The effects of feeling threatened on attitudes toward immigrants

ARTICLE IN PRESS International Journal of Intercultural Relations 29 (2005) 1–19 www.elsevier.com/locate/ijintrel The effects of feeling threatened on attitudes toward immigrants Walter G. Stephana,, C. Lausanne Renfrob, Victoria M. Essesc, Cookie White Stephand, Tim Martine a Department of Psychology, New Mexico State University, Las Cruces, New Mexico Department of Psychology, East Tennessee State University, Johnson City, Tennessee, USA c Department of Psychology, University of Western Ontario, London, Ont., Canada d Department of Sociology, New Mexico State University, Las Cruces, New Mexico e Department of Psychology, University of New Mexico, Albuquerque, New Mexico b Abstract Three studies tested the integrated threat theory by examining the causal role that threats play in attitudes toward immigrants. In Study I, students were presented with information about an immigrant group indicating that it posed realistic threats, symbolic threats, both types of threat or no threats to the ingroup. Attitudes toward the immigrant group were most negative when it posed both realistic and symbolic threats to the ingroup. In Study II, information was presented indicating that an immigrant group possessed negative traits, positive traits, or a combination of positive and negative traits. The results indicated that the negative stereotypes led to significantly more negative attitudes toward the immigrant group than the other types of stereotypes. In the third study, group descriptions leading to high levels of intergroup anxiety led to negative attitudes toward foreign exchange students. Empathizing with the foreign exchange students reduced these negative attitudes. The implications of the results of these studies for theory and practice are discussed. r 2005 Elsevier Ltd. All rights reserved. Keywords: Attitudes; Immigrants; Threats; Empathy; Prejudice Corresponding author. 2097 Aliali Pl., Honolulu, HI 96821, USA. E-mail address: [email protected] (W.G. Stephan). 0147-1767/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijintrel.2005.04.011 ARTICLE IN PRESS 2 W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 1. Introduction Immigration has been an important social issue throughout history, and remains so today. In the modern era, immigration has occurred among nation states as a result of the changing distribution of employment opportunities, population imbalances, natural disasters, and the actions of the nation states themselves (Faist, 2000; Tilly, 1978; United Nations Population Fund, 1993). Immigration often benefits both the immigrant group and the host nation, yet it is almost invariably accompanied by problems. One common problem is prejudice on the part of citizens of the host country, leading to hostility and discrimination toward immigrant groups (Stephan, Ybarra, Martinez, Scharzwald, & Tur-Kaspa, 1998). Such discrimination has negative economic, political, and social effects on both the host country and the immigrant group, which is one reason why a number of researchers have attempted to understand the causes of prejudice toward immigrant groups. Over the course of the last several decades, many theorists and researchers have suggested that fear and perceptions of threat play an important role in prejudice toward outgroups in general and immigrants in particular. One domain of threats, realistic threats, concern threats to the political and economic power of the ingroup, as well as threats to the well-being of the ingroup or its members (Ashmore & Del Boca, 1976; Bobo, 1988; Coser, 1956; LeVine & Campbell, 1972; Sherif, 1966). The basic argument is that ‘‘Real threat causes hostility to the source of the threat’’ (LeVine & Campbell, 1972, p. 30). For example, Ashmore and Del Boca (1976) found that perceived realistic threats to Euro-Americans by African-Americans were highly correlated with Euro-Americans’ evaluations of African-Americans. A second domain of threats, symbolic threats, concerns group differences in values, beliefs, morals, and attitudes. In short, symbolic threats are threats to the worldview of the ingroup. Symbolic threats form the basis for the theories of symbolic and modern racism (McConahay, 1986; Sears, 1988), ambivalence-amplification theory (Katz, Wachenhut, & Hass, 1986), and symbolic beliefs (Esses, Haddock & Zanna, 1993). In one set of studies, Esses et al. (1993) found that symbolic beliefs were related to Canadians’ prejudice toward a variety of different outgroups, including Pakistanis in Canada. Recently, an attempt has been made to combine these types of threats, as well as other related theories and research, into a more comprehensive threat model of prejudice (Stephan & Stephan, 2000). Four threats are included in the model: realistic threats, symbolic threats, intergroup anxiety, and negative stereotypes. In this model the first type of threat, realistic threat, differs from the idea of threat embodied in realistic group conflict theory in two ways. It is broader, encompassing any threat to the welfare of the group or its members, and the focus is on subjectively perceived conflict between groups. The second type of threat consists of symbolic threats. Previous theories, such as symbolic and modern racism, argue that perceiving that one’s values are threatened by an outgroup is a form of prejudice, whereas in the threat model it is proposed that believing that one’s values are threatened is a cause of prejudice (McConahay, 1986; Kinder & Sears, 1981; Sears, 1988). The third type of threat concerns negative stereotypes. Negative outgroup stereotypes can create perceptions of threat among ingroup members when they serve as a basis for ARTICLE IN PRESS W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 3 negative expectations concerning the behavior of members of the stereotyped group (Hamilton, Sherman, & Ruvolo, 1990). To the extent that the expectations are negative, conflict-laden or unpleasant interactions are likely to be anticipated. For instance, when outgroup members are perceived to be aggressive, untrustworthy, or unintelligent, ingroup members may feel threatened by the prospect of interacting with them. The fourth type of threat concerns intergroup anxiety. Several theorists have argued that people feel personally threatened in intergroup interactions because they are concerned about negative outcomes for the self, such as being embarrassed, rejected, or ridiculed (Gaertner & Dovidio, 1986; Gudykunst, 1995; Stephan & Stephan, 1985). Bringing together these threats makes it possible to employ a more comprehensive approach to the specific problems existing between any given set of groups and provides clear testable hypotheses about the relationships among these variables. To date, this model has been used to examine attitudes toward such diverse groups as immigrants (Stephan et al., 1998, Stephan, Ybarra, & Bachman, 1999), Blacks and Whites in the US (Stephan et al., 2002), women’s attitudes toward men (Stephan, Stephan, Demitrakis, Yamada, & Clason, 2000), and people with AIDS and cancer (Berrenberg, Finlay, Stephan, & Stephan, 2003). Much of the evidence concerning the relationship between threats and prejudice has been correlational (for a review see Esses, Jackson, & Armstrong, 1998). By their very nature, such correlational studies cannot provide definitive evidence concerning the role of threats as causes of prejudice toward outgroups. Fortunately, several studies have used experimental manipulations of threat to assess the causal role of threats as determinants of prejudice. For instance, Maio, Esses, and Bell (1994) presented people with positive or negative media descriptions of a fictitious immigrant group’s values and traits, along with information on how other people had reacted to this group. The participants’ responses depended on the personal relevance of the immigration to the participants (i.e., whether the immigrants would be moving to the perceiver’s province in Canada or another province far away). The negative descriptions combined with high personal relevance led to the most negative attitudes. In a subsequent set of studies, the economic threats posed by immigrants to Canadians citizens were manipulated (Esses et al., 1998; Esses, Dovidio, Jackson, & Armstrong, 2001). In these studies, high realistic threats led to more negative attitudes toward immigrants. Another study demonstrating the causal role of threats was conducted by Branscombe and Wann (1994). They found that threats to ingroup identity led to derogation of the threatening outgroup for people who strongly identified with their ingroup, although the outgroup in this case was not an immigrant group. The present set of studies was designed to establish the causal roles of the four types of threat included in the integrated threat theory: realistic threats, symbolic threats, negative stereotypes, and intergroup anxiety. 2. Study I In the first study, the effects of realistic and symbolic threats on prejudice were examined. It was anticipated that people who were presented with information that an outgroup posed either a realistic or symbolic threat to the ingroup would respond ARTICLE IN PRESS 4 W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 with negative attitudes. The effects of these types of threat on attitudes were examined by asking participants to read accounts of imminent immigration to the United States involving members of the Tutsi tribe of Rwanda. The Tutsi tribe was chosen to avoid contamination of the results by participants having had previous contact with the immigrant group. In order to create high levels of personal relevance, the accounts indicated that the Tutsis would be immigrating to the state in which the participants resided (Maio et al., 1994). 2.1. Method 2.1.1. Participants Eighty-eight introductory psychology students at New Mexico State University participated in this study in partial fulfillment of a course requirement. Forty-two of the participants were female and 44 were male. Two participants did not provide gender information. There were 40 Caucasians, 35 Hispanics, and 13 participants who listed themselves as having other ethnic backgrounds or did not provide information on ethnicity. Of the Caucasian participants, 100% were American born, while 86% of the Hispanics were American born. All of the fathers of the Caucasians were also American born, as were 72% of the Hispanic participants’ fathers. 2.1.2. Procedure After participants arrived at the laboratory, the experiment was introduced as a study of perceptions of the mass media. Participants were told that they would be asked to read a reprinted newspaper article and answer several questions about their opinion of the article and the issues it covered. Participants were led to believe that several articles were being investigated and, on that particular day, they would read an article about immigration. They then read the article and completed the questionnaires in individual cubicles. After reading the article and completing the questionnaire, each participant was given a written debriefing. Finally, participants were allowed to ask questions and thanked for their participation. 2.1.3. Design and materials The design was a 2 (Symbolic Threat, Present vs. Absent)
2 (Realistic Threat, Present vs. Absent) factorial. In order to manipulate perceived realistic and symbolic threats, four versions of a fictional news magazine article were created describing a proposed plan to allow Rwandans from the Tutsi minority to immigrate to New Mexico to escape the ravages of the Rwandan civil war. All versions of the article began with a description of the background of the civil war in Rwanda. The articles reported that many people were killed during the war and that the war created many refugees, a substantial number of whom would be immigrating to New Mexico. In one version of the article, the Tutsi were described as posing realistic threats. A portion of this article read: The Rwandans who will immigrate will need a wide range of jobs in the local economy soon after their arrival. Additionally, the federal government will not be ARTICLE IN PRESS W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 5 able to cover the expense of the immigration. This may increase the tax burden on states to cover the costs of moving the Tutsi. In addition, the Tutsi will require immediate financial support and health care. Despite the efforts of the U.N. in Rwanda, many of the Tutsi have been exposed to communicable diseases, including AIDS, Hepatitis B, and Ebola. The Tutsi were not prone to violence before the civil war, but some experts have warned that life during this virtually lawless period has increased the risk of violent criminal behavior. In a second version of this article the Tutsi were presented as posing symbolic threats. A portion of this article read: According to sources in the U.N., the values and beliefs of the Tutsi are rather different from those of most Americans . . . [A] lack of experience with democracy has led to a distrust of democratic principles . . . Very few Tutsi have had an opportunity to own a business or work in an industrialized setting. Therefore they do not share the work ethic of most Americans. . . [T]hey worship and celebrate a variety of tribal gods, in sharp contrast to the mainstream religions of the U.S. Traditional culture in Rwanda has bred a mistrust of science and technology. The third version of the article consisted of a combination of all of the realistic and symbolic threats, while the fourth version of the article consisted only of the background information on the Rwandan civil war and the immigration of the Tutsi. This last version of the article was used as a control condition. The students were asked to read one version of this article as part of a study of the fairness with which the national media dealt with stories of regional interest. After reading one version of the article, all students completed a questionnaire that included items specifically related to the cover story and a set of attitude measures. The first five questions addressed media fairness, and were used to strengthen the cover story. Next, the students were asked a series of questions regarding their evaluations of the Tutsi. The attitude scale was a condensed version of the attitude scale developed for use in earlier studies of attitudes toward immigrants (Stephan et al., 1999; Stephan et al., 1998). The questions asked the students to indicate the degree to which they had the following attitudes toward the Tutsi: Dislike, acceptance, friendliness, resentment, respect, and approval. The response format consisted of a 10-point scale anchored by ‘‘none at all’’ and ‘‘extreme.’’ The responses were reverse scored where necessary and averaged to form an index in which high scores indicated negative attitudes. The Cronbach Alpha for this measure was .85. 2.2. Results A 2 (Symbolic Threat, Present vs. Absent)
