evidence of decreased prejudiced feelings towards African American students by Caucasian students following desegregation, and though some African American students showed a decrease in negative feelings towards Caucasians, others showed an increase (Stephan, 1978) Furthermore, contact can sometimes even increase negative feelings between group members In another study on th e effects of desegregation, Caucasian and African American students with segregated backgrounds showed more negative attitudes to both ingroup and outgroup members following desegregation (Stephan & Rosenfield, 1978) Increasing contact between groups is the basic strategy underlying the hypotheses developed by Williams (1947, as cited in Utsey, Ponterotto, & Porter, 2008) and Allport (1954, as cited in Nelson 2006) Importantly, Williams identified specific preconditions that must be met for contact to b e effective, as did Allport According to Williams (1947, as cited in Utsey, Ponterotto, & Porter, 2008), in order for contact between groups to be successful at reducing prejudice, the groups must be equal in status, they must share a common goal, and the ir activities must promote meaningful interpersonal interaction Building off of this work, Allport also asserted that groups must be in equal in status and share a common goal, but he added that they must be dependent on one another to achieve their goal and that there must be legitimate authority figures supporting and encouraging intergroup cooperation (Allport, 1954, as

PAGE 13

13 cited in Utsey, Ponterotto, & Porter, 2008; Nelson, 2006) Other researchers have continued to build off of the contact hypothesis, add ing or identifying other factors which influence the success of a contact based prejudice reduction strategy (e.g., Amir, 1969; Pettigrew, 1998; Stephan, 1985, as cited in Nelson, 2006) A meta analysis of 515 prejudice studies provided evidence in suppo rt of Allport's contact hypothesis; rigorous, tightly controlled tests of Allport's contact hypothesis showed larger effect sizes in prejudice reduction relative to less rigorous, convenience sample survey studies (Pettigrew & Tropp, 2006) Further, studie s meeting all four of Allport's proposed optimal contact preconditions tended to have larger effect sizes in prejudice reduction than did studies which met only some of Allport's preconditions (Pettigrew & Tropp, 2006). Although there seems to be some pro mise in reducing prejudice according to the approaches described above, I would like to discuss an alternative approach for several reasons First, although the theory behind previous work may not be confined to interracial conflict, it does seem to have a disproportionately large focus on interracial prejudice, and in particular, prejudice between African Americans and Caucasians in the U.S On the other hand, the threat based approach proposed by Cottrell and Neuberg (2005) is not specific to any particul ar group or location Rather, the threat based approach deals with qualitatively distinct types of threats, as well as with emotional reactions to those threats, which are applicable across time and location Second, the previous approaches to the study of prejudice have been shown to obscure the qualitatively distinct types of prejudices and emotion profiles elicited by others through a focus on valence (Cottrell & Neuberg, 2005) This has important implications

PAGE 14

14 for prejudice reduction, for if there are mu ltiple types of prejudices (Cottrell & Neuberg, 2005), there may also need to be multiple types of prejudice reduction strategies, tailored to the type of threat and emotion profiles specific to a given type of prejudice Third, the previous attempts at pr ejudice reduction in the literature do not seem to take into account the underlying process of prejudice; the proposed study, on the other hand, does take into account the process of prejudice laid out by Neuberg and Cottrell (2006) Hence, I feel it would be useful to take a threat based approach to reducing prejudice Threat B ased Prejudice and the Sociofunctional Approach Humans are highly social animals, and for good reason; high social interdependence has many survival benefits (Cottrell & Neuberg, 2 005; Neuberg & Cottrell, 2008; Park, Faulkner, & Schaller, 2003; Sundie, Cialdini, Griskevicius, & Kenrick, 2006) Through cooperation with others, we can more easily acquire resources such as food and water, find or construct shelter, and protect ourselve s from dangerous others; however, there are also costs to group living (Cottrell & Neuberg, 2005; Neuberg & Cottrell 2008), such as exposure to contagious disease and exploitation by others who would "free ride" on other group members' efforts, partaking o f the group's resources or protection, but failing to contribute to group resources or protection themselves. To manage the threats posed by group living, it would have been necessary to evolve a system that could detect and respond to threats in appropri ate ways (Faulkner et al., 2004; Kurzban & Leary 2001; Park, Faulkner, & Schaller, 2003) Many species have an evolved system to detect and respond to various threats to survival and reproduction (Neuberg & Cottrell, 2006; Neuberg & Cottrell, 2008) Differ ent threats evoke qualitatively different types of emotions, which then facilitate appropriate

PAGE 15

15 cognitive and behavioral responses that serve to mitigate or eliminate the threat (Neuberg & Cottrell, 2006; Tooby & Cosmides, 1990) A cue that connotes threat will lead to a perception of threat This threat perception leads to a threat response syndrome, which includes the emotions, attitudes, and behavioral inclinations associated with that threat The threat response syndrome can lead to a behavioral response -in terms of intergroup interactions and prejudice, the prejudiced behavior or discrimination (see Figure 1 1 ; Neuberg & Cottrell, 2006) For example, think back to the gorilla from the zoo Imagine the gorilla escaped from his cage and was now speeding towards you with teeth bared; the rapid approach and angry expression serve as cues to danger and lead to a perception of physical safety threat This perception of threat to physical safety leads to fear and a desire to run away Ultimately, the threat r esponse syndrome would lead to the behavior response of actually running away This threat management system did not evolve to manage social threats, as it exists in other nonsocial animals, but was likely co opted to help individuals detect and mitigate t hreats associated with group living (Cottrell & Neuberg, 2005; Neuberg & Cottrell, 2008). Health threats Health threats are one such class of threats associated with group living From an evolutionary perspective, one of the most basic goals of any organ ism is to reproduce, but it needs to survive long enough in order to successfully reproduce (Neuberg & Cottrell, 2006; Cosmides & Tooby, 1997) Good health is obviously essential to survival and reproduction Furthermore, good health is desired not only in the self, but in one's "ideal" interaction partners as well (Cottrell, Neuberg, & Li, 2007) Thus, potential health threats should be particularly salient Our ancestors who attended to and avoided such threats would have enhanced their survival and

PAGE 16

16 repro ductive success greatly relative to those who did not (Faulkner et al., 2004; Neuberg & Cottrell 2006) Indeed, one could argue that those who failed to attend to such threats died out and are not our ancestors Here is where the aforementioned threat mana gement system comes into play A health threat cue would lead to a perception of health threat This perception of health threat would lead to emotional reactions to the threat, namely disgust (Park, Faulkner, & Schaller, 2003) Disgust would lead to behav ioral responses that would reduce or remove the threat Behavioral responses to disgust typically include wrinkling the nose (to protect nose and eyes from pathogens), protruding of the tongue (as if to eject the threatening substance), and general avoidan ce behavior (Cosmides & Tooby, 1997; Park, Schaller, & Crandall, 2007). Among the cues to health threats that our ancestors would have learned to detect and which still lead to a perception of threat today are skin lesions and bodily fluid secretions, which are suggestive of contagious disease (Kurzban & Leary, 2001; Neuberg & Cottrell 2006; Park, Faulkner, & Schaller, 2003) Although obesity is not typically perceived as a disease that can be spread through human contact with pathogens, it may be perc eived as a health threat because of heuristic conceptions that morphological deviations -for example, obesity, or extreme thinness -may be indicative on a general level of disease or parasitic infection (Kurzban & Leary, 2001; Neuberg & Cottrell, 2006; Park, Faulkner, & Schaller, 2003; Park, Schaller, & Crandall, 2007) In an experiment by Park, Schaller, and Crandall (2007), participants implicitly associated obesity with notions of disease, and individuals who felt particularly vulnerable to disease h ad more negative attitudes towards obese people relative to

PAGE 17

17 participants who did not feel vulnerable (Park, Schaller, & Crandall, 2007) Furthermore, this effect was pronounced following exposure to photos of obese people (Park, Schaller, & Crandall, 2007) Prejudice Reduction Revisited If prejudice functions according to the threat management system, it can be thought of as a chain of responses (Neuberg & Cottrell, 2006; see Figure 1 1 ) Like with any chain, it is only necessary to break one link in order to break the chain In terms of prejudice, the different links on this chain are the links between the cue of threat and perception of threat (Link A), the link between perception of threat and threat response syndrome (Link B), and the link between the t hreat response syndrome and behavior response (Link C) Breaking any of these links should reduce prejudiced behavior, if not prejudiced attitudes and emotions Additionally, eliminating the cue to threat should prevent the threat management system from en gaging But is it more effective to break one of the links of the threat management system or prevent the system from starting? If it is more effective to break one of the links, which link is the weakest? Given that qualitatively different threats lead to qualitatively different emotion profiles and behavioral inclinations, the answers to these questions may depend on the threat in question For example, if the angry facial expression of an outgroup member serves as a cue to physical harm to an observer, it may be fairly simple to break the link between cue and perception of threat ( Figure 1 2 ) by showing that the outgroup member is angry at someone or something other than the observer, or by showing that the outgroup member does not mean the observer any harm despite their anger On the other hand, if a target other has many sores and lesions on their skin serving as cues to pathogen threat, it may be very difficult to break the link between

PAGE 18

18 cue and perception of threat because the cue is a reliable predi ctor of threat (i.e., it is diagnostic of threat) and the association between cue and perception would be very strong The link between cue of health threat and perception of health threat may be very strong indeed, due to a tendency for such disgust evo king cues and actual threat to covary Although it is known that the relationship between cues to health threats and actual health threats may be spurious (Park, Faulkner, & Schaller, 2003; Park, Schaller, & Crandall, 2007), it may be the case that the str ength of association between cue and perception of health threats varies according to the diagnosticity of the threat cue If so, then an attempt to reduce prejudice by breaking the link between cue and perception of threat when the cue is highly diagnosti c of threat (e.g., when the cue is a rash or oozing sores) may be less successful than an attempt to reduce prejudice by breaking that same link when the cue is not diagnostic of a contagious health threat (e.g., when the cue is obesity) Current Research and Hypotheses Overview Conceptually, the purpose of this study is to examine the effectiveness of a prejudice reduction intervention that breaks or weakens the link between cue of threat and perception of threat ( Figure 1 2 ) Further, this study will exa mine the effectiveness of this type of strategy at different levels of cue diagnosticity, that is, when the cue is highly indicative of a health threat versus when it is not This study will also test for differences between participants with varying perce ptions of their own susceptibility to disease

