Content-Type: text/html

Racial Stereotyping and Mass Mediated Contact


Racial Stereotyping and Mass Mediated Contact














Carol M. Liebler
S.I. Newhouse School of Public Communications
Syracuse, NY  13244-2100
<a href="/cgi-bin/wa?LOGON=A3%3Dind0109B%26L%3DAEJMC%26E%3D8bit%26P%3D6589220%26B%3D--%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D_114152108%253D%253D_%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a>



Richard D. Waters
University of Georgia











Paper presented to the Minorities and Communication Division,
Association for Education in Journalism and Mass Communication,
Washington, DC, August, 2001





Abstract


Racial Stereotyping and Mass Mediated Contact




This study examines in-group bias, and the extremity-complexity and contact hypotheses in relation to media exposure, and crime and success stereotyping. A cross-sectional survey of African, Anglo, Asian and Latinos (n=491) illustrated that not only did stereotyping vary by group,  but that interpersonal and mediated communication are both important factors to consider when conceptualizing and operationalizing contact, as is the type of media contact.


 Racial Stereotyping and Mass Mediated Contact





Although the mass media today reflect a society that is much less tolerant of blatant prejudices, they do not reflect a society in which all intergroup issues have been resolved and in which prejudice and racial inequality are things of the past.
Healey (1998)  p. 111




Introduction

        Discussion of the media's role in race relations dates back at least to the Kerner Commission of 1968.  Since that time, the mass communications literature has become rich in content studies of media stereotyping. Yet the degree to which the mass media improve or hamper understanding across racial and cultural lines remains relatively unexplored. While negative and stereotypical images persist, our media landscape is evolving to reflect an increasingly diverse population: we witness more varied representations in advertising, news and entertainment than ever before. Whether audiences' perceptions reflect this more balanced imagery, or remain consistent with traditional stereotypes and prejudices, warrants further attention.
        The present study examines three related social psychological conceptualizations: in-group bias, and the extremity-complexity and contact hypotheses.  At the heart of this study is analysis of the linkage between media use and stereotyping, or more specifically, how people perceive people like themselves and people from other groups.  Suggested here is that members of different racial/ethnic groups will not only have different media exposure patterns, but that the relationship between media usage and stereotyping will vary by group as well.  This study further expands the research in this area by drawing upon a diverse sample of African American, Anglo American, Asian American and Latino adults.
        Stereotypes have been defined frequently in the mass communications literature.  Definitions typically include such terms as oversimplification and generalization.   But when speaking of media stereotypes, it is important to distinguish among racial and ethnic groups in terms of both audiences and content.  It is noteworthy that while media stereotypes can be found in both entertainment and news content, these stereotypes vary across groups (Gandy, 1998).  Moreover, while any stereotype is limiting, media portrayals have included a variety of both relatively positive and negative depictions. For example, Blacks have often been stereotyped as criminals and entertainers; Asian Americans as both a yellow peril and a model minority; and Native Americans as savages, drunks and noble keepers of the earth (e.g. Wilson & Gutierrez, 1995).
        Such stereotyping manifests in many ways in our everyday lives. This study examines them in relation to in-group bias which suggests that people tend to prefer people like themselves. Related to this discussion is the contact hypothesis which states that the more contact people have with others unlike themselves, the less prejudiced they will be toward members of that group.  The equal status contact hypothesis goes one step further, predicting prejudice will be reduced only when people are of comparable status. These concepts are applied here to the mass media and their audiences, and are discussed below in the literature review.
        It is logical to assume that the manner in which different groups are portrayed in the media will have an impact on how people perceive members of those groups.  In their study of racial perceptions, Domke, McCoy and Torres (1999, p. 575) used priming theory to argue that "media content interacts with individual predispositions to guide information processing and subsequent judgments."  Their study revealed that news coverage influenced which racial cognitions were activated.  The authors suggested that even when specific stereotypes are not present in news coverage, the way the news is framed affects whether individuals apply existing stereotypes.
        Cultivation theorists have long argued that media provide a common world view and set of values, and that these are reflected among heavy media users, most particularly  television viewers.  Research has provided some evidence that media can cultivate perceptions of racial integration (Matabane, 1988); socioeconomic standing (Armstrong, Neuendorf & Brentar, 1992); and of inequality (Gandy & Baron, 1998). Particularly germane to the present research  is a series of studies in the Netherlands by Vergeer and colleagues (2000).  They have found that exposure to fictional and entertainment media, to commercial television, and to newspapers causes people to perceive ethnic minorities as more of a threat.  For the latter, exposure to newspapers explained nearly 12% of the variance, even after controlling for respondents' socio-structural variables (Vergeer, Lubbers & Scheepers, 2000).  In their concluding remarks, the authors suggest that future research investigate the cumulative expos
ure of a variety of mass media, and take into account interpersonal contact; the present study incorporates both factors within the context of the contact hypothesis.
        Accordingly, this study poses four research questions:
        RQ1:    To what extent is amount and nature of media use related to perceptions                  of one's own racial group?