2 (Realistic Threat, Present vs. Absent) ANOVA was performed on the measure of attitudes toward Tutsi immigrants. It revealed a marginally significant main effect for Symbolic threat, F ð1; 84Þ ¼ 3:03, po:10 and a significant interaction between symbolic and realistic threats F ð1; 84Þ ¼ 5:94, po:02 (Table 1). Neither the symbolic threat nor the realistic ARTICLE IN PRESS 6 W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 Table 1 Means (and standard deviations) for attitudes toward tutsi immigrants. Study I Symbolic threat Absent Present Realistic threat Absent Present 3.64a (1.18) 3.41a (1.50) 3.19a (1.82) 4.60b (1.76) Note. High numbers indicate more negative attitudes toward Tutsi immigrants. Means with different superscripts are significantly different from one another. threat by themselves caused more negative attitudes than the control condition, but when realistic threats and symbolic threats were combined they did lead to more negative attitudes. Follow-up contrasts indicated that the difference between the condition in which the two threats were combined and each of the other three conditions was significant (Realistic Present–Symbolic Present vs. Realistic Present–Symbolic Absent, tð42Þ ¼ 2:15, po:05, Realistic Present–Symbolic Present vs. Realistic Absent–Symbolic Present, tð43Þ ¼ 2:42, po:05, and Realistic Present– Symbolic Present vs. Realistic Absent–Symbolic Absent, tð43Þ ¼ 2:62, po:05). A follow-up 2
2
2 analysis was conducted to determine whether or not the pattern of responses for Hispanic participants was different from those of Caucasian participants. There were no interactions involving ethnicity, nor was the main effect of ethnicity significant. 2.3. Discussion of Study I The results of Study I indicated that perceptions of threat can play a causal role in the creation of negative attitudes toward immigrants. Although the results indicated that the symbolic threats produced a marginally significant increase in prejudice toward immigrants, this effect was subsumed by the interaction effect, which indicated that it was a combination of realistic and symbolic threats that led to the most negative attitudes toward the immigrant group. Neither symbolic threats nor realistic threats presented in isolation had a significant negative effect on attitudes. It may be that when only one type of threat was presented the participants did not feel threatened enough for this to affect their attitudes. However, when the students perceived that the immigrants posed a threat to both their worldviews and their economic, political, and physical well-being, they responded with negative attitudes. 3. Study II In the second study we examined the role of another threat, negative stereotypes, as a cause of prejudice. A substantial number of studies has shown that negative ARTICLE IN PRESS W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 7 stereotypes are associated with prejudice toward outgroups (Eagly & Mladinic, 1989; Esses et al., 1993; Stangor, Sullivan, & Ford, 1991; Stephan et al., 1993, 1998, 1999; Stephan, Ageyev, Coates-Shrider, Stephan, & Abalakina, 1994). The assumption in these studies has been that negative stereotypes cause prejudice, although some researchers have suggested that causality runs in the opposite direction (Boniecki & Brown, 1998). In addition, in the second study we also examined the effects of race because race has not been systematically investigated in previous studies of threat. Given the prejudice toward African-Americans in American society (Jones, 1997), it seemed likely that US participants would express more negative attitudes toward Black than toward White immigrants. To assess this possibility, the immigrants were said to be of either African or European background. 3.1. Method 3.1.1. Participants The participants were 142 students at New Mexico State University. The sample included 87 females, 53 males, and 2 participants who did not provide gender information. There were 56 Caucasians, 57 Hispanics, 28 participants who listed themselves as having other ethnic backgrounds and 12 who did not provide ethnicity information. Of the Caucasian participants, 98% were American born, while 95% of the Hispanic participants were American born. The students completed the study for partial credit in their psychology classes. 3.1.2. Procedure When participants arrived at the laboratory, they were told that the study concerned the objectivity of the national media in reporting on regional events. It was explained that they would be reading one news article among many. The one they were assigned had to do with immigration to New Mexico by a large group (40,000) of immigrants from the war torn island of East Timor. They were requested to read the article and then to respond to questions about the objectivity of the article and its contents. After reading the article and responding to the questions, the participants were thoroughly debriefed. 3.1.3. Design and materials The design is a 2 (Race of Immigrants—Black vs. White)
3 (Stereotype Valence—Positive, Negative, Mixed) factorial. Stereotype valence was manipulated by varying the traits attributed to the immigrant group. The article began with a description of the civil violence in East Timor. It indicated that the conflict had created a large number of refugees, some of whom would be relocating in New Mexico. Similar to the Tutsi tribe in Study 1, the East Timorese were chosen to avoid previous contact influencing attitudes. In one condition, the people of East Timor were said to have migrated originally from Africa and the picture accompanying this story was of Black refugees (from the civil war in Rwanda). In the other condition, the people of East Timor were said to have come ARTICLE IN PRESS 8 W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 from Portugal during colonial times and the picture accompanying this story was of White refugees (from the war in Kosovo). The article further described the reactions of the UN. High Commissioner on Refugees and the reactions of a US general who was a member of the peacekeeping force in East Timor. In the positive stereotype condition, part of this description read: They [the East Timorese] are very open and warm even to people from outside of their own group and they have been very friendly and respectful to those of us who are attempting to help them. In