PAGE 19

19 It may be possible to break the link between cue and perception of threat through repeated exposure to the threat over time with no negative consequences Consider, for example, hospital emergency room perso nnel and doctors who no longer experience disgust at the potential health threats of their patients despite awareness of the associated cues and threats to physical health In this experiment, however, there was not sufficient time to habituate the partici pants to the threat cue Instead, a much shorter intervention was used; because I am particularly interested in perceived health threats, I attempted to "wash away" the link between cue and perception of threat Participants (one per session) in this stu dy came to the lab and were exposed to a target who either appeared to be obese (a low diagnosticity health threat cue), have a rash or skin disease (a high diagnosticity health threat cue), or neither (no health threat) The targets in this study were con federates posing as fellow participants with whom the participant had been paired to work in a dyad Under the pretense of preparation for a taste testing task, the experimenter had the participants and confederates scrub theirs hands with hand sanitizer ( one at a time) and measured how long the participants spent washing their hands as well as how much hand sanitizer was used This constituted the Time 1 hand washing data In the intervention conditions, the experimenter had the participants wash their han ds first, and then the confederates washed their hands In the no intervention condition, the experimenter had the participants wash their hands and then ostensibly forgot to repeat the procedure with the confederates Participant and confederates then beg an the taste testing task Participants and Confederates shared candy from the same bowl for the taste testing task This sharing ensured that the confederates had touched the candy that the

PAGE 20

20 participants would be eating The participants were again taken t o wash their hands to collect Time 2 hand washing data The participants then filled out questionnaires assessing threat perceptions, disgust, and feelings of general prejudice Although this study focuses specifically on health threats and its associate d emotion profiles and prejudice, the major concepts could also be applied to various other threats and prejudices I do not mean to suggest that targets of any type of prejudice could simply wash their hands to reduce prejudiced feelings towards them; the hand washing intervention used here is specific to health threats because of associations between cleanliness and healthfulness However, the strategy of breaking the link between cue and perception of threat could also apply to, for example, threats to p hysical safety or reciprocity It would be expected that a different threat will have a different pattern of results than the ones suggested here concerning the effectiveness of a given prejudice reduction strategy Predictions I predicted that there woul d be differences in general prejudice, disgust, and behavioral reactions towards the different targets These differences were expected to depend on diagnosticity of the health threat cue and the presence of the intervention I made no a priori predictions about the effect of participants' perceived vulnerability to disease, as there is more than one plausible outcome For example, it could have been that the intervention would have the strongest effect on participants who scored high on the Perceived Vulne rability to Disease scale (PVD; Duncan, Schaller, & Park, in press) because they should have been the ones for whom cleanliness matters the most Additionally, high PVD participants have been shown to have more negative attitudes towards obese people relat ive to low PVD participants (Park, Schaller, & Crandall,

PAGE 21

21 2007), so their perceptions of threat, levels of disgust, and threat mitigating behaviors would have had more room to decrease; however, low PVD participants may have shown a floor effect because the ir disgust and perception of threat were already so low On the other hand, it could have been that for high PVD participants, merely seeing a target wash his or her hands was not strong enough to reduce the participants' disgust and perception of threat, and so would not have reduce the participant's threat mitigating behavior Moreover, participants low on PVD may have only needed a small intervention, so hand washing would have been sufficient to decrease perception of threat, disgust, and threat mitiga ting behavior Because multiple effects of PVD were possible, I tested for effects of PVD post hoc I have derived several specific hypotheses from this threat based approach to prejudice reduction concerning the other two factors I predicted: (1) The i ntervention would reduce feelings of prejudice, perception of threat, disgust, and threat mitigating behavioral reactions towards the target, and would be differentially effective depending on the diagnosticity of the threat cue; that is, the intervention would be differentially effective depending on whether the health threat cue exhibited by the target was obesity or skin disease Thus, I predicted a two way interaction between intervention (intervention vs no intervention/control) and threat cue (skin d isease vs obesity vs control) on feelings of prejudice, perception of threat, disgust, and associated behavioral reactions (e.g., amount of time spent and soap used in hand washing) I also predicted: (2) The intervention would be more effective when th e cue was non diagnostic of a health threat compared to when the cue was diagnostic of a health

PAGE 22

22 threat; that is, the difference between intervention and no intervention in feelings of prejudice, perception of threat, disgust, and the amount of time and soa p spent in hand washing would be greater for participants paired with an obese confederate than for participants paired with a skin diseased confederate Prejudice, threat perception, and disgust would be lower for participants exposed to the obesity (i.e. non diagnostic) cue than for participants exposed to the skin disease (i.e., diagnostic) cue, but only for participants in the intervention conditions Further, the amount of time and soap used in hand washing would depend on intervention (intervention v s no intervention) a nd threat cue (skin disease vs. obesity vs. control) Specifically, I predicted that, controlling for Time 1 hand washing data, Time 2 amount of soap and washing time (in seconds) would be higher in the no intervention control conditi on than in the intervention condition, but only for participants exposed to the obesity threat cue.

PAGE 23

23 Figure 1 1 Threat Management System (Neuberg & Cottrell, 2006) Figure 1 2 Threat Management System: The link between Cue and Perception of Threat

PAGE 24

24 CHAPTER 2 METHOD Participants Participants were 97 undergraduates at University of Florida Participants were compensated with credit towards a psychology course research participation requirement Design and Variables Independent variables The desi gn of this study is a 3 (health threat cue: skin disease vs obesity vs control) 2 (intervention: intervention vs no intervention/control) All factors are between subjects Threat cue was exhibited by a target, that is, a confederate posing as another participant in the study Skin disease was the diagnostic health threat cue, whereas obesity was the non diagnostic health threat cue A confederate exhibiting neither of these cues was used as a health threat cue control Conceptually, the intervention b reaks or weakens the link between cue and perception of threat This was accomplished by having the confederate wash their hands before the taste testing, as hand washing is a behavior that should reduce the spread of pathogens Knowing that the confederat e has washed his or her hands should reduce or remove perception of threat Participant gender and confederate gender were matched Dependent variables The dependent variables of interest were the amount of time participants spent scrubbing their hands with hand sanitizer, how much hand sanitizer the participants used, measures of perception of threat, measures of participants' feelings of disgust, and measures of prejudice

PAGE 25

25 Hand washing measures were included as dependent variables because hand washin g is a behavior intended to mitigate health threats Threat mitigating behavior occurs at the end of the threat management system proposed by Neuberg and Cottrell (2006), and as such, the presence and magnitude of threat mitigating behavior may be useful i ndicators of whether the link between cue and perception of threat has been weakened Furthermore, research has shown that hand washing may decrease or alleviate feelings of moral disgust and that there is a strong connection between physical and moral dis gust (e.g., Schnall, Benton, & Harvey, 2008; Zhong & Liljenquist, 2006) In a study by Schnall, Benton, and Harvey (2008), participants who washed their hands after exposure to a disgust eliciting film made less severe moral judgments than participants who did not wash their hands, presumably because less severe negative reactions to immoral behavior resulted from feelings of disgust that had been reduced due to hand washing Perception of threat was used as a dependent variable because perceptions of thr eat should occur in response to a cue of threat (Neuberg & Cottrell, 2006), and as such, should be reliable indicators of whether the link between cue and perception of threat has been weakened A questionnaire assessing perceptions of "specific threat" on a likert scale (Cook, 2009; Cottrell & Neuberg, 2005) was given to the participant The items of interest in this questionnaire were, "In general, I think that this group member increases the risk of physical sickness for me," "In general, I think that th is group member harms my medical health," and, "In general, I think that this group member poses problems for me" (Appendix B) The person referred to in these items was the confederate

PAGE 26

26 Feelings of disgust were used as a dependent variable because of th e general association between disgust and health threats (Cottrell & Neuberg, 2005; Zhong & Liljenquist, 2006), as well as more specific associations between disgust and extreme morphological deviations, such as obesity (Park, Schaller, & Crandall, 2007) Disgust occurs in response to a perception of threat in the threat management system, and so may serve as an indicator of whether the link between cue and perception of threat has been weakened A questionnaire assessing "specific emotions in response to t hreats" on a Likert scale (Cook, 2009; Cottrell & Neuberg, 2005) was given to the participants The items of interest in this questionnaire were "when I think about my impressions of this person, I feel physically disgusted by him/her," and, "when I think about my impressions of this person, I feel physically sickened by him/her" (Appendix C) The person referred to in these items is the confederate Finally, a feeling thermometer (Appendix D) and a social distancing scale (Appendix E) were used as genera l prejudice measures Feeling thermometers have been used in previous research to measure prejudice and global feelings towards a target (e.g., Miller, Smith, & Mackie, 2004) Higher ratings indicate warmer and more favorable views towards the target, wher eas lower ratings indicate colder, more negative feelings and higher levels of prejudice An adaptation of the Bogardus Social Distancing Scale (Crandall, 1991), which has also been used in previous research to measure prejudice and global behavioral react ions towards a target (e.g., Cook, 2009; Crandall, 1991) was used as an additional general measure of prejudice On this scale, participants reported willingness to actively distance themselves from another person by responding to items such as "He/she is the kind of person that I tend to avoid," on a

PAGE 27

27 Likert scale ranging from 1 to 7, with higher ratings corresponding to greater social distancing A validity study showed that "best friends" and "worst enemies" of participants received average ratings on thi s scale of 1.44 and 6.46, respectively ( t (18) = 30.24, p < .001), while several reliability studies had alphas ranging from .84 to .89, with an average of .88 (Crandall, Glor, & Britt, 1997) Appropriate items were reverse scaled (e.g., "I would like him/h er to be a close personal friend") and items were averaged to form a single social distancing score Procedure Participants came to the lab individually A confederate, posing as a participant, also arrived at the lab at the time the study was supposed to start This other confederate served as the personification of the threat cue manipulation The confederate either appeared to be obese through use of a fat suit or appeared to have some sort of rash or skin disease through the use of make up Upon arriva l, participants and confederates read and signed an informed consent document informing them that they may or may not have to complete tasks that would involve food preparation or testing After signing the informed consent, they were told that the princip al investigators were interested in examining factors that affect intragroup processes in various tasks Participants were told that they would work as a pair with the other participant on one of a variety of randomly assigned tasks Participants were told that to help another experimenter collect data, the study would involve tasks from another study Participants were told that the researcher of this other study was interested in taste preferences, so one of the tasks to which participants would be ra ndomly assigned was a taste testing study All participants were told that they and their partner had been randomly assigned to the taste testing