        RQ2:    To what extent is amount and nature of media use related to perceptions of              racial groups other than one's own?

        RQ3:    To what extent do media and interpersonal contact vary with relative                    group status?

        RQ4:    To what extent do perceptions vary with type of media content?

Theory
        In-group Bias, Extremity-Complexity and the Contact Hypothesis
        In 1954, Gordon Allport discussed in-group formation in his well-known book, The Nature of Prejudice.  Defining an in-group as "any cluster of people who can use the term 'we' with the same significance," (p. 37), Allport argued that in-groups are psychologically primary.  While such primacy does not automatically lead to out-group hostility, Allport recognized that for some people rejecting the out-group is a "salient need" (p. 49).
        More recently, Brewer (1999) reviewed forty years of social psychological research on intergroup relations and concluded that Allport had been right:
Many discriminatory perceptions and behaviors are motivated primarily by the desire to promote and maintain positive relationships within the in-group rather than by any direct antagonism toward out-groups...However, the very factors that make in-group attachment and allegiance important to individuals also provide a fertile ground for antagonism and distrust of those outside the in-group boundaries. p. 441-442

        One contemporary example of in-group versus out-group perception is Wong, Lai, Nagasawa and Lin's (1998) analysis of Asian Americans' self perceptions and perceptions by other groups.  They found that among a college student sample, Asian Americans "perceived themselves as more prepared, motivated and more likely to have greater career success than whites" (p. 113).  Interestingly, however, the results held for white, African American, Hispanic and Native American students' perceptions as well; Wong et al. conclude that the study provides support for the model minority hypothesis.
        The extremity-complexity hypothesis offers one explanation for why evaluations of in-group and out-group members often differ.  It suggests that people will have more complex schema for members of their own group, and more extreme evaluations for members of out-groups.   In other words, the more dimensions people use in evaluating others, the less extreme (stereotypical) their evaluation  (Linville & Jones, 1980; Jussim, Coleman & Lerch, 1987).
        With all probability, audience communication patterns play a key role in such evaluation, whether it be in-group attachment or out-group hostility. Work testing the contact hypothesis has emphasized interpersonal contact, but it clearly has implications for mass media exposure (see Tan, Fujioka & Lucht, 1997;  Fujioka, 1999).   In its most basic form, the contact hypothesis, which is derived from Allport's (1954) work, suggests an inverse relationship between contact and prejudice:  As contact among members of different groups increases, prejudice decreases.  However, numerous scholars have argued that such an interpretation is too simplistic; if reducing prejudice were so easy, we would see much less of it in many places around the world.  Rather, the effectiveness of contact is related to a number of situational factors.
        First off, and central to the current study, is the nature of the contact.  According to Forbes, whose 1997 book offers a comprehensive review of the literature in this area, "how to distinguish favorable from unfavorable contact is the central problem of contact theory" (p. 24). His review suggests that casual, impersonal contact may actually lead to increased stereotyping whereas prolonged, personal contact may have the opposite effect. In addition to the nature of the contact being a factor, for it to have a positive impact people must be of equal status. Otherwise, the power imbalance may actually reinforce prejudice. Other criteria for effective contact are normative support for positive intergroup interaction and a sense of cooperative interdependence (Forbes, 1997).
        Tan, Fujioka & Lucht (1997) and  Fujioka (1999) have examined the contact hypothesis within a mass communications context.   The first study tested the extremity-complexity hypothesis, and found that frequency of television viewing did not lead to more positive stereotyping.  The authors conclude that their data offer "some support for vicarious contact (via television) hypotheses based on learning theory, and no support for the vicarious contact based on cognitive extremity-complexity theory" (p.280). However, the study was limited in scope.  Its sample was white college students who were asked their perceptions of Native Americans.  Furthermore, as the authors point out, the study failed to distinguish among content types.
        Two years later, Fujioka (1999) expanded upon this inquiry.  She studied white and Japanese students' perceptions of African Americans, testing whether students with frequent direct contact (white students) would have different perceptions than those with limited direct contact (Japanese students).  She found that evaluation of television programs had significant influence on stereotypes, but that cumulative television exposure did not.  However, perceived positive portrayals had a greater impact on the Japanese students than on the white students.  As Fujioka points out, however, the study had some considerable limitations.  Not only was a student sample used, but it drew upon international students who came from a racially homogeneous setting and whose experience with racial issues is quite different than that of their American counterparts.