the negative stereotype condition, the parallel description read: They seem to reserve their warmth and openness for members of their own group and have been hostile and disrespectful to those of us who are attempting to help them. In the mixed stereotype conditions, the participants read that the East Timorese had both positive and negative qualities, as follows: Some of them are very open and warm toward us, but others reserve that warmth and openness only for members of their own group. Many are respectful to those of us who are attempting to help, while others are hostile and disrespectful. After responding to several items consistent with the cover story (e.g., In your opinion was the article you read objective in its coverage of the issue?), the participants completed the primary dependent variable, an evaluation of the East Timorese. This measure contained five questions. Following a stem which read, ‘‘My attitude toward most East Timorese is,’’ the participants were asked to indicate the degree to which six evaluative statements could be applied to the East Timorese. These statements were set up in a 10-point format running from ‘‘no (dislike) at all’’ to ‘‘Extreme (dislike).’’ The evaluative dimensions were: dislike, admiration, acceptance, rejection, and sympathy. The items were reverse coded where necessary and summed to form an index in which high scores indicate negative attitudes. The Cronbach alpha for this measure was .83. As a manipulation check, nine questions about the stereotype-related traits of the East Timorese were included in the post-experimental questionnaire. The questions asked the participants to indicate the percentage of East Timorese who possessed the traits mentioned in the articles. The following traits were listed: hard-working, dishonest, warm, hostile, respectful, inflexible, strong, clannish, and insecure. The response format consisted of a 10-point scale ranging from 0% to 100% in 10% increments. The positively phrased items were reverse scored to create an index reflecting the degree to which the East Timorese were negatively stereotyped. The Cronbach alpha for this measure was .76. 3.2. Results The data were analyzed by performing a 2 (Immigrant Race—Black or White)
3 (Stereotype Valence—Negative, Positive, or Mixed) ANOVA. The manipulation ARTICLE IN PRESS W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 9 check indicated a significant effect for the valence of the stereotype F ð2; 130Þ ¼ 15:30, po:001. The Stereotype Valence main effect indicated that the participants correctly perceived that more negative stereotypes had been attributed to the East Timorese in the Negative Stereotype condition than in the other two conditions (see Table 2). Follow-up contrasts indicated that the mean for the Negative Stereotype condition was significantly different from the Positive Stereotype condition, tð81Þ ¼ 5:14, po:001, and from the Mixed Stereotype condition, tð92Þ ¼ 4:34, po:001. However, the Positive Stereotype condition was not rated more favorably than the Mixed Stereotype condition, tð93Þo1, ns. The Race main effect F ð1; 130Þo1, ns. and the interaction of Race and Valence F ð2; 130Þo1, ns. were non-significant for the manipulation check. An ANOVA on the attitude measure indicated a significant effect for Stereotype Valence, F ð2; 135Þ ¼ 14:74, po:0001. Participants in the Negative Stereotype condition had the most negative attitudes toward the East Timorese (see Table 3). Follow-up contrasts indicated that the mean for the Negative Stereotype condition was significantly different from the Positive Stereotypes condition, tð80Þ ¼ 2:51, Table 2 Means (and standard deviations) for the stereotype valence manipulation check. Study II Immigrant race Valence of stereotypes Positive Negative Mixed Black 3.40 (1.21) 4.99 (1.47) 3.91 (1.04) White 3.71 (.80) 4.88 (1.37) 3.77 (1.13) 3.55 (1.02) 4.95 (1.42) 3.83 (1.09) Marginal means Note. High numbers indicate more negative stereotypes. Table 3 Means (and standard deviations) for attitudes toward east timorese immigrants. Study II Immigrant race Black White Marginal means Valence of stereotypes Positive Negative Mixed 3.33 (1.55) 3.70 (1.35) 4.22 (1.64) 4.49 (1.27) 3.78 (1.73) 3.19 (1.30) 3.51a (1.45) 4.33b (1.49) 3.44a (1.51) Note. High numbers indicate more negative attitudes. Means with different superscripts are significantly different from one another. ARTICLE IN PRESS 10 W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 po:05, and the Mixed Stereotype condition, tð92Þ ¼ 2:83, po:01. Again, the Positive Stereotype condition was not rated more favorably than the Mixed Stereotype condition, tð92Þ ¼ :221, po1, ns. The Race main effect F ð1; 135Þ o1, ns. and the interaction of Race and Stereotype Valence F ð2; 135Þo1, ns. were non-significant. As in the first study, when the ethnicity of the participants was included in the analyses, none of the effects involving this factor was significant. 3.3. Discussion of Study II The results of Study II indicate that the creation of a negative stereotype led to negative attitudes toward an immigrant group. There are a number of reasons why negative stereotypes might lead to prejudice. Under some circumstances, simple cognitive consistency could account for the relationship. People who have negative stereotypes of a group may come to hold negative attitudes toward the group in order to maintain evaluative consistency between their stereotypes and attitudes. However, cognitive consistency does not seem to be a compelling explanation of the results of the present study because the negative stereotypes were said to be held by others. There should have been little need to create consistency between someone else’s views of this group and their own attitudes toward the group. A more motivational explanation would suggest that negative stereotypes set up negative expectations for the behavior of outgroup members (Hamilton et al., 1990) and these negative expectations lead people to feel threatened by interactions with outgroup members. Consistent with this line of reasoning, Allen (1996) found that the traits attributed by Blacks and Whites to the other racial group were rated higher on anxiety by objective raters than the traits attributed to the racial ingroup. The anxiety evoked by the negative traits of the outgroup may then lead to negative attitudes toward that group. Thus, the results are consistent with the premise of the integrated threat theory, which suggests that negative stereotypes can function as threats that cause prejudice. The attribution of positive stereotypes to the immigrant group did not have an impact on the ratings of the stereotypes of this group or on attitudes toward this group. It seems likely that these results were obtained because negative trait information about outgroups is more likely to be deeply processed and better retained than positive information (Ybarra, Stephan, & Schaberg, 2000). Race did not play a role in attitudes toward the immigrant group in this study. This finding is important because it suggests that racism may not be as important in determining attitudes toward immigrants as the traits possessed by immigrant groups. This finding echoes previous findings in the literature indicating that in many instances differences in beliefs are more important determinants of prejudice than race (Triandis & Davis, 1965). 