PAGE 28

28 task To reduce suspicion and discourage questions about the relevancy of taste testing to group processes, p articipants were also told that the taste testing task would be used as a control condition for the experimenters interested in group processes, because there is little collaborative effort in taste testing. The participants and confederates were told this information simultaneously to ensure that the participant is exposed to the threat cue After this cover story had been given, particpants were instructed to wash their hands with sanitizer before the taste testing task, supposedly to ensure that the stud y was in compliance with university health codes In reality, this was for collecting data on the hand washing dependent variables; the experimenter timed how long the participants spent scrubbing their hands, as well as how much hand sanitizer (measured i n "pumps" of a hand sanitizer bottle) was used In the intervention conditions, participants and confederates were walked over to the other side of the room, one at a time, to where the sanitizer was The participant was always taken to wash his or her han ds first so that the participant would not see or be influenced by the amount of sanitizer or time the confederate spent scrubbing his or her hands In the control (no intervention) conditions, the confederate did not wash his or her hands; the experimente r had the participant wash his or her hands first, but then pretended to be distracted and forgot to have the confederate wash his or her hands before beginning the taste testing To account for individual differences in hand washing habits, these measures were taken twice The hand washing data collected before the taste testing was the Time 1 hand washing data were used as covariates in the analyses to control for individual differences

PAGE 29

29 After washing their hands, participants and confederates sat togeth er at a table and were given instructions for the taste testing Participants were informed that they would be taste testing two different candies, M&M's and Skittles, and that they would be rating them along several dimensions on scales provided to them ( Appendix A) Participants were instructed by the experimenter to taste at least one candy of each color before filling out the questionnaire for that candy This process was to be completed for both candies For example, participants were instructed to tas te at least one red, orange, yellow, green, blue, and brown M&M, fill out the taste questionnaire for M&M's, and then move on to Skittles, repeating the same process The experimenter brought out two bowls filled with the two different candies, as well as napkins and questionnaires The experimenter brought out one bowl at a time before bringing out the questionnaires When the experimenter sat down each bowl, the confederate sorted the candies, placing them on napkins by candy type and color This was to ensure that the participant saw the confederate touching most, if not all, of the candy before the participant took any of it If participant looked confused or asked as to why the confederate was sorting or touching all the candy, the confederate responde d that they simply thought it would make things easier to sort the candy first since the experimenter told them they had to taste at least one of each color After all the bowls and questionnaires had been brought to the table, the experimenter instructed the participant and confederate to begin the taste testing task The confederate and participant completed the taste testing task according to the instructions After completing the taste testing task, participants in both the intervention and no interven tion conditions were told that they would complete the task again for another

PAGE 30

30 set of foods This time, however, there would be a small amount of food preparation involved, so they would need to wash their hands again The same procedures as mentioned above were followed to collect Time 2 hand washing data, with the exception that at Time 2, regardless of intervention condition, the experimenter always remembered to have the confederate also wash his or her hands The experimenter left ostensibly to get th e food and materials needed for the next taste testing task The experimenter returned a minute later to inform the participants that they did not have the other food necessary to complete the second taste testing task, so they would skip it The participa nts and confederates were then separated and seated at computers in private cubicles to complete the rest of the dependent measures Participants were told that the questionnaires to be filled out on the computer were for the researchers interested in intr agroup processes They were reminded that the intragroup processes researchers were interested in a large variety of factors, so they might be asked a lot of questions, including questions about their personal beliefs, current thoughts and feelings, and fe elings towards their partner This served to reduce salience of the disgust and prejudice questions and reduce suspicion Participants then completed measures of threat perception (Appendix B), disgust (Appendix C), a feeling thermometer measure of prejudi ce (Appendix D), and a social distancing measure of prejudice (Appendix E) The questionnaire also included the PVD scale (Appendix F), which was used as a covariate in some of the following analyses Additionally, the questionnaire included a demographic information page (Appendix G) After completing this last questionnaire, participants were probed for suspicion and debriefed.

PAGE 31

31 Predictions I predicted that there would be differences in perception of threat, disgust, general measures of prejudice, and beh avioral reactions towards the different targets These differences would depend on diagnosticity of the cue to health threat and the presence of the intervention, which removes immediacy of the threat Hypothesis 1a : There would be an interaction between intervention and health threat cue on perception of threat in response to the confederate Hypothesis 1b: There would be an interaction between intervention and health threat cue on feelings of disgust towards the confederate Hypotheses 1c and 1d: Ther e would be an interaction between intervention and health threat cue on a social distancing measure of prejudice (1c), as well as on a feeling thermometer measure of prejudice (1d), towards the confederate Hypotheses 1e and 1f: There would be an interact ion between intervention and health threat cue on (1d) time, in seconds, the participants spent washing their hands and (1e) the amount (or number of pumps) of sanitizer used. Hypothesis 2a: The intervention would be more effective at reducing perceptions of threat for non diagnostic health threat cues than for diagnostic health threat cues; the reduction of threat perception (compared to when there was no intervention) would be greater when the confederate appeared obese compared to when the confederate ap peared to have a skin disease Hypothesis 2b: The intervention would be more effective at reducing feelings of disgust for non diagnostic health threat cues than for diagnostic health threat cues; the reduction of disgust (compared to when there was no in tervention) would be greater when the confederate appeared obese compared to when the confederate appeared to have a skin disease Hypothesis 2c and 2d: The intervention would be more effective at reducing feelings of prejudice for non diagnostic health t hreat cues than for diagnostic health threat cues; that is, the reduction of prejudiced feelings (compared to when there was no intervention) would be greater when the confederate appeared obese compared to when the confederate appeared to have a skin dise ase Hypotheses 2e and 2f: The intervention would result in a greater reduction of time, in seconds, spent washings hands (2d) and amount of sanitizer used (2e) for participants who interacted with an ostensibly obese confederate compared to participants who interacted with an ostensibly skin diseased confederate

PAGE 32

32 CHAPTER 3 RESULTS Preliminary Analyses To assess perceived threat, three items from the Questions of Specific Threat scale representing health threat and threat in general were combined to fo rm an overall on harms my medical log transformed Marginal means were then transformed back into their original metric (e.g., see Tables 3 3 & 3 4) To assess disgust, two items from the Questions of Specific Emotions in Response to Threats scale represent ing physical disgust were combined to form an about my impressions of this person, I feel p was detected and excluded from analyses Additionally, four items from the Questions of Specific Emotions in Response to Threats scale representing general negative emotions or dislike were combined to form an o To assess social distancing, a general measure of prejudice, all items from Crandall's (1991) adaptation of the Bogardus Social Distancing Scale were combined to

PAGE 33

33 All items were reverse scored except fo A feeling thermometer provided an additional measure of general prejudice The feeling thermometer was presented to participants on the computer as an image of a scale 1000 pixels wide (0 to 999) The left side of the scale corresponded to cool and unfavorable feelings (recorded as 0) and the right side of the scale corresponded to warm and favorable feelings (recorded as 999) Participants used the cursor to click anywhere on the scale to report how warm or cold they felt towards the confederate and the x coordinate of where the participant clicked on the image was recorded The lower the value of the x coordinate, the cooler and less favorably the participant rated his/her feelings towards the co nfederate (see Appendix D). As measures of behavioral response to perceived threat and disgust, the amount of time and hand sanitizer participants used scrubbing their hands at Time 2 which took place after the taste testing were analyzed The amount of time participants spent scrubbing their hands at Time 2 was measured in seconds The amount of sanitizer participants used was measured in pumps from the hand sanitizer bottle Because the data for seconds spent hand washing at Time 2 were positively sk ewed, the data were natural log transformed Marginal means were then transformed back into their original metric (e.g., see Tables 3 10 & 3 11) Correlations and descriptive statistics for all dependent variables are presented in Table 3 1. The natural logs of perceived threat, seconds washing at Time 1, and seconds washing at Time 2 were used for the correlations The means presented in

PAGE 34

34 Table 3 1 however, are presented in their original metric The disgust outlier was removed for all correlations invol ving disgust Self Report Measures As there were no a priori predictions made about the effects of perceived vulnerability to disease (PVD) on the dependent measures, the self report measures were analyzed both with and without participants' PVD scale scores as a covariate I first present results for the self report measures analyzed without the PVD as a covariate Contrast coding was used to test for specific effects and these effects were entered as predictors into a multiple regression equation The effects entered as predictors into the regression equation were the intervention effect (confederate washes control/no threat cue condition to the average of the obese and rash threat cue conditions), and the obese vs rash effect (comparing the obese and rash threat cue conditions) The interactions tested were the interaction between intervention and control vs else and the interaction between intervention and obese vs rash Contrast codes for these predictors are presented in Table 3 2 This regression equation was used to test for effects in all self report and behavioral measures This multiple regression approach is analogous to performing traditional AN(C)OVAs. Perceived threat To test for these effects with the dependent variable of perceived threat, the natural log of perceived threat was entered as a dependent variable into the regression equation There was a significant two way interaction between interven tion and threat cue, t (91) = 2.29, p < .05, pr = .23 When the confederate appeared obese, there seemed to be no difference in perceived threat between the conditions in which the confederate did and did not wash his or her hands

PAGE 35

35 before the taste testing ( M = 1.43, SD = 1.61, and M = 1.58, SD = 1.82, respectively) However, when the confederate appeared to have a rash, participants reported greater perceived threat when the confederate did wash his or her hands ( M = 1.77, SD = 1.58) than when the confedera te did not wash his or her hands ( M = 1.14, SD = 1.25) In other words, the intervention appeared to backfire when the confederate appeared to have a rash, resulting in increased perceived threat rather than decreased perceived threat To examine the eff ect of PVD on perceived threat, participants' PVD score was added as a covariate in the regression equation The significant, two way, intervention obese vs rash interaction was again significant That is, controlling for PVD, there was a significant in teraction between intervention and threat cue, specifically involving the obese and rash confederate conditions, t (90) = 2.11, p <.05, pr = .22 Controlling for PVD, when the confederate appeared obese, there was no significant difference in perceived thr eat between the conditions in which the confederate did and did not wash his or her hands before the taste testing ( M = 1.47, SE =1.14 and M = 1.56, SE =1.12 respectively) However, when the confederate appeared to have a rash, participants reported gr eater perceived threat when the confederate did wash his or her hands ( M = 1.76, SE = 1.13) than when the confederate did not wash his or her hands ( M = 1.13, SE =1.11 ) In other words, controlling for PVD, the intervention appeared to backfire when the c onfederate appeared to have a rash, resulting in increased rather than decreased perceived threat Disgust For disgust, there was a significant two way interaction between intervention and threat cue, t (90) = 2.24, p < .05, pr = .23 When the confedera te