Hypotheses
H1:     The greater the status differential, the less the interpersonal and     mass mediated contact.

        This hypothesis draws upon the concept of in-group bias. People are less likely to seek out those who they see as very different from themselves.

H2:     As media exposure increases, so will the inclination to negatively      stereotype groups other than one's own.

        This hypothesis is based on the extremity-complexity hypothesis which predicts that people will have more extreme evaluations for groups other than their own.  It is further supported by contact theory which suggests that impersonal, casual contact (such as media exposure) may actually increase prejudices and the inclination to stereotype.  Moreover, since much of media content draws upon racial stereotypes, it appears all the more likely that heavy users will be primed to stereotype.
        This relationship will be further mediated by two situational factors:  the type of contact and the group. Therefore:

H3:     The inclination to negatively stereotype groups other than one's        own will vary by media content type.

        This hypothesis recognizes that not all media contact is created equal. For example, local TV news frames African Americans as criminals Entman (1990, 1992), while the Internet provides diverse content from a variety of perspectives.  Similarly, televised sporting events provide a look at success for many different groups, the most prominent depending upon the sport.  It  is expected that in-group-out-group bias will function so that such stereotyping will be more evidenced for members of groups other than one's own.

Method
        Research Design
        This study employed a cross-sectional telephone survey in Metropolitan Atlanta in July, 1999.  Ten interviewers, all of whom had previous telephone survey or telemarketing experience, received extensive Kish training prior to data collection. An initial pretest over a two-day period resolved minor problems related to instrumentation and survey administration.  The 49 diversity-related items used in the present study were part of a larger 73-item instrument.
Sample
        A systematic random sample was drawn from the population of the ten Central North Georgia counties that comprise the Atlanta Metro area: Cherokee, Clayton, Cobb, Dekalb, Douglas, Fayette, Fulton, Gwinnett, Henry and Rockdale.  Ninety-two percent of households in the area have at least one telephone line, according to the BellSouth Corporation.  Combined, these counties have a diverse population with approximately 29 percent of the population ethnic minority.
        The sampling frame was a CD-Rom phone directory which was determined to be the most comprehensive list available of residential telephone numbers available for the Atlanta metro area. The CD lists approximately  2.7 million phone numbers.  The sampling interval  of 1,928 was determined based on Lavrakas (1993), whereby:
Estimated Size of Sampling Pool = (FSS)/(HR)(1-REC)(1-LE)
FSS is the field sample size; in the present study, the target was 500 completed surveys.  HR is the percentage of phone numbers expected to be working, which Lavrakas estimates at .60 for large metropolitan areas such as Atlanta. REC is respondent exclusion rate (here based on age restrictions) and estimated at .20.  Nonresponses (LE) was estimated at .25, using the general population refusal rate for metropolitan areas.  Based on the formula, 1,667 phone numbers were selected randomly for inclusion in the study.