4. Study III Anxiety in response to outgroup members emerges at an early age (Feinman, 1980) and appears to have a variety of detrimental effects on intergroup relations. ARTICLE IN PRESS W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 11 People who are anxious about the negative consequences of intergroup interaction may not behave naturally or effectively with outgroup members and they may be prejudiced toward them (Gudykunst, 1995; Stephan & Stephan, 1985). Intergroup anxiety has proven to be one of the strongest and most consistent predictors of negative attitudes in correlational studies (Corenblum & Stephan, 2001; Bizman & Yinon, 2001; Britt, Boniecki, Vescio, Biernat, & Brown, 1996; Greenland & Brown, 1999; Islam & Hewstone, 1993; Stephan et al., 1999, 2000, 2002; Stephan & Stephan, 2001; Stephan & Renfro, 2002). Very few experimental studies have tested the idea that intergroup anxiety causes prejudice. Maio et al. (1994), found that providing information about the emotional reactions of ingroup members and the traits of an immigrant group led host nationals to have more negative attitudes toward the immigrant group if the outgroup had high personal relevance for the ingroup members. In the Maio et al. (1994) study, the information about the emotional reactions of other ingroup members to the outgroup provided information on intergroup anxiety. However, because the emotional reaction information was combined with information on stereotypes, the study does not provide a clear test of effects of intergroup anxiety on attitudes. Thus, the primary goal of the third study was to clarify the relationship between intergroup anxiety and intergroup attitudes. A secondary goal of the third study was to examine a factor that may reduce the negative outgroup attitudes elicited by intergroup anxiety. In several studies, it has been shown that empathizing with outgroup members can reduce prejudice (Batson et al., 1997; Finlay & Stephan, 2000). In these studies, participants were asked to take the role of members of stigmatized outgroups and later their attitudes toward these groups were assessed. In the present study, we wished to determine if empathizing with the members of an immigrant group would decrease the negative attitudes created by intergroup anxiety. 4.1. Method 4.1.1. Participants Eighty-three introductory psychology students participated in this study in partial fulfillment of a course requirement. Sixty-two of the participants were female and 21 were male. There were 35 Caucasians, 33 Hispanics, and 15 participants who listed themselves as having other ethnic backgrounds or did not provide information on ethnicity. All of the Caucasians were American born as were 91% of the Hispanics. 4.1.2. Procedure When the participants arrived for the study, they were told that the University was interested in obtaining information on students’ reactions to a large group of foreign exchange students that would be arriving on campus the following semester. The participants were asked to read a report from an ad hoc administrative committee that had assessed the impact of the arrival of these students. The report contained a background description of the civil war in East Timor and its causes and ARTICLE IN PRESS 12 W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 consequences. It was explained that university students there had supported one faction in the struggle and that the other side had caused great damage to the local university and had threatened, as well as beaten, students and professors. The report indicated that the university in East Timor had been shut down as a consequence and many East Timorese college students would soon be attending school in the US. The report also contained information on the reactions of students and faculty at UCLA where the first contingent of East Timorese was said to have been sent the prior semester. In addition, there were ‘‘interviews’’ with some of these East Timorese students on their experiences during the conflict and their adjustment to life on the UCLA campus. The information on the reactions of students and faculty at UCLA to the East Timorese students was used to manipulate intergroup anxiety. In the Low Intergroup Anxiety condition, the participants read that the students and faculty at UCLA said that the East Timorese students were, ‘‘really easy to be around’’ and ‘‘very predictable.’’ Other comments included, ‘‘I usually understand what they are thinking or feeling,’’ ‘‘I don’t feel at all threatened,’’ I never worry that they won’t like me,’’ ‘‘I never feel self-conscious when I’m with them,’’ and ‘‘I never have to worry about what I should say or do.’’ In the High Intergroup Anxiety condition, the student and faculty reactions contained the following statements, ‘‘they make me nervous,’’ ‘‘you never know what to expect,’’ ‘‘I just don’t know how to interact with them,’’ ‘‘[I feel] ill at ease with them,’’ ‘‘I don’t understand them at all,’’ ‘‘I worry about what I should do or say,’’ ‘‘I always feel sort of threatened,’’ ‘‘I’m always worried they won’t like me,’’ ‘‘I can’t figure out what they are thinking or feeling,’’ and ‘‘I feel self-conscious whenever I’m with them.’’ The report also included ‘‘interviews’’ with East Timorese students. For example, in commenting on their experiences in East Timor, the report quoted students who said, ‘‘It was getting to the point where people were afraid to be around us just because we were associated with the university (in East Timor).’’ ‘‘I’m grateful to the United States for letting me continue my education here. It was getting too dangerous for anyone associated with the university to be in East Timor.’’ ‘‘I was in the science lab when the fires started. We all got out safely, but all we could do was watch as our research burned.’’ In commenting about their transition to UCLA, the students were quoted as saying, ‘‘I miss my family,’’ ‘‘it has been hard for me to adjust to life in America,’’ ‘‘I worry about my family,’’ ‘‘I have a hard time understanding my teachers,’’ ‘‘It has been difficult for me to make friends,’’ and ‘‘I don’t understand what the [American] students are saying.’’ In order to manipulate the target of empathy, the participants were given two different sets of instructions before reading the report. The instructions read, ‘‘As you read this report put yourself in the place of the East Timor refugees (or in the other condition, ‘‘UCLA students’’). Consider how they feel and what they are thinking about their situation. This should help you understand how these students will respond to New Mexico State University (or in the other condition, ‘‘how NMSU students will react’’).’’ In the condition in which the participants were asked to empathize with the East Timorese refugees, they were expected to pay special attention to comments concerning the difficulties the East Timorese students ARTICLE IN PRESS W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 13 had experienced. In the condition in which the participants were asked to empathize with the UCLA students, they were expected to pay special attention to the comments of the UCLA students concerning their interactions with East Timorese students. 4.1.3. Measures After the participants completed reading the report, they were asked to complete a questionnaire about the East Timorese students. The questionnaire contained measures of attitudes toward East Timorese students, intergroup anxiety, behavioral intentions, and demographic information. When the students had completed the questionnaire, they were debriefed. The attitude measure was adapted from the measures used in previous studies in this series and contained 8 items. For 6 of the items, the 10-point response format ran from ‘‘No ______ at all,’’ to ‘‘Extreme ______’’. These six items were: admiration, dislike, acceptance, sympathy, rejection, and warmth.’’ The remaining two items employed a 10-point format running from ‘‘Extremely favorable,’’ to ‘‘Extremely unfavorable.’’ These two items asked the participants to indicate their attitudes toward ‘‘immigration of East Timorese students to New Mexico State University’’ and ‘‘East Timorese refugee students in general.’’ The items were reverse scored where necessary and averaged to create a scale in which high scores indicate negative attitudes. The Cronbach alpha for this measure was .88. The intergroup anxiety measure was also adapted from the previous studies. It contained 14 items in a 10-point semantic differential format. Half of the items referred to feelings that were mentioned in the experimental protocols, including ill at ease, threatened, anxious, worried, nervous, self-conscious, and disapproval. The other half of the items referred to feelings that were not included in the experimental protocols, including unsafe, awkward, uncomfortable, unsure, confident, trusting, and unfriendly. High scores on these measures indicate greater intergroup anxiety. The Cronbach alpha for the items that were included in the protocols was .84, while the Cronbach alpha for the items that were not included was .85. The behavioral intention measure included 15 items in a 10-point response format. The participants were asked how likely it would be that they would engage in these behaviors with the East Timorese students when they arrived on their campus. The items included, ‘‘share my class notes with an East Timorese student,’’ ‘‘study with an East Timorese student,’’ ‘‘play on the same intramural team,’’ and ‘‘eat a meal with an East Timorese student.’’ High scores on this measure indicate greater willingness to interact with East Timorese students. The Cronbach alpha for this measure was .93. 4.2. Results A 2 (High vs. Low Intergroup Anxiety)
2 (Empathy with East Timorese vs. UCLA students) ANOVA was employed to analyze the data. ARTICLE IN PRESS 14 W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 4.2.1. Manipulation check The analyses for both the intergroup anxiety items that were included in the manipulation and those that were not included were nearly identical. In both analyses, only the main effect for intergroup anxiety was significant F ð1; 78Þ ¼ 6:08, po.05 for the included items, and, F ð1; 77Þ ¼ 8:75, po.005 for the items that were not included. In both cases, intergroup anxiety was greater in the High than the Low Intergroup Anxiety condition. 4.2.2. Attitudes The principal analysis was the attitude measure (Table 4). The intergroup anxiety main effect was significant, F ð1; 78Þ ¼ 12:09, po.001. This main effect indicated that East Timorese students were more favorably evaluated by participants in the Low Anxiety condition than in the High Intergroup Anxiety condition. There was also a significant interaction between Intergroup Anxiety and Empathy, F ð1; 78Þ ¼ 4:16, po:05 (Table 4). This interaction indicated that in the High Intergroup Anxiety condition, the East Timorese students were more favorably evaluated when the participants empathized with them than when they were empathizing with the UCLA students. In contrast, in the Low Intergroup Anxiety condition, the East Timorese students were evaluated more favorably when the participants empathized with the UCLA students than when they were empathizing with the East Timorese students. As a consequence, the significant difference between the High and Low Intergroup Anxiety conditions that existed when the participants empathized with the UCLA students, tð38Þ ¼ 4:67, po:001, disappeared when the students empathized with the East Timorese students, tð40Þ ¼ :90, p4:30. Again, when the ethnicity of the participants was included in the analyses, none of the effects involving this factor was significant. 4.2.3. Behavioral Intentions The results for the behavioral intentions measure paralleled those for the anxiety measures. Again, there was only a main effect for Anxiety F ð1; 76Þ ¼ 4:59, po:05, indicating that the participants were more willing to interact with East Timorese students in the Low than in the High Intergroup Anxiety condition. Table 4 Means (and standard deviations) for attitudes toward east timorese students. Study III Empathy East Timorese Students UCLA Students Intergroup anxiety Low High 1.87 (1.49) 1.29a (.97) 2.28 (1.47) 2.88b (1.17) Note. High numbers indicate more negative attitudes East Timorese students. Means with different superscripts are significantly different from one another. ARTICLE IN PRESS W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 15 4.3. Discussion of Study III In support of the integrated threat theory, this study found that high levels of intergroup anxiety led to negative evaluations of foreign students and less willingness to interact with them. In the case of attitudes, it was also found that empathizing with the foreign students reduced the negative effects on attitudes of high levels of manipulated intergroup anxiety. The most favorable evaluations occurred when the participants were empathizing with other American students and the foreign students were presented as eliciting low levels of intergroup anxiety. In this condition, the participants would have been focused primarily on how easy other American students found interacting with the East Timorese students to be. The beneficial effects of empathy on attitudes toward the foreign students under conditions of high intergroup anxiety is consistent with other studies finding that empathy for outgroup members reduces prejudice (Batson et al., 1997; Stephan & Finlay, 2000). The findings for behavioral intentions suggest that empathy alone may not be enough to change behaviors, although perhaps empathy over a more extended time period would be effective in modifying behavioral intentions. 