PAGE 36

36 appeared obese, participants reported greater disgust when the confederate did not wash his or her hands ( M = 1.36, SD = 0.77) than when the confederate did wash his or her hands ( M = 1.09, SD = 0.20) However, when the confederate appeared to have a ra sh, participants reported greater perceived threat when the confederate did wash his or her hands ( M = 1.50, SD = 0.65) than when the confederate did not wash his or her hands ( M = 1.11, SD = 0.27) The intervention appeared to decrease participants' feeli ngs of disgust when the confederate appeared obese, but appeared to backfire when the confederate appeared to have a rash, resulting in increased rather than decreased disgust The pattern of results for disgust, in the case of the rash confederate conditi on, resembles the pattern of results for perceived threat To examine the effects of PVD on disgust, participants' PVD score was added as a covariate in the regression equation The significant, two way, intervention obese vs rash interaction was again significant That is, controlling for PVD, there was a significant interaction between intervention and threat cue, specifically involving the obese and rash confederate conditions, t (89) = 2.38, p < .05, pr = .24 C ontrolling for PVD, when the confederate appeared obese, participants reported greater disgust when the confederate did not wash his or her hands ( M = 1.37, SE = 0.14) than when the confederate did wash his or her hands ( M = 1.06, SE = 0.17) However, when the confederate appeared to have a rash, participants reported greater perceived threat when the confederate did wash his or her hands ( M = 1.51, SE = 0.15) than when the confederate did not wash his or her hands ( M = 1.12, SE = 0.13) Controlling for PVD the intervention appeared to decrease participants' feelings of disgust when the confederate appeared obese, but appeared to backfire when the confederate appeared

PAGE 37

37 to have a rash, resulting in increased rather than decreased disgust The pattern of resul ts for disgust, in the case of the rash confederate condition, resembles the pattern of results for perceived threat Negative affect For negative affect, none of the specific effects entered as a predictor into the regression equation was significant This seems to suggest that any differences in general prejudice (social distancing and feeling thermometer scores), as well as behavior responses to perceived threat and emotions in response to that threat (amount of time spent and hand sanitizer used scru bbing hands at Time 2) are not likely due to negative affect towards the confederate in general, but specifically due to disgust towards the confederate To examine the effects of PVD on general negative affect, participants' PVD score was entered as a c ovariate in the regression equation Even after adding PVD as a covariate, none of the specific effects that were entered as predictors into the regression equation was significant, nor was there a significant effect of PVD itself Once again, this seems t o suggest that any differences in general prejudice (social distancing and feeling thermometer scores), as well as behavior responses to perceived threat and emotions in response to that threat (amount of time spent and hand sanitizer used scrubbing hands at Time 2) are not likely due to negative affect in general, but specifically due to disgust towards the confederate Social distancing For social distancing, there was a marginally significant two way interaction between intervention and threat cue, t ( 91) = 1.71, p = .09, pr = .18 In the obese confederate conditions, participants reported greater social distancing when the confederate did not wash his or her hands ( M = 3.98, SD = 1.07) than when the

PAGE 38

38 confederate did wash his or her hands ( M = 3.71, SD = 1.02) However, in the skin diseased confederate conditions, participants reported greater social distancing when the confederate did wash his or her hands ( M = 3.98, SD = 0.94 ) than when the confederate did not wash his/her hands ( M = 3.41, SD = 0.73) Although this interaction is not significant, the pattern resembles the pattern of results for perceived threat; the intervention appears to backfire when the confederate appears to have a rash or skin disease, resulting in increased rather than decrease d social distancing when the confederate washes his or her hands To examine the effects of PVD on social distancing, participants' PVD score was added as a covariate in the regression equation After adding PVD as a covariate, the marginally significant two way, intervention obese vs rash interaction was no longer marginally significant, t (90) = 1.62, p = .11, pr = .17 Additionally, none of the other specific effects was significant Feeling thermometer For the feeling thermometer, there was a significant two way inter action between intervention and threat cue, t (91) = 2.37, p < .05, pr = .24 In the obese confederate conditions, participants reported cooler feelings toward the confederate when the confederate did not wash his or her hands ( M = 489, SD = 217) than when the confederate did wash his/her hands ( M = 621, SD = 114) However, in the skin diseased confederate conditions, participants reported cooler feelings toward the confederate when the confederate did wash his/her hands ( M = 458, SD = 135) than when the con federate did not wash his/her hands ( M = 557, SD = 198) Unlike the other dependent variables presented so far, the feeling thermometer is scored such that higher scores would correspond to less prejudice Hence, the pattern of results again

PAGE 39

39 resembles the pattern of results for the above dependent variables; the intervention appears to backfire when the confederate appears to have a rash or skin disease, resulting in cooler rather than warmer feelings towards the confederate when the confederate washes his or her hands To examine the effects of PVD on the feeling thermometer, participants' PVD score was added as a covariate in the regression equation The significant, two way, intervention obese vs rash interaction was again significant That is, contr olling for PVD, there was a significant interaction between intervention and threat cue, specifically involving the obese and rash confederate conditions, t (90) = 2.27, p < .05, pr = .23 In the obese confederate conditions, participants reported cooler fe elings toward the confederate when the confederate did not wash his or her hands ( M = 492, SE = 47) than when the confederate did wash his or her hands ( M = 617, SE = 56) However, in the skin diseased confederate conditions, participants reported cooler f eelings toward the confederate when the confederate did wash his or her hands ( M = 459, SE = 50) than when the confederate did not wash his or her hands ( M = 558, SE = 43) Remember that unlike the other self report measures, the feeling thermometer is sco red such that higher scores would correspond to less prejudice Hence, the pattern of results again resembles the pattern of results for the above dependent variables; the intervention appears to backfire when the confederate appears to have a rash or skin disease, resulting in cooler rather than warmer feelings towards the confederate when the confederate washes his or her hands When PVD was allowed to interact with the effects for the manipulated variables, the two way interaction between PVD and threa t cue was significant (i.e., comparing the

PAGE 40

40 control/no threat cue condition to the average of the obese and rash confederate conditions), t (85) = 2.05, p < .05, pr = .22 This interaction was probed by conducting simple effects tests at 1 SD from the mean PVD score At 1 SD below the PVD mean, the control group had a predicted feeling thermometer score of 491, whereas the obese and rash group had a predicted feeling thermometer score of 564 (a difference of 73 pixels) At 1 SD above the PVD mean, the contro l group had a predicted feeling thermometer score of 681, whereas the obese and rash group had a predicted feeling thermometer score of 507 (a difference of 174 pixels) Thus, the difference between the control condition and the average of the obese and r ash conditions is larger for participants with high PVD than it is for participants with low PVD Further, participants with high PVD scores show higher predicted feelings of warmth to a confederate who exhibits no threat cue than they do to a confederate who appears to be obese or have a rash On the other hand, participants with low PVD show lower predicted feelings of warmth to a confederate who exhibits no threat cue than they do to a confederate who appears to be obese or have a rash After controll ing for PVD interactions, the significant, two way, intervention obese vs rash interaction remained significant That is, controlling for PVD, there was a significant interaction between intervention and threat cue, specifically involving the obese and rash confederate conditions, t (85) = 2.03, p < .05, pr = .21 Behavioral Measures Time spent scrubbing hands As for the self report measures, contrast coding was used to test for specific effects and these effects were entered as predictors into a multi ple regression equation In Model 1 of this multiple regression, no covariates were

PAGE 41

41 entered into the regression equation The amount of time that participants spent scrubbing their hands with sanitizer at Time 2 vari able There was a significant two way interaction between intervention and threat cue, t (91) = 2.20, p < .05, pr = .23 In the rash confederate condition, participants spent more time scrubbing their hands with hand sanitizer at Time 2 when the confederate did not wash his or her hands at Time 1 ( M = 7.60 s, SD = 1.63 s) than when the confederate did wash his or her hands at Time 1 ( M = 6.57 s, SD = 1.42 s) However, in the obese confederate condition, participants spent more time scrubbing their hands at T ime 2 when the confederate did wash his or her hands at Time 1 ( M = 9.37 s, SD = 1.55 s) than when the confederate did not wash his or her hands at Time 1( M = 6.79 s, SD = 1.38 s) The intervention appears to backfire when the confederate appears obese, re sulting in the participant spending more time scrubbing his or her hands at Time 2 when the confederate washes his or her hands at Time 1 This pattern of results is the opposite of the pattern of results for the self report measures (without PVD as a cova riate) mentioned above In Model 2 of this regression equation, the amount of time participants spent scrubbing their hands at Time 1 was mean centered and entered as a covariate to control for individual differences in hand washing There was a signific ant main effect of the covariate, t (90) = 6.42, p < .001, pr = .56, such that participants who spent more time scrubbing their hands at Time 1 spent more time scrubbing their hands at Time 2 After controlling for individual differences in time spent scrub bing hands, the two way interaction between intervention and threat cue was no longer significant, t (90) = 1.61, p = .11, pr = .17, though the pattern of results remained the same Controlling for washing

PAGE 42

42 at Time 1, participants in the rash confederate con dition spent more time scrubbing their hands with hand sanitizer at Time 2 when the confederate did not wash his or her hands at Time 1 ( M = 7.85 s, SD = 1.08 s) than when the confederate did wash his or her hands at Time 1 ( M = 6.67 s, SD = 1.10s ) Howev er, participants in the obese confederate condition spent more time scrubbing their hands at Time 2 when the confederate did wash his or her hands at Time 1 ( M = 7.98 s, SD = 1.11 s) than when the confederate did not wash his or her hands at Time 1( M = 6.9 9 s, SD = 1.09 s) The intervention appears to backfire when the confederate appears obese, resulting in the participant spending more time scrubbing his or her hands at Time 2 when the confederate washes his or her hands at Time 1 This pattern of results resembles the pattern of the results for washing at Time 2 before the addition of the covariate and is the opposite of the pattern of results for the self report measures mentioned above In Model 3 of this regression, washing at Time 1 was allowed to i nteract with the effects of the manipulated variables In this model, the effect of the covariate remained significant, t (85) = 5.96, p < .001, pr = .54, but there were no significant interactions either with or without the Time 1 covariate Amount of sa nitizer used for scrubbing hands As with the washing at Time 2 dependent variable, contrast coding was used to test for specific effects and these effects were entered as predictors into a multiple regression equation where the dependent variable was the amount of hand sanitizer participants used at Time 2, measured in number of pumps None of these effects was significant in Model 1 In Model 2 of this regression equation, the number of pumps participants used to scrub their hands at Time 1 was mean cen tered and entered as a covariate to control

PAGE 43

43 for individual differences in hand washing There was a significant main effect of the covariate, t (90) = 4.03, p < .001, pr = .39, such that participants who used more hand sanitizer at Time 1 used more hand san itizer at Time 2 No other effects were significant

PAGE 44

44 Table 3 1. Correlations for Dependent Variables and Descriptive Statistics Perceived Threat Disgust Negative Affect Social Distancing Feeling Thermometer Correlations Perceived Threat Disgust .49** Negative Affect .34** .55* Social Distancing .26* .26** .62** Feeling Thermometer .17 .18 .49** .41** Pumps (time1) .16 .25* .19 .14 .13 Pumps (time2) .144 .32** .07 .08 .03 Seconds (time1) .01 .004 .04 .002 .17 Seco nds (time2) .06 .10 .11 .11 .24* PVD .02 .10 .15 .01 .02 Descriptives Mean 1.34 1.23 2.61 3.60 555.86 SD 1.53 0.55 0.79 0.88 175.91 ** p < .01. p < .05.