Data Collection
        A written protocol was provided for the ten interviewers and all interviewers participated in a training session.  A scripted introduction and Kish selection sheet helped to control for interviewer bias. Data were then collected over a two-week period (July 10, 1999 to July 24,1999), following a two day pre-test.  All  phone calls were made between 9 a.m. and 9 p.m.  The Kish method screened out individuals younger than 16 years old.
Measurement
        The primary dependent variable of this study is stereotypes.  In operationalizing that concept this study drew upon the earlier work of Tan, Fujioka & Lucht (1997) and  Fujioka (1999).  A ten point scale was used for each of seven items that provided bipolar adjectives:  lazy/hardworking, violent/non-violent, intelligent/unintelligent, likely to commit a crime/un likely to commit a crime, educated/uneducated, likely to use drugs/unlikely to use drugs, wealthy/not wealthy.  For example, interviewers read a series of statements such as this one: "Now I am going to read you a list of several adjectives.  Using a 10 point scale, please tell me where you would place Asian Americans if 1= lazy and 10 = hardworking."  The statements were read for each of four groups (Anglo Americans, African Americans, Asian Americans and Latino/Hispanics).  Positive and negative responses varied on the ten point scale to avoid response set.
        Contact, a key independent variable, was operationalized in terms of media exposure.  Respondents were asked how much time they spent each day with each of the following media:  television news, prime time television, televised sports, newspapers, and the Internet.  Respondents were also asked how frequently they saw each of four groups (African Americans, Asian Americans, Latinos and Anglo Americans) in the media (1 = not at all and 5 = very often).
        Contact was further operationalized in terms of interpersonal contact.  Respondents were asked for the same four groups, "On a day-to-day basis, how frequently do you interact with each of the following groups, where 1 = not at all and 5 = very often?"
        Respondents were also asked their race, education level, sex and income level.  Racial/ethnic categories mirrored those then used on the US Census, with the full recognition that Hispanic denotes an ethnic group not a race.
        Group status was based on two socio-economic indicators gleaned from US Census data:  percentage with a college degree or higher and percent living in poverty (Table 1).  These data were used in categorically to determine the status differential between groups. According to these data, Anglo Americans and Asian Americans are of comparable status. The greatest differences according to these data are between African Americans and Asian Americans, and African Americans and Anglo Americans.