5. General discussion The results of these three studies indicate that perceiving threats posed by immigrant groups can lead to prejudice. In the first study, a combination of realistic and symbolic threats led to negative attitudes. In the second study, negative stereotypes led to negative attitudes, while in the third study high levels of intergroup anxiety led to negative attitudes. Together, these studies suggest that one reason that people are hostile toward immigrant groups is that they feel threatened by them. This general conclusion is consistent with a variety of theories that argue that threats cause prejudice. These theories include realistic group conflict theory (Coser, 1956; LeVine & Campbell, 1972; Sherif, 1966), symbolic racism (Sears, 1988), modern racism (McConahay, 1986), aversive racism (Gaertner & Dovidio, 1986), and anxiety/uncertainty management theory (Gudykunst, 1995). The results are also consistent with previous experimental studies indicating that economic threats, value differences, and negative traits cause prejudice (Maio et al., 1994; Esses et al., 1998, 2001). However, none of these previous studies isolated the separate threats as was done in the present set of studies. In the world outside the laboratory, immigrant groups would seem particularly likely to evoke the types of threat outlined in the integrated threat theory. They always need jobs and they may require additional resources from the societies to which they are immigrating. Almost by definition, they are culturally different from the societies to which they are immigrating. Due to ethnocentrism, most stereotypes of national outgroups are negative, and perhaps because immigrants are commonly from less well-developed countries than the ones to which they are immigrating, the stereotypes of immigrants tend to be particularly negative. In addition, interacting with immigrants is often difficult for citizens of the host culture due to differences in cultural values and language, and this adds intergroup anxiety as an element of threat. ARTICLE IN PRESS 16 W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 To the extent that the results of these and related studies of threat apply to outgroups in general, they have important implications for educators and intergroup relations trainers who present information about outgroups during their programs. Clearly, they should take care not to present the outgroups in a manner that threatens ingroup members. Intergroup relations programs that stress group differences, such as multicultural education, diversity training, and intercultural relations training (Banks, 1988, 1997; Ellis & Sonnenfield, 1994; Hollister, Day, & Jesaitis, 1993; Landis & Bhagat, 1996), run the risk of creating negative attitudes in their training. This may be particularly true when one of the avowed aims of the program is to strengthen ingroup identity, as is often the case with multicultural education. One solution to this dilemma is to be very careful to present information on group differences in a non-evaluative manner and to stress the benefits of group differences (Stephan & Stephan, 2001). Information on group differences should be accompanied by information on group similarities to dilute the impact of the dissimilarity information. As suggested by Allport (1954), emphasizing the common interests and common humanity shared by all groups may also prove beneficial. One of the most powerful antidotes to perceiving threat is intergroup contact, under the conditions specified by the contact hypothesis (Amir, 1976; Pettigrew & Tropp, 2000; Stephan & Stephan, 1996). It is also important to present information on within-group diversity in an attempt to undercut group stereotypes by creating more differentiated perceptions of outgroups. Perceptions of group differences are often exaggerated, but it may be possible to reduce these perceived differences by presenting more realistic information about outgroups. The third study in this series also suggests that empathizing with members of the outgroup may have a beneficial effect on attitudes (Stephan & Finlay, 1999). This is a valuable finding because so many intergroup relations programs explicitly or implicitly employ empathy as a technique of improving intergroup relations (Stephan & Stephan, 2001). In summary, these three studies provide experimental support for the integrated threat theory by indicating that threats posed by immigrant groups can lead to prejudice. Future research on this topic could elucidate other effects of threats, in addition to their effects on attitudes, and examine what types of people are most likely to perceive threats (Stephan & Renfro, 2002). Practitioners may benefit in their future work from giving greater consideration to the important role that perceptions of threat play in intergroup relations. Acknowledgement The authors wish to thank W. Larry Gregory for his assistance in analyzing the data. References Allen, B. P. (1996). African Americans’ and European Americans’ mutual attributions: Adjective Generation Technique stereotyping. Journal of Applied Social Psychology, 26, 884–912. Allport, G. W. (1954). The nature of prejudice. Reading, MA: Addison-Wesley. ARTICLE IN PRESS W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 17 Amir, Y. (1976). The role of intergroup contact in change of prejudice and race relations. In P. Katz, & D. A. Taylor (Eds.), Towards the elimination of racism (pp. 245–308). New York: Pergamon. Ashmore, R. D., & Del Boca, F. K. (1976). Psychological approaches to understanding intergroup conflict. In P. A. Katz (Ed.), Towards the elimination of racism (pp. 73–124). New York: Pergamon. Banks, J. A. (1988). Multicultural education (2nd ed.). Boston: Allyn & Bacon. Banks, J. A. (1997). Educating citizens in a multicultural society. New York: Teachers College Press. Batson, C. D., Polycarpou, M. P., Harmon-Jones, E., Imhoff, H. J., Mitchener, E. C., Bednar, L. L., Klein, T. R., & Highberger, L. (1997). Empathy and attitudes: Can feeling for a member of a stigmatized group improve feelings toward the group? Journal of Personality and Social Psychology, 72, 105–118. Berrenberg, J. L., Finlay, K. A., Stephan, W. G., & Stephan, C. W. (2003). Prejudice toward people with Cancer or AIDS: Applying the Integrated Threat Model. Journal of Biobehavioral Research, 7, 75–86. Bizman, A., & Yinon, Y. (2001). Intergroup and interpersonal threats as determinants of prejudice: The moderating role of in-group identification. Basic and Applied Social Psychology, 23(3), 191–196. Bobo, L. (1988). Group conflict, prejudice, and the paradox of contemporary racial attitudes. In P. A. Katz, & D. A. Taylor (Eds.), Eliminating racism: Profiles in controversy (pp. 85–116). New York: Plenum. Boniecki, K. A., & Brown, L. M. (1998). The influence of prejudice on stereotype formation: The justification hypothesis revisited. Paper presented at the American Psychological Association meeting, Washington, DC, May 23. Branscombe, N. R., & Wann, D. L. (1994). Collective self-esteem consequences of outgroup derogation when a valued social identity is on trial. European Journal of Social Psychology, 24, 641–657. Britt, T. W., Boniecki, K. A., Vescio, T. K., Biernat, M., & Brown, L. M. (1996). Intergroup anxiety: A person X situation approach. Personality and Social Psychology Bulletin, 22(11), 1177–1188. Corenblum, B., & Stephan, W. G. (2001). White