PAGE 45

45 Table 3 1. Continued Pumps (time1) Pumps (time2 ) Seconds (time1) Seconds (time2) PVD Correlations Perceived Threat Disgust Negative Affect Social Distancing Feeling Thermometer Pumps (time1) Pumps (time2) .40** Seconds (time1) .07 .12 Seconds (time2) .20 .284** .57** PVD .14 04 .13 .03 Descriptives Mean 1.08 1.08 8.77 7.73 3.68 SD 0.31 0.28 1.46 1.51 0.83 ** p < .01. p < .05.

PAGE 46

46 Table 3 2. Contrast Coding for Multiple Regression Predictors Intervention Absent Intervention Present Control Obese Rash Control Obe se Rash Intervention 1 1 1 1 1 1 Control vs. Else 2 1 1 2 1 1 Obese vs. Rash 0 1 1 0 1 1 Intervention Control vs. Else 2 1 1 2 1 1 Intervention Obese vs. Rash 0 1 1 0 1 1 Table 3 3. Self Report Measures (without PVD as a Covariat e) Control Obese Rash M SD M SD M SD Perceived Threat Does Not Wash Hands 1.33 1.62 1.58 1.82 1.14 1.25 Wash Hands 1.16 1.31 1.43 1.61 1.77 1.58 Disgust Does Not Wash Hands 1.30 0.67 1.36 0.77 1.11 0.27 Wash Hands 1.00 0 1.09 0.2 1.50 0.65 N egative Affect Does Not Wash Hands 2.67 1.00 2.99 1.21 2.47 0.74 Wash Hands 2.37 0.66 2.59 0.42 2.79 0.82 Social Distancing Does Not Wash Hands 3.47 0.79 3.98 1.07 3.41 0.73 Wash Hands 3.46 1.05 3.71 1.02 3.98 0.94 Feeling Thermometer Does Not Wash H ands 571.52 146.02 489.40 217.20 556.78 198.29 Wash Hands 599.92 245.79 621.27 113.67 458.31 134.96

PAGE 47

47 Table 3 4. Natural Log Transformed Perceived Threat (with PVD as a mean centered covariate) Control Obese Rash M SE M SE M SE Does Not Wash Ha nds 0.29 0.09 0.44 0.11 0.12 0.10 Wash Hands 0.15 0.12 0.38 0.14 0.57 0.12 Table 3 5 Perceived Threat (PVD covariate, mean centered) Control Obese Rash M SE M SE M SE Does Not Wash Hands 1.34 1.09 1.56 1.12 1.13 1.11 Wash Hands 1.16 1.13 1.47 1.14 1.76 1.13 Table 3 6. Disgust (PVD covariate, mean centered) Control Obese Rash M SE M SE M SE Does Not Wash Hands 1.29 0.10 1.37 0.14 1.12 0.13 Wash Hands 1.00 0.15 1.06 0.17 1.51 0.15 Table 3 7. General Negative Affect (PVD covariate mean centered) Control Obese Rash M SE M SE M SE Does Not Wash Hands 2.66 0.17 3.01 0.23 2.48 0.21 Wash Hands 2.37 0.25 2.57 0.27 2.79 0.25 Table 3 8. Social Distancing (PVD covariate, mean centered) Control Obese Rash M SE M SE M SE Do es Not Wash Hands 3.48 0.18 3.97 0.24 3.4 0.22 Wash Hands 3.46 0.25 3.74 0.28 3.98 0.25

PAGE 48

48 Table 3 9. Feeling Thermometer (PVD covariate, mean centered) Control Obese Rash M SE M SE M SE Does Not Wash Hands 570.73 34.9 491.96 47.18 557.86 42.76 Wa sh Hands 600.14 50.22 617.16 55.54 458.76 50.23 Table 3 10. Natural Log Transformed Seconds Scrubbing Hands at Time 2 (Without Time 1 covariate) Control Obese Rash M SD M SD M SD Does Not Wash Hands 2.11 0.39 1.92 0.32 2.03 0.49 Wash Hands 2.09 0.43 2.24 0.44 1.87 0.35 Table 3 11. Seconds Scrubbing Hands at Time 2 (Without Time 1 covariate) Control Obese Rash M SD M SD M SD Does Not Wash Hands 8.21 1.48 6.79 1.38 7.6 1.63 Wash Hands 8.11 1.54 9.37 1.55 6.47 1.42 Table 3 12. Pumps 2 (Without Time 1 covariate) Control Obese Rash M SD M SD M SD Does Not Wash Hands 1.000 0 1.200 0.41 1.000 0 Wash Hands 1.077 0.28 1.182 0.41 1.154 0.38

PAGE 49

49 Table 3 13. Behavioral Measures (Time 1 covariate, mean centered) Control Obese Rash M SE M SE M SE Seconds (Time 2) Does Not Wash Hands Mean Seconds (Time 1) 8.21 1.07 6.99 1.09 7.85 1.08 Wash Hands Mean Seconds (Time 1) 8.35 1.10 7.98 1.11 6.67 1.10 Pumps (Time 2) Does Not Wash Hands Mean Pumps (Time 1) 1.02 0.05 1.23 0.07 1.03 0.06 Wash Hands Mean Pumps (Time 1) 1.11 0.07 1.11 0.08 1.07 0.07

PAGE 50

50 Table 3 14 Regression Perceived Threat (ln) B SE t Sig. pr Model 1 Intercept 0.33 0.05 7.00 .00 Intervention 0.03 0.05 0.73 .47 0.08 Control vs. Else 0.05 0.03 1.71 09 .18 Obese vs. Rash 0.03 0.06 0.51 .62 .05 Intervention Control vs. Else 0.05 0.03 1.62 .11 .17 Intervention Obese vs. Rash 0.14 0.06 2.29 .02 .23 Model 2 Intercept 0.33 0.05 7.01 .00 Intervention 0.04 0.05 0.86 .39 .09 Contr ol vs. Else 0.05 0.03 1.7 .09 .18 Obese vs. Rash 0.03 0.06 0.58 .56 .06 Intervention Control vs. Else 0.06 0.03 1.75 .08 .18 Intervention Obese vs. Rash 0.13 0.06 2.11 .04 .22 PVD (centered) 0.05 0.05 1.00 .32 .11

PAGE 51

51 Table 3 15. Regressio n Disgust B SE t Sig. pr Model 1 Intercept 1.23 0.06 21.39 .00 Intervention 0.03 0.06 0.51 .62 .05 Control vs. Else 0.04 0.04 1.00 .32 .11 Obese vs. Rash 0.04 0.07 0.56 .58 .06 Intervention Control vs. Else 0.06 0.04 1.54 .13 .16 Inter vention Obese vs. Rash 0.16 0.07 2.24 .03 .23 Model 2 Intercept 1.22 0.06 21.38 .00 Intervention 0.04 0.06 0.64 .53 .07 Control vs. Else 0.04 0.04 1.01 .32 .11 Obese vs. Rash 0.05 0.07 0.68 .50 .07 Intervention Control vs. Else 0.05 0.0 4 1.37 .17 .14 Intervention Obese vs. Rash 0.18 0.07 2.38 .02 .24 PVD (centered) 0.08 0.07 1.16 .25 .12

PAGE 52

52 Table 3 16. Regression General Negative Affect B SE t Sig. pr Model 1 Intercept 2.65 0.09 28.38 .00 Intervention 0.07 0.09 0 .70 .49 .07 Control vs. Else 0.07 0.06 1.03 .31 .11 Obese vs. Rash 0.08 0.12 0.69 .50 .07 Intervention Control vs. Else 0.04 0.06 0.68 .50 .07 Intervention Obese vs. Rash 0.18 0.12 1.52 .13 .16 Model 2 Intercept 2.65 0.09 28.26 .00 Intervention 0.07 0.09 0.75 .46 .08 Control vs. Else 0.07 0.06 1.03 .31 .11 Obese vs. Rash 0.08 0.12 0.64 .52 .07 Intervention Control vs. Else 0.04 0.06 0.60 .55 .06 Intervention Obese vs. Rash 0.19 0.12 1.57 .12 .16 PVD (centered) 0. 05 0.11 0.49 .63 .05

PAGE 53

53 Table 3 17 Regression Feeling Thermometer B SE t Sig. pr Model 1 Intercept 3.67 0.10 37.97 .00 Intervention 0.05 0.10 0.51 .61 .05 Control vs. Else 0.10 0.07 1.54 .13 .16 Obese vs. Rash 0.08 0.12 0.64 .53 .07 Intervention Control vs. Else 0.03 0.07 0.41 .68 .04 Intervention Obese vs. Rash 0.21 0.12 1.71 .09 .18 Model 2 Intercept 3.67 0.10 37.80 .00 Intervention 0.06 0.10 0.56 .58 .06 Control vs. Else 0.10 0.07 1.53 .13 .16 Obese vs. Rash 0.08 0.12 0.66 .51 .07 Intervention Control vs. Else 0.03 0.07 0.47 .64 .05 Intervention Obese vs. Rash 0.20 0.13 1.62 .11 .17 PVD (centered) 0.05 0.11 0.41 .68 .04