Results
        Descriptive Data
        Of the 1,667 phone numbers included in the study, 216 were deleted due to disconnections, fax machines or a non-residential number. A total of 386 questionnaires were completed based on the first call; an additional 175 were completed after a call-back appointment.  A total of 561 respondents participated in the study for a completion rate of 45 percent.
         For the purposes of this analysis, however, respondents were included only if they clearly identified themselves as white, African American, Asian American, or Latino (n = 491).  Of the 491 respondents, 47% were Anglo American (n=231), 21% were African American (104) , 20% were Asian American (n=97) and 12% identified themselves as Hispanic or Latino (n=59).  In this sample, then, minorities were over-represented relative to the population in the Atlanta metro area.   This may be because Fulton and Cherokee counties were over-represented in the sample, both of which have significant minority populations.  Men comprised 51% of the sample, women 49%.
        Of note, income levels for the groups in the sample did not match those of census data.  In this sample, 38% of African Americans had incomes over $45,000, followed by 35% of Asian Americans, 20% of Anglo Americans and 12% of Latinos. This, again, is a function of the oversampling of two counties. Therefore, the differences in status in the sample are different than those for the population at large.
        Media exposure was significantly different among the four racial groups (See Table 2).  Anglos spent more time with newspapers than did members of other groups (x=.71 hours/day);  Asian Americans spent the most time of any group on the Internet (x=2.4 hours/day) and watching TV news (x=.61 hours); African Americans spent the most time with prime time (x=2.8 hours/day); and Latinos spent the most time watching TV sports (x=1.0).  However, regardless of race or ethnicity, respondents spent more time with prime time television (x=2.1 hours/day) than with any other content measured.
        The means and standard deviation for the statements measuring stereotypes are provided in Table 3, and are discussed further below.  Of note, the means indicate that respondents did stereotype.  For example, on a ten point scale (some responses were recoded so that in all instances a one indicates a negative perception and 10 a relatively positive one), the highest score (8.21) was for Asian American wealth, and the lowest score was for African American violence (3.38).
Index Construction
        For African Americans, Asian Americans and Latinos, twelve items were consolidated into two indices, resulting in a total of six dependent measures (two per group).   One index is a success measure and the other is an anti-social/crime measure. Chronbach's alpha coefficients, which ranged from .62 to .76, are provided in Table 4 along with the variables for each index.  For Anglo Americans, satisfactory reliability was achieved for the anti-social/crime measure only.  These results indicate that while the same measures were asked about each group, they functioned differently across groups.
Hypothesis Testing
        H1 predicted that interpersonal and media contact would vary with status.
This hypothesis was partially supported.  When it came to how often respondents saw members of the four groups in media content, responses appeared consistent with content studies:  Anglo Americans were seen most often , followed by African Americans, Latinos and Asian Americans. However, as shown in Table 5, although responses did significantly vary by group, they did not necessarily covary with status.
        Interpersonal contact significantly varied as well.  Not surprisingly, more time was spent with members of one's own group than with members of other groups (See Table 6).  Interestingly, Latinos reported interacting with African Americans (x=2.9; 1= not at all) more often than with any other groups than their own, as might be expected based on status.  But the opposite did not strictly hold true:  African Americans reported interacting with Asian American (3.7) as often as with Latinos (3.6) .  This may be a function of the relatively high status of the African Americans in the sample compared to those in the general population.  Asian Americans and Anglo American reported interacting with each other more than with any other group, a finding consistent with the hypothesis.
        H2 predicted that as media exposure increased, the tendency to negatively stereotype would as well.  Table 7 provides the correlation coefficients between total media exposure and perceptions of the groups.  For African Americans, as media exposure increased, the tendency to negatively stereotype Asian Americans and Latinos increased.  This was true for both the success and crime variables, as indicated by the moderately strong negative correlations.  When looking at the other groups, a more complex picture emerges.  For Anglos, increased media exposure was related to a more positive view of African Americans and a more negative one of Asian Americans as criminals/antisocial.  For Asian Americans, increased media exposure was related to considering African Americans as more successful, while the only significant correlation for Latinos was between media exposure and a positive view of Anglos as law abiding.
        Media exposure was further measured in terms of how frequently respondents saw members of the different groups in the media (Table 8).  Interestingly, all groups were more likely to view Anglos as criminals/anti social as their mediated contact with them increased, although the relationship for Latinos was not significant.  For the latter, despite the small sample size (n=59), moderately strong negatively correlations were found between Latinos mediated contact with Asians and their view of them.  In other words, as such exposure increased, the tendency to negatively stereotype increased.  Also of note, as contact increased African Americans were viewed more positively when it came to crime. In-group bias may be somewhat in evidence here, as Anglos were the only group to negatively stereotype themselves as contact increased.
        The third hypothesis predicted that stereotyping would vary with media content type.  To test this hypothesis, a series of multiple regressions was run on the stereotype variables controlling for demographic variables in the first block (sex, age, education and income) and frequency of interpersonal contact in the second block.  In the third block were the primary independent variables of interest:  number of hours per day with primetime TV, TV news, TV sports, newspapers and the Internet.
        Results indicate that the relationship between media exposure and stereotyping varies not only by content type, but also with the group.  First off, in Table 9 are the results for African Americans (n= 104).  Little or no variance (<10%) was explained for perceptions of members of their own racial group, although watching prime time TV and TV news was related (i.e., significant betas at p<.05) to a more negative view of African Americans as criminals.  Interestingly, the more interpersonal and media contact African Americans had with Anglos, the less positive the perception (R2=.26).  Watching TV news was related to perceiving Asian Americans as anti-social/criminal, as was using the Internet.  However, as interpersonal contact and watching TV increased, African Americans were more likely to perceive Asian Americans as successful.  Of the media variables, only watching prime time was related to stereotyping of Latinos, with a more negative perception as exposure increased.
        For Anglo Americans (n=231) watching televised sports produced positive perceptions of African American, Asian American and Latino success; for other media results varied by group (Table 10). Of note, watching TV news was related to perceiving African Americans as criminals but this finding did not hold for other groups. Only two of the seven equations resulted in R2  higher than .20:  Anglo American anti-social crime and Latino success.   In the first case, watching primetime and using the Internet was negatively related (meaning a more negative perception as exposure increased), while the beta for TV news was positive. The more mediated contact Anglos had with Latinos, the greater the likelihood they perceived them as successful.
        In the Asian American subgroup (n=97), no significant results were found when the  Asian American dependent variables were regressed on the media variables (Table 11).  And while a number of the betas in the equations for Asian Americans were significant at p<.05, by far the most variance was explained for Anglo and Latino antisocial/crime (R2=.61 and .59, respectively).   Even here, however, the direction of the betas differed.   The betas for prime time and TV sports were negative for Anglos, but positive for Latinos. In fact, Asians were more likely to perceive Latinos as successful as contact increased; the same does not hold true for Anglos.
        For Latinos, the betas for Latino success were negative for four of the five media variables. This indicates that as media exposure increased, Latinos were less likely to perceive members of their in-group as successful:  the equation explained 36% of the variance, with significant negative betas for prime time, TV sports, newspaper and TV news (Table 12).  Betas for the Internet were positive in all but one instance (Asian American crime), and significant for African American success and Anglo American crime.