fears and Native apprehensions: An integrated threat theory approach to intergroup attitudes. Canadian Journal of Behavioral Science, 33(4), 251–268. Coser, L. (1956). The functions of social conflict. New York: Free Press. Eagly, A. H., & Mladinic, A. (1989). Gender stereotypes and attitudes toward women and men. Personality and Social Psychology Bulletin, 15, 543–558. Ellis, C., & Sonnenfield, J. A. (1994). Diverse approaches to managing diversity. Human Resource Management, 33, 79–109. Esses, V. M., Dovidio, J. F., Jackson, L. M., & Armstrong, T. L. (2001). The immigration dilemma: The role of perceived competition, ethnic prejudice, and national identity. Journal of Social Issues, 57(3), 389–412. Esses, V. M., Haddock, G., & Zanna, M. P. (1993). Values, stereotypes, and emotions as determinants of intergroup attitudes. In D. M. Mackie, & D. L. Hamilton (Eds.), Affect, Cognition and Stereotyping (pp. 137–166). San Diego: Academic Press. Esses, V. M., Jackson, L. M., & Armstrong, T. L. (1998). Intergroup competition and attitudes toward immigrants and immigration: An instrumental model. Journal of Social Issues, 54(3), 699–724. Faist, T. (2000). The volume and dynamics of international migration and transnational social spaces. New York: Oxford University Press. Feinman, S. (1980). Infant response to race, size, proximity, and movement of strangers. Infant Behavior and Development, 3, 187–204. Finlay, K. A., & Stephan, W. G. (2000). Reducing prejudice: The effects of empathy on intergroup attitudes. Journal of Applied Social Psychology, 30(8), 1720–1737. Gaertner, S. L., & Dovidio, J. F. (1986). The aversive form of racism. In J. F. Dovidio, & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 61–90). Orlando: Academic Press. Greenland, K., & Brown, R. (1999). Contact and intergroup anxiety between British and Japanese Nationals. European Journal of Social Psychology, 29, 503–521. Gudykunst, W. B. (1995). Anxiety/uncertainty management (AUM) theory: Development and current status. In R. L. Wiseman (Ed.), Intercultural communication theory (pp. 8–51). Thousand Oaks, CA: Sage. ARTICLE IN PRESS 18 W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 Hamilton, D. L., Sherman, S. J., & Ruvolo, C. M. (1990). Stereotype-based expectancies: Effects on information processing and social behavior. Journal of Social Issues, 46, 35–60. Hollister, L., Day, N. E., & Jesaitis, P. T. (1993). Diversity programs: Key to competitiveness or just another fad. Organization Development Journal, 11, 49–59. Islam, M. R., & Hewstone, M. (1993). Dimensions of contact as predictors of intergroup anxiety, perceived out-group variability and out-group attitude: An integrative model. Personality and Social Psychology Bulletin, 19(6), 700–710. Jones, J. M. (1997). Prejudice and racism (2nd ed.). New York: McGraw-Hill. Katz, I., Wachenhut, J., & Hass, R. G. (1986). Racial ambivalence, value duality, and behavior. In J. F. Dovidio, & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 35–60). New York: Academic Press. Kinder, D. R., & Sears, D. O. (1981). Prejudice and politics: Symbolic racism vs. racial threats to the good life. Journal of Personality and Social Psychology, 40, 414–431. Landis, D., & Bhagat, R. S. (Eds.). (1996). Handbook of Intercultural Training, (2nd ed.). Thousand Oaks, CA: Sage. LeVine, R. A., & Campbell, D. T. (1972). Ethnocentrism: Theories of conflict, ethnic attitudes, and group behavior. New York: Wiley. Maio, G. R., Esses, V. M., & Bell, D. W. (1994). The formation of attitudes toward new immigrant groups. Journal of Applied Social Psychology, 24, 1762–1776. McConahay, J. G. (1986). Modern racism, ambivalence, and the modern racism scale. In J. F. Dovidio, & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 91–125). Orlando, FL: Academic Press. Pettigrew, T. F., & Tropp, L. R. (2000). Does intergroup contact reduce prejudice: Recent meta-analytic findings. In S. Oskamp (Ed.), Reducing prejudice and discrimination (pp. 93–114). Mahwah, NJ: Erlbaum. Sears, D. O. (1988). Symbolic racism. In P. A. Katz, & D. A. Taylor (Eds.), Eliminating racism (pp. 53–84). New York: Plenum. Sherif, M. (1966). Group conflict and cooperation. London: Routledge & Kegan Paul. Stangor, C., Sullivan, L. A., & Ford, T. E. (1991). Affective and cognitive determinants of prejudice. Social Cognition, 9, 359–380. Stephan, W. G., Ageyev, V. S., Coates-Shrider, L., Stephan, C. W., & Abalakina, M. (1994). On the relationship between stereotypes and prejudice: An international study. Personality and Social Psychology Bulletin, 20, 277–284. Stephan, W. G., Ageyev, V. S., Stephan, C. W., Abalakina, M., Stefanenko, T., & Coates-Shrider, L. (1993). Soviet and American stereotypes: A comparison of methods. Social Psychology Quarterly, 56, 54–64. Stephan, W. G., Boniecki, K. A., Ybarra, O., Bettencourt, A., Ervin, K. S., Jackson, L. A., McNatt, P. S., & Renfro, C. L. (2002). Racial attitudes of Blacks and Whites: An integrated threat theory analysis. Personality and Social Psychology Bulletin, 28(9), 1242–1254. Stephan, W. G., & Finlay, K. (1999). The role of empathy in improving intergroup relations. Journal of Social Issues, 55(4), 729–744. Stephan, W. G., & Renfro, C. L. (2002). The role of threats in intergroup relations. In D. Mackie, & E. R. Smith (Eds.), Beyond prejudice: Differentiated reactions to social groups. Mahwah, NJ: Erlbaum. Stephan, W. G., & Stephan, C. W. (1985). Intergroup anxiety. Journal of Social Issues, 41, 157–175. Stephan, W. G., & Stephan, C. W. (1996). Intergroup relations. Boulder, CO: Westview Press. Stephan, W. G., & Stephan, C. W. (2000). An integrated threat theory of prejudice. In S. Oskamp (Ed.), Claremont symposium on applied social psychology (pp. 23–46). Hillsdale, NJ: Erlbaum. Stephan, W. G., & Stephan, C. W. (2001). Improving intergroup relations. Thousand Oaks, CA: Sage. Stephan, C. W., Stephan, W. G., Demitrakis, K., Yamada, A. M., & Clason, D. (2000). Women’s attitudes toward men: An integrated threat theory analysis. Psychology of Women Quarterly, 24, 63–73. Stephan, W. G., Ybarra, O., & Bachman, G. (1999). Prejudice toward immigrants: An integrated threat theory. Journal of Applied Social Psychology, 29, 2221–2237. ARTICLE IN PRESS W.G. Stephan et al. / International Journal of Intercultural Relations 29 (2005) 1–19 19 Stephan, W. G., Ybarra, O., Martinez, C., Scharzwald, J., & Tur-Kaspa, D. (1998). Prejudice toward immigrants to Spain and Israel: An integrated threat theory analysis. Journal of Cross-Cultural Psychology, 29, 559–576. Tilly, C. (1978). Migration in modern European history. In W. H. McNeill, & R. S. Adams (Eds.), Human migration: Patterns and policies. Bloomington, IN: Indiana University Press. Triandis, H. C., & Davis, E. (1965). Race and belief as determinants of behavioral intentions. Journal of Personality and Social Psychology, 7, 316–328. United Nations Population Fund. (1993). The state of the world population 1993. New York: United Nations. Ybarra, O., Stephan, W. G., & Schaberg, L. (2000). Misanthropic memory for the behavior of group members. Personality and Social Psychology Bulletin, 26, 1515–1525.