PAGE 54

54 Table 3 18. Regression Social Distancing B SE t Sig. pr Model 1 Intercep t 549.53 19.09 28.78 .00 Intervention 10.3 19.09 0.54 .59 .06 Control vs. Else 18.09 12.97 1.4 .17 .15 Obese vs. Rash 23.9 24.27 0.99 .33 .10 Intervention Control vs. Else 1.95 12.97 0.15 .88 .02 Intervention Obese vs. Rash 57.59 24.27 2 .37 .02 .24 Model 2 Intercept 549.34 19.18 28.64 .00 Intervention 9.25 19.36 0.48 .63 .05 Control vs. Else 18 13.03 1.38 .17 .14 Obese vs. Rash 23.13 24.45 0.95 .35 .10 Intervention Control vs. Else 2.73 13.17 0.21 .84 .02 Intervent ion Obese vs. Rash 56.08 24.67 2.27 .03 .23 PVD (centered) 9.04 22.44 0.4 .69 .04

PAGE 55

55 Table 3 19. Regression Natural Log Transformed Seconds Scrubbing Hands, Time 2 B SE t Sig. pr Model 1 Intercept 2.04 0.04 47.24 .00 Intervention 0.03 0 .04 0.57 .57 .06 Control vs. Else 0.03 0.03 0.99 .33 .10 Obese vs. Rash 0.06 0.06 1.17 .24 .12 Intervention Control vs. Else 0.02 0.03 0.53 .60 .06 Intervention Obese vs. Rash 0.12 0.06 2.20 .03 .23 Model 2 Intercept 2.03 0.04 56. 46 .00 Intervention 0.003 0.04 0.07 .94 .01 Control vs. Else 0.04 0.03 1.61 .11 .17 Obese vs. Rash 0.02 0.05 0.34 .74 .04 Intervention Control vs. Else 0.01 0.03 0.22 .83 .02 Intervention Obese vs. Rash 0.07 0.05 1.61 .11 .17 PVD (centered ) 0.61 0.10 6.42 .00 .56

PAGE 56

56 Table 3 20 Regression Pumps 2 B SE t Sig. pr Model 1 Intercept 1.10 0.03 38.48 .00 Intervention 0.04 0.03 1.24 .22 .13 Control vs. Else 0.03 0.02 1.64 .11 .17 Obese vs. Rash 0.06 0.04 1.57 .12 .16 Interven tion Control vs. Else 0.002 0.02 0.08 .94 .01 Intervention Obese vs. Rash 0.04 0.04 1.18 .24 .12 Model 2 Intercept 1.07 0.03 37.99 .00 Intervention 0.003 0.03 0.11 .92 .01 Control vs. Else 0.02 0.02 0.91 .37 .10 Obese vs. Rash 0.06 0.03 1.79 .08 .19 Intervention Control vs. Else 0.02 0.02 1.13 .26 .12 Intervention Obese vs. Rash 0.04 0.03 1.18 .24 .12 PVD (centered) 0.36 0.09 4.03 .00 .39

PAGE 57

57 CHAPTER 4 DISCUSSION While a view of prejudice as affect on a negative positive conti nuum may be common and intuitive, this view masks the complexity of the different emotional and behavioral reactions that could potentially result from a social interaction (Cottrell & Neuberg, 2005). Failing to recognize and account for this complexity o f emotional reactions to others may hinder efforts to reduce prejudice. Therefore, this study assumes a threat based approach to prejudice. According to the threat based approach to prejudice, a cue of threat leads to a perception of threat. This perc eption of threat leads to emotions and cognitions associated with that threat, which then lead to behavioral inclinations that would remove or mitigate the threat. By weakening the link between cue and perception of threat, prejudice may be reduced, but t he effectiveness of this strategy may depend on the type of threat cue, specifically, how diagnostic the threat cue is of actual threat. In this study, a prejudice reduction intervention designed to break the link between cue and perception of health thre at was tested for its effectiveness depending on the diagnosticity of the health threat cue. Confederates in this study appeared either to have a rash (high diagnosticity health threat cue) or to be obese (low diagnosticity health threat cue) and, in the intervention present conditions, attempted to break the link between cue and perception of threat by scrubbing their hands with hand sanitizer. This intervention served to break the link between cue and perception of threat because while the health threat cue remained present (i.e., the confederate still appeared to have a rash or to be obese after scrubbing his or her hands), hand washing should make

PAGE 58

58 perceivers feel less concerned about the spread of pathogens, and so the perception of threat should be di minished or eliminated. Summary of Findings Hypotheses 1a 1f predicted that the intervention would be differentially effective for their respective dependent variables (perceived threat, disgust, social distancing, feeling thermometer, and amount of tim e/sanitizer used in scrubbing hands at Time 2). Hypotheses 2a 2f are qualifications about the direction of the effects in the interactions from Hypotheses 1a 1f. Hypotheses 2a 2f predicted that the intervention would be more effective for non diagnostic health threat cues than for diagnostic health threat cues; that is, the intervention would be more effective when the confederate appeared obese than when the confederate appeared to have a rash. Hypotheses 1a and 2a Perceived Threat. There was a sig nificant two way interaction between intervention and threat cue for perceived threat, t (91) = 2.29, p < .05, pr = .23; the amount of perceived threat participants felt depended on both intervention and threat cue, as predicted by Hypothesis 1a. However, the marginal means for this interaction did not support Hypothesis 2a. The intervention did not appear to reduce perceived threat when the confederate appeared obese. Furthermore, the intervention appeared to backfire when the confederate had a rash, sh owing greater perceived threat when the confederate did wash his or her hands than when the confederate did not wash his or her hands. This interaction remained significant and the pattern of the results remained the same even after controlling for PVD. Hypotheses 1b and 2b Disgust. There was a significant two way interaction between intervention and threat cue for disgust, t (90) = 2.24, p < .05, pr = .23; the

PAGE 59

59 amount of disgust participants felt depended on both intervention and threat cue, as predi cted by Hypothesis 1b. However, the marginal means of this interaction provide only partial support for Hypothesis 2b. While the intervention did appear to reduce disgust when the confederate appeared obese, the intervention appeared to increase disgust when the confederate appeared to have a rash. Thus, as was the case for perceived threat, the intervention seemed to backfire when the confederate appeared to have a rash. This interaction remained significant and the pattern of results remained the sam e even after controlling for PVD. Hypotheses 1c and 2c Social Distancing. There was a marginally significant two way interaction between intervention and threat cue for social distancing, t (91) = 1.71, p = .09, pr = .18. The amount of reported soci al distancing from the confederate a general measure of prejudice may not have depended on intervention and threat cue, as predicted by Hypothesis 1c, but the pattern of marginal means for social distancing resembles the pattern of marginal means for d isgust and, in the case of the rash conditions, the pattern of marginal means for perceived threat. For participants in the obese confederate condition, there was less reported social distancing when the confederate washed his or her hands; however, for p articipants in the rash confederate condition there was more reported social distancing when the confederate washed his or hands. Thus, although this interaction was only marginally significant, the pattern of results for social distancing resembles the p attern of results for the other self report measures so far, in that the intervention seems to backfire for participants interacting with a confederate who appears to have a rash. However, after controlling for PVD, this interaction was no longer marginal ly significant.

PAGE 60

60 Hypotheses 1d and 2d Feeling Thermometer. There was a significant two way interaction between intervention and threat cue for the feeling thermometer, t (91) = 2.37, p < .05, pr = .24; participants' feelings of warmth or coolness towa rds the confederate depended on both intervention and threat cue, as predicted by Hypothesis 1d. However, the marginal means of this interaction provide only partial support for Hypothesis 2d. While the intervention did appear to result in warmer feeling s (relative to when the intervention was absent) towards confederates who appeared obese, the intervention appeared to result in cooler feelings (relative to when the intervention was absent) towards confederates who appeared to have a rash. Thus, the int ervention seems to be working for participants interacting with confederates who appear obese, but backfiring for participants interacting with confederates who appear to have a rash. This interaction remained significant and the pattern of results remain ed the same even after controlling for PVD. Hypotheses 1e and 2e Time spent scrubbing hands. Although the results for the self report measures showed only partial support for the hypotheses, the pattern of results across dependent variables was fair ly consistent. The question of whether and in what conditions the intervention works becomes more difficult to answer with the introduction of the behavioral measures. There was a significant two way interaction between intervention and threat cue on the amount of time, measured in seconds, the participants spent scrubbing their hands at Time 2, t (91) = 2.20, p < .05, pr = .23. However, after controlling for individual differences in hand washing time by entering seconds at Time 1 as a covariate, thi s interaction was no longer significant, t (90) = 1.61, p = .11, pr = .17. Furthermore, there was a significant main effect of the

PAGE 61

61 covariate, t (90) = 6.42, p < .001, pr = .56, indicating that participants who spent more time scrubbing their hands at Time 1 spent more time scrubbing their hands at Time 2. Hence, there is only partial support for the hypothesis that the amount of time participants spent washing their hands at Time 2 depended on intervention and threat cue (Hypothesis 1e). Although the int eraction was not significant, the marginal means of this interaction revealed an interesting pattern of results a pattern that is contrary to the pattern of results for the self report measures described above. While the self report measures suggested t hat the intervention was somewhat effective (or at least, did not backfire) in the obese confederate conditions, and that the intervention backfired in the rash confederate conditions, this pattern of results for the obese versus rash confederate condition s is reversed for the behavioral dependent measure of time (in seconds) participants spent scrubbing their hands at Time 2. When the confederate appeared to have a rash, the intervention resulted in shorter hand scrubbing times at Time 2; however, when th e confederate appeared to be obese, the intervention resulted in longer hand scrubbing times at Time 2. Thus, had there been support for the hypothesis that the intervention was differentially effective (depending on intervention and threat cue) for the d ependent variable of time spent scrubbing at Time 2, the marginal means for this interaction would suggest effects contrary to the results for the self report measures. Although these results fail to support the hypothesis that the intervention would be m ore effective for the obese confederate conditions than for the rash confederate conditions (Hypothesis 2e), they nevertheless present a pattern of results that is interesting in light of the results for the self report measures. Specifically,

PAGE 62

62 contrary to the pattern of results for the self report measures, the intervention appeared to be effective for the rash confederate conditions, but appeared to backfire for the obese confederate conditions. Hypotheses 1f and 2f. The predicted interaction between intervention and threat cue for the dependent variable of amount of hand sanitizer, measured in pumps, was not significant; there was not sufficient evidence to support Hypothesis 1f, which predicted that the amount of sanitizer used scrubbing hands at Tim e 2 depended on intervention and threat cue. This interaction was also non significant when controlling for individual differences by entering pumps at Time 1 as a covariate. The only significant effect for the dependent variable of pumps of sanitizer us ed was a main effect of the Time 1 pumps covariate, indicating that participants who used more pumps scrubbing their hands at Time 1 used more pumps scrubbing their hands at Time 2. Hence, there is not sufficient evidence to support Hypotheses 1f and 2f. These hypotheses predicted that the intervention would be differentially effective at decreasing the amount of hand sanitizer used at Time 2 depending on intervention and threat cue (Hypothesis 1f) and that the intervention would be more effective fo r the obese confederate conditions than for the rash confederate conditions (Hypothesis 2f). Implications The results of this study show at least partial support for the idea that interventions designed to break the link between cue and perception of he alth threat may be effective at stopping the chain of reactions in the threat management system (see Figure 1 1 ) and, therefore, may be effective at reducing prejudice. However, the effectiveness of this kind of strategy for breaking the link between cue and perception of health threat would likely depend on the diagnosticity of a health threat cue.