Discussion
        This study has illustrated the complexities of the relationship between media exposure and racial/ethnic stereotyping.  Drawing upon the concept of in-group bias, and the extremity-complexity and contact hypotheses, this study has shown that group membership matters: stereotyping varies by group.  Furthermore, the results of this study indicate that interpersonal and mediated communication are both important factors to consider when conceptualizing and operationalizing contact.
        In this sample, groups varied in their interpersonal and mediated contact to one another; for the latter this variation was not necessarily attributable to status (H1) but seemingly a function of the visibility of different groups in the media.  In-group bias appeared to manifest for interpersonal contact, with respondents seeking out in-group members more frequently than out-group.  Thus, when given the choice it appears that respondents preferred people like themselves, and the closer in status the better.  However, such a choice is not always available in media content, whether it be news or entertainment.
        Results of correlations between media exposure and perceptions of in-group and out-group members indicated some support for the extremity-complexity hypothesis.  As total media exposure increased for African Americans, negative stereotyping increased for all three out-groups, but not for the in-group.  This may be because African Americans' schema for Anglo, Asian and Latinos are more general, and therefore their evaluations more extreme. When it came to in-group perceptions, similar findings were found for Asian and Latino respondents.
        Overall, total media exposure was more likely to be related to negative perceptions than positive ones, a finding congruent with the contact hypothesis:  Media imagery glosses over the complexities of different racial and ethnic groups, thereby facilitating a prejudicial or stereotypical interpretation.  However, the data suggest that the media landscape may indeed be changing.  Respondents in all four groups were more likely to have a positive impression of Latinos as their frequency of seeing them in the media increased.  The opposite, interestingly, held true for perceptions of Anglos and crime.
And the perception of Asian Americans by Latinos was more negative the more  they saw of them.  While these data do not extend beyond correlational evidence, they do provide support for the notion that selective exposure and perception may lead members of different racial and ethnic groups to a variety of interpretations or readings of media content.
        As suggested by previous research (Tan, Fujioka & Lucht,  1997; Vergeer, Lubbers & Scheepers, 2000), it is important to consider the role of different types of content in relation to stereotyping.  In the present study, five different types of media content were examined:  primetime television, televised sports, newspapers, TV news, and the Internet.  The amount of variance explained indicated that such media exposure may be more influential for some groups than others, and that not all types of media content have the same impact.  For example, the five types of media examined here appeared less influential for Anglo and Asian Americans than they did for African and Latinos.  But this is not to say that in some instances strong relationships weren't found.  That media variables were significant even after controlling for interpersonal contact indicates that media content is indeed an additional frame of reference about racial and ethnic groups other than one's own --  sometimes a
 positive one,  sometimes not.
        This tendency is also at times true for the in-group: as media contact increased, Latinos perceived members of their group as less successful.  Such a finding suggests that "positive" representations of Latinos as productinve members of society are still be lacking across media content.
        Yet other results further indicate that some media may be doing a "better" job in their portrayals than are others.  For example, watching prime time was related to negative perceptions among African and Anglo American respondents.  Anglos were more likely to be perceived as antisocial/criminal as viewing TV sports increased, a finding that held for all groups but Anglo respondents: again, an indicator of in-group bias. These findings seem to indicate that it important not to speak only of "media stereotypes," but to take into consideration the very significant differences found in our increasingly diverse media landscape.
        This study was limited by several factors.  The sample proved not to be fully representative of the Atlanta MSA, thereby confounding the status measure. The size of the Latino sample, while proportionate to census figures, was smaller than ideal for the subgroup analyses done here.  The dependent measures were constrained by the fact that they were part of a measuring instrument on an entirely different topic.  More complex measures of stereotypes are warranted.
        Nonetheless, this study extended the research in this area.  It provided a multi-group comparison with a diverse sample of adults.  The design maximized comparisons among all groups; too often in the literature whites are used as the standard by which  to compare another group, most frequently African Americans.   It is important that research in this area be truly multicultural, and not be limited to samples of predominately white students.
        This research has highlighted that the contact hypothesis is a fruitful line of inquiry for mass communications researchers.  Future research should build upon this, providing a stronger test of the role of status in relation to media contact. The extremity-complexity hypothesis should be further integrated to guide researchers toward more complex measures of stereotypes as they relate to media content.  Finally, this research shows that in-group bias is alive and well.  The circumstances under which  the media contribute to producing this bias should be explored in more depth.