PAGE 63

63 Furthermore, this concept may apply to other types of prejudice besides just health related prejudice. For example, the effectiveness of a prejudice reductio n intervention that breaks the link between cue and perception of physical safety threat would likely depend on the diagnosticity of the physical safety threat cue, just as is the case for health threat cues. However, it is still ambiguous when this typ e of intervention will be effective and when it will backfire. For the most part, results for the self report measures suggested that the intervention would be effective for an obese target but would backfire for a skin diseased target. On the other hand results for one of the behavioral measures, the amount of time participants spent scrubbing their hands at Time 2, suggested that the intervention would be effective for a skin diseased target but backfire for an obese target. The other behavioral measu re, amount of hand sanitizer participants used scrubbing their hands at Time 2, yielded results that showed insufficient evidence to conclude that the intervention was effective at all, for either obese or skin diseased targets. It may be that the interve ntion was effective at breaking or weakening the link between cue and perception of threat when the confederate appeared to have a rash when the threat cue would be diagnostic of health threat but only proximally. That is, shortly after the interventi on and while participants were still in close proximity to the confederate, the intervention may have been effective at breaking or weakening the link between cue and perception of threat. This would account for participants in the rash confederate condit ion spending less time scrubbing their own hands at Time 2 when the intervention was present than when it was absent. However, participants did not

PAGE 64

64 complete the self report measures until the end of the study when the confederate and participant were sepa rated. Participants' remembrance of the confederate scrubbing his or her hands with hand sanitizer at Time 1 (i.e., the intervention) could have actually increased the salience of the rash, thereby amplifying the threat cue but not weakening the link betw een cue and perception of threat. Basically, this intervention may have weakened the link between cue and perception of threat proximally, but only served to amplify the threat cue distally or retrospectively. However, why the intervention would be proxi mally effective and distally ineffective for the rash targets (those exhibiting a diagnostic health threat cue), but proximally effective and distally ineffective for obese targets (those exhibiting a non diagnostic health threat cue) is unclear. Moreover this explanation of proximal versus distal effectiveness is speculative and cannot be addressed by the current study. While a major implication of this study is that interventions that break the link between cue and perception of threat may be effecti ve at reducing prejudice, there are other ideas that this research may seem to imply but that I do not endorse. First, I do not mean to suggest that the simple act of hand washing is strong enough to protect targets from health threat based prejudice. Pe ople may err on the side of caution in domains of perceived physical health or safety threats, and so may experience residual perceived threat and disgust even after the link between cue and perception of threat has been weakened. It may also be the case that targets of prejudice are targets of more than one type of prejudice. Therefore, an intervention which aims to reduce a particular type of prejudice though potentially more effective than an intervention strategy that does not take the different sou rces and types of prejudice into account

PAGE 65

65 will only be effective at reducing some of the prejudice towards targets of multiple types of prejudice. Furthermore, people may have complex and multi faceted relations and histories with one another, both at th e group and individual level, and such a simple behavior may be insufficient on its own to reduce conflict and increase trust between them. Therefore, I am not suggesting that targets of perceived health threat based prejudice simply wash their hands more often and all prejudice towards them will vanish. Rather, I suggest that interventions which break the link between cue and perception of a particular threat (and these interventions may need to be much stronger than a hand washing intervention) have the potential to reduce that particular type of prejudice, and that further investigation of such interventions would be a fruitful research path. Second, and perhaps most importantly, I am not suggesting that targets of health threat based prejudice are r esponsible for the prejudice directed towards them. Similarly, I am not suggesting that it is incumbent on the targets of prejudice, rather than the actors, to reduce that prejudice. Still, it may be possible and effective for a target of prejudice to ca rry out the intervention him or her self, breaking the link between his or her own potential threat cue and another person's perception of threat. Indeed, that is essentially how the intervention was implemented in the current study. Limitations There are several limitations in this study. This study relied on a convenience sample of university undergraduates, which limits the generalizability of these findings. Furthermore, the current sample size ( N = 97) is substantially smaller than the proposed and ideal final sample size ( N = 300) for adequate statistical power to detect the

PAGE 66

66 predicted effects. Finally, this study could also have benefited from additional self report and behavioral measures. The strength and variation of the manipulations in this study are major limitations. Although all confederates were trained to apply the rash make up similarly, over time there was divergence from the standard in both the subtlety and general appearance of the rash manipulation. For example, some confede rates applied the rash make up more heavily than did others, and in some cases the make up looked more like bruises or burns than rashes. Furthermore, all confederates had to wear the same fat suit and adjust it to their own bodies as much as possible. B ecause there was substantial variation between confederates in height, the same fat suit made the shorter confederates appear more overweight than the taller confederates. Although such variation can be good for the purposes of stimulus sampling (Wells & Windschitl, 1999), this variation may have weakened the strength of the manipulations. Follow up studies should ensure greater uniformity in the manipulations. Another limitation was that the data for self report measures tended to cluster towards the lower end of the scales. Thus, there was very little perceived threat, disgust, and social distancing to reduce in the first place. Follow up studies should employ stronger manipulations to avoid floor effects. Finally, data collection for amount of tim e and hand sanitizer used was problematic. Not all pumps of hand sanitizer were equal. That is, some participants pressed the hand sanitizer pump all the way down while others only pressed the hand sanitizer pump part of the way down, resulting in differ ences between how much hand sanitizer was actually dispensed per pump. Unfortunately, data entry did not reflect

PAGE 67

67 these differences, with half pumps, whole pumps, and any fraction of pumps in between zero and one recorded as simply one pump. Thus, there m ay actually be greater variability in amount of hand sanitizer used than the data actually show. Regarding the amount of time participants spent scrubbing their hands, their knowledge that the experimenter was watching them may have affected the amount of time they spent scrubbing their hands. When participants saw and asked about the stopwatch the experimenter was using to record the amount of time they spent scrubbing their hands, they were told that this information had to be reported to the universi ty, to make sure that they were in compliance with university health codes. After receiving this information, some participants were unsure at Time 1 whether or not they w testing. Thus, the amount of time participants spent scrubbing hands in the experiment may not be analogous to the amount of time they would have spent scrubbing their hands under similar threat cue c onditions in a non laboratory setting. Further, the data at Time 1 for some participants may have been inappropriate for controlling for individual differences in amount of time spent scrubbing hands, due to the mistaken belief that they should wash their hands for a certain minimum amount of time. Follow up studies should employ more accurate methods of collecting the hand washing data. Follow up studies should also employ a variety of other behavioral measures, for example, allowing participants to cho ose whether or not to wash their hands, or allowing them to choose how close or distant they would seat themselves to the confederate.

PAGE 68

68 APPENDIX A TASTE TESTING QUESTIONNAIRE Please write the name of the candy you are rating: __________________________ _____ Please rate the candy on each of the following dimensions: Taste 1 2 3 4 5 6 7 Terrible Perfect Smell 1 2 3 4 5 6 7 Extremely Extremely Unappetizing Appetizing Appearance 1 2 3 4 5 6 7 Extremely Extremely Unappetizing Appetizing Crunchines s 1 2 3 4 5 6 7 Not enough Too Much Chewiness 1 2 3 4 5 6 7 Not enough Too Much Texture 1 2 3 4 5 6 7 Terrible Perfect Size 1 2 3 4 5 6 7 Too Small Too Big

PAGE 69

69 Quality/Freshness 1 2 3 4 5 6 7 Very bad Very good Can you taste differences between the colors ? 1 2 3 4 5 6 7 Not at all Very much Overall, how much do you like this candy? 1 2 3 4 5 6 7 Not at all Very much

PAGE 70

70 APPENDIX B QUESTIONS OF SPECIFIC THREAT Please consider your general thoughts about the person who was assigned to be your par tner in this experiment: In general, I think that this person poses a challenge to me 1 2 3 4 5 6 7 Not at all Very much In general, I think that this person increases my risk of physical sickness 1 2 3 4 5 6 7 Not at all Very much In general, I t hink that this person poses problems for me 1 2 3 4 5 6 7 Not at all Very much In general, I think that this person promotes values that directly oppose my own 1 2 3 4 5 6 7 Not at all Very much In general, I think that this person harms my medical health 1 2 3 4 5 6 7 Not at all Very much In general, I think that this person advocates values that are morally inferior to my own 1 2 3 4 5 6 7 Not at all Very much

PAGE 71

71 APPENDIX C QUESTIONS OF SPECIFIC EMOTIONS IN RESPONSE TO THREATS Please think abo ut your impressions of the person who was assigned to be your partner in this experiment. Consider your impressions of your partner in this experiment when answering the following questions: When I think about my impressions of this person, I feel respec t for him/her 1 2 3 4 5 6 7 Not at all Very much When I think about my impressions of this person, I feel positive toward him/her 1 2 3 4 5 6 7 Not at all Very much When I think about my impressions of this person, I feel morally sickened by him/he r 1 2 3 4 5 6 7 Not at all Very much When I think about my impressions of this person, I feel dislike for him/her 1 2 3 4 5 6 7 Not at all Very much When I think about my impressions of this person, I feel admiration for him/her 1 2 3 4 5 6 7 Not at all Very much When I think about my impressions of this person, I feel physically disgusted by him/her 1 2 3 4 5 6 7 Not at all Very much When I think about my impressions of this person, I feel negative towards him/her

PAGE 72

72 1 2 3 4 5 6 7 Not at all Very much When I think about my impressions of this person, I feel morally disgusted by him/her 1 2 3 4 5 6 7 Not at all Very much When I think about my impressions of this person, I feel a liking for him/her 1 2 3 4 5 6 7 Not at all Very much W hen I think about my impressions of this person, I feel physically sickened by him/her 1 2 3 4 5 6 7 Not at all Very much

PAGE 73

73 APPENDIX D FEELING THERMOMETER Please use the following scale to indicate how 'coolly' (or negative) or how 'warmly' (or positive) you feel toward the person you have been assigned to work with in this study. Make your choice by clicking along the appropriate point anywhere along the line below.

PAGE 74

74 APPENDIX E SOCIAL DISTANCING SCALE Please think about the person who was assigned to be your partner in this experiment when completing the following scales. Indicate the degree to which you agree or disagree with the following statements by circling a number on each scale below: He/she appears to be a likeable person. 1 2 3 4 5 6 7 Str ongly Agree Strongly Disagree I would like him/her to be a close personal friend. 1 2 3 4 5 6 7 Strongly Agree Strongly Disagree 1 2 3 4 5 6 7 Strongly Agree Strongly Disagree I would like him/her to come and work at the same place I do. 1 2 3 4 5 6 7 Strongly Agree Strongly Disagree He/she is a person who is similar to me. 1 2 3 4 5 6 7 Strongly Agree Strongly Disagree I would like to have him/her marry into my family. 1 2 3 4 5 6 7 Strongly Agree Strongly Disagree He/she is the kind of person that I tend to avoid 1 2 3 4 5 6 7 Strongly Agree Strongly Disagree

PAGE 75

75 APPENDIX F PERCEIVED VULNERABILITY TO DISEASE SCALE Please answer the following questions: It really bothers me w hen people sneeze without covering their mouths. 1 2 3 4 5 6 7 Not at all Very much If an illness is going around, I will get it. 1 2 3 4 5 6 7 Not at all Very much I am comfortable sharing a water bottle with a friend. 1 2 3 4 5 6 7 Not at all Very much I don't like to write with a pencil someone else has obviously chewed on. 1 2 3 4 5 6 7 Not at all Very much My past experiences make me believe I am not likely to get sick even when my friends are sick. 1 2 3 4 5 6 7 Not at all Very much I h ave a history of susceptibility to infectious diseases. 1 2 3 4 5 6 7 Not at all Very much I prefer to wash my hands pretty soon after shaking someone's hand. 1 2 3 4 5 6 7

PAGE 76

76 Not at all Very much In general, I am very susceptible to colds, flus, and ot her infectious diseases. 1 2 3 4 5 6 7 Not at all Very much I dislike wearing used clothes because you don't know what the past person who wore it was like. 1 2 3 4 5 6 7 Not at all Very much I am more likely than the people around me to catch an in fectious disease. 1 2 3 4 5 6 7 Not at all Very much My hands do not feel dirty after touching money. 1 2 3 4 5 6 7 Not at all Very much I am unlikely to catch a cold, flu, or other illness even if it is going around. 1 2 3 4 5 6 7 Not at all V ery much It does not make me anxious to be around sick people. 1 2 3 4 5 6 7 Not at all Very much My immune system protects me from most illnesses that other people get. 1 2 3 4 5 6 7 Not at all Very much

PAGE 77

77 I avoid using public telephones because of the risk that I may catch something from the previous user. 1 2 3 4 5 6 7 Not at all Very much

PAGE 78

78 APPENDIX G DEMOGRAPHICS Please answer the following questions. You may skip any question that you do not feel comfortable answering, but please remember t hat all responses are confidential and will not be linked back to you. Age: Gender: Height: Weight: Which of the following religions best describes you? ______ Buddhist ______ Catholic ______ Fundamentalist/Evangelical Christian ______ Hindu ______ Jewish ______ Muslim ______ Protestant (Methodist, Lutheran, Episcopalian, etc.) ______ Other: _________________________________________ ______ No religious affiliation Which of the following best describes you? ______ Bisexual ______ Gay/lesbian ______ Straight ______ Questioning

PAGE 79

79 8. How much money does your family make (i.e., income before taxes) in an average year? (Please estimate if unknown.) ____________________________________ 9. How much money do you make (i.e., income before taxes) in an average year? (Please estimate if unknown. Negative income is fine.) ____________________________________

PAGE 80

80 REFERENCES Amir, Y. (1969). Contact hypothesis in ethnic relations. Psychological Bulletin, 71 (5), 319 342. doi:10.1037/h0027352. Cook, C. L. (2009). A Threat Based Approach to Subgroups: Do Different Subgroups of Gay Men Elicit Different Prejudices? Florida, Gainesville, FL. Cosmides, L., & Tooby, J. (1997). Evolutionary psychology: A primer. 2009, http://www.psych.ucsb.edu/research/cep/primer.html Cottrell, C. A., & Neuberg, S. L. (2005). Different emotional reactions to different groups: A sociofunctional threat based approach to "prejudice". Journal of Personality and Social Psychology, 88 (5), 770 770. doi:10.1037/0022 3514.88.5.770. Cottrell, C. A., Neuberg, S. L., & Li, N. P. (2007). What do people desire in others? A sociofunctional perspective on the importance of different valued characteristics. Journal of Personality and Social Psychology, 92 (2), 208 208. doi:10.1037/0022 3514.92.2.208. Crandall, C. S. (1991). Multiple stigma and AIDS: Illness stigma and attitudes toward homosexuals and IV drug users in AIDS related stigmatization. Journal of Community & Applied Social Psychol ogy. Special Issue: Social Dimensions of AIDS, 1 (2), 165 172. Crandall, C. S., Glor, J., & Britt, T. W. (1997). AIDS related stigmatization: Instrumental and symbolic attitudes. Journal of Applied Social Psychology, 27, 95 123. doi:10.1111/j.1559 1816.1997.tb00625.x. Duncan, L., Schaller, M., & Park, J. (2009). Perceived vulnerability to disease: Development and validation of a 15 item self report instrument. Personality and Ind ividual Differences 47 (6), 541 546. doi:10.1016/j.paid.2009.05.001. Faulkner, J., Schaller, M., Park, J. H., & Duncan, L. A. (2004). Evolved disease avoidance mechanisms and contemporary xenophobic attitudes. Group Processes & Intergroup Relations, 7 (4), 333 353. doi:10.1177/1368430204046142. Kurzban, R., & Leary, M. R. (2001). Evolutionary origins of stigmatization: The functions of social exclusion. Psychological Bulletin, 127 (2), 187 208. doi:10.1177/1368430204046142. Kenrick, D. T., Neuberg, S. L., & Cialdini, R. B. (2007). Social psychology: Goals in interaction (4th ed.) Boston, MA, US: Allyn & Bacon.

PAGE 81

81 Miller, D. A. Smith, E. R., & Mackie, D. M. (2004). Effects of intergroup contact and political predispositions on prejudice: Role of intergroup emotions. Group Processes & Intergroup Relations, 7 (3), 221 237. doi:10.1177/1368430204046142. Nelson, T. D. (2006). The psychology of prejudice (2nd ed.) Boston, MA, US: Allyn & Bacon Neuberg, S. L., & Cottrell, C. A. (2006). Evolutionary bases of prejudices. In M. Schaller, J. A. Simpson & D. T. Kenrick (E ds.), Evolution and social psychology. (pp. 163 187). Madison, CT, US: Psychosocial Press. Neuberg, S. L., & Cottrell, C. A. (2008). Managing the threats and opportunities afforded by human sociality. Group Dynamics: Theory, Research, and Practice.Special Issue: Evolutionary Approaches to Group Dynamics, 12 (1), 63 72. Park, J. H., Faulkner, J., & Schaller, M. (2003). Evolved disease avoidance processes and contemporary anti social behavior: Prejudicial attitudes and avoidance of people with physical disabi lities. Journal of Nonverbal Behavior, 27 (2), 65 87. doi:10.1177/1368430204046142. Park, J. H., Schaller, M., & Crandall, C. S. (2007). Pathogen avoidance mechanisms and the stigmatizatio n of obese people. Evolution and Human Behavior, 28 (6), 410 414. doi:10.1177/1368430204046142 Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psychology, 49 65 85. doi:10.1177/1368430204046142. Pettigrew, T. F., & Tropp, L. R. (2006). A meta analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90 (5), 751 783. doi:10.1177/1368430204046142. Schnall, S., Benton, J., & Harvey, S. (2008). With a clean conscience: Cleanliness reduces the severity of moral judgments. Psychological Science, 19 (12), 1219 1 222. doi:10.1177/1368430204046142. Stephan, W. G. (1978). School desegregation: An evaluation of predictions made in brown v. board of education. Psychological Bulletin, 85 (2), 217 238. doi:10.1177/1368430204046142. Stephan, W. G., & Rosenfield, D. (1978). Effects of desegregation on race relations and self esteem. Journal of Educational Psychology, 70 (5), 670 679. doi:10.1177/1368430204046142. Sundie, J. M., Cialdini, R. B., Griskevicius, V., & Kenrick, D. T. (2006). Evolutionary social influence. In M. Schaller, J. A. Simpson & D. T. Kenrick (Eds.), Ev olution and social psychology. (pp. 287 316). Madison, CT, US: Psychosocial Press.

PAGE 82

82 Tooby, J., & Cosmides, L. (1990). On the universality of human nature and the uniqueness of the individual: The role of genetics and adaptation. Journal of Personality.Spec ial Issue: Biological Foundations of Personality: Evolution, Behavioral Genetics, and Psychophysiology, 58 (1), 17 17. Utsey, S. O., Ponterotto, J. G., & Porter, J. S. (2008). Prejudice and racism, year 2008 -still going strong: Research on reducing prejud ice with recommended methodological advances. Journal of Counseling & Development.Special Issue: Multicultural Counseling, 86 (3), 339 347. Wells, G., & Windschitl, P. (1999). Stimulus sampling and social psychological experimentation. Personality and Soci al Psychology Bulletin 25 (9), 1115 1125. doi:10.1177/1368430204046142. Zhong, C., & Liljenquist, K. (2006). Washing away your sins: Threatened morality and physical cleansing. Science, 3 13 (5792), 1451 1452. doi:10.1177/1368430204046142.

PAGE 83

83 BIOGRAPHICAL SKETCH Tatiana Orozco Schember was born in Mesa, Arizona in 1986. She graduated with highest honors from Hamilton High School in 2004 and enrolled at Arizona State University in August of 2004. Sh e graduated cum laude with a Bachelor of Science in Psychology from Arizona State University in May of 2008. She enrolled in the graduate program for social p sychology at the Univ ersity of Florida in August of 2008 Her based approach. She is p ursuing her doctoral degree in social p sychology at the University of Florida.