References

        Allport, G. W. (1954). The nature of prejudice. Cambridge, MA: Addison-Wesley.
        Armstrong, G. B., Neuendorf, K. A. & Brentar, J. E. (1992). TV entertainment, news, and racial perceptions of college students. Journal of Communication, 42(3), 153-176.
        Brewer, M. B. (1999). The psychology of prejudice: Ingroup love or out-group hate? Journal of Social Issues, 55(3), 429-444.
        Domke, D., McCoy, K., & Torres, M. (1999). News, media, racial perceptions, and political cognition. Communication Research, 26(5), 570-607.
        Entman, R. (1990).  Modern racism and the images of blacks in local television news.  Critical Studies in Mass Communication, 7, 332-345.
        Entman, R, (1992).  Blacks in the news:  Television, modern racism and cultural change.  Journalism Quarterly, 69 (2), 341-361.
        Forbes, H. D. (1997). Ethnic conflict: Commerce, culture, and the contact hypothesis. New Haven, CT: Yale University Press.
        Fujioka, Y. (1999). Television portrayals and African-American stereotypes: Examination of television effects when direct contact is lacking. Journalism and Mass Communication Quarterly, 76(1), 52-75.
        Gandy Jr., O. H. (1998). Communication and Race.  A Structural Perspective.  London:  Arnold.
        Gandy Jr. O. H. & Baron, J. (1998).  Inequality:  It's all in the way you look at it.  Communication Research, 25, 505-527.
        Healey, J. F. (1998).  Race, Ethnicity, Gender and Class.   Thousand Oaks:  Pine Forge.
        Jussim, L., Coleman, L. M., & Lerch, L. (1987). The nature of stereotypes: A comparison and integration of three theories. Journal of Personality and Social Psychology, 52(3), 536-546.
        Lavrakas, P.J. (1993).  Telephone survey methods.  New York:  Sage Publications.
        Linville, P. W., & Jones, E. E. (1980). Polarized appraisals of out-group members. Journal of Personality and Social Psychology, 38(5), 689-703.
        Matabane, P. W. (1988). Television and the black audience: Cultivating moderate perspectives on racial integration. Journal of Communication38(4), 21-32.
        Tan, A., Fujioka, Y., & Lucht, N. (1997). Native American stereotypes, TV portrayals, and personal contact. Journalism and Mass Communication Quarterly, 74(2), 265-284.
        Vergeer, M., Lubbers, M. & Scheepers, P. (2000). Exposure to newspapers and attitudes toward ethnic minorities: A longitudinal analysis. The Howard Journal of Communication, 11(2), 127-143.
        Wong, P., Lai, C. F., Nagasawa, R., & Lin, T. (1998). Asian-Americans as a model minority: Self-perceptions and perceptions by other racial groups. Sociological Perspectives, 41(1), 95-118.