HIV/AIDS appeared in the United States

..... Rural Populations..... Urban-Rural Differences in Motivation to Control Prejudice Toward People With HIV/AIDS: The Impact of Perceived Identifiability in the Community Janice Yanushka Bunn, PhD; 1 Sondra E. Solomon, PhD; 2 Susan E. Varni, MA; 2 Carol T. Miller, PhD; 2 Rex L. Forehand, PhD; 2 and Takamaru Ashikaga, PhD 1 ABSTRACT: Context: HIV/AIDS is occurring with increasing frequency in rural areas of the United States, and people living with HIV/AIDS in rural communities report higher levels of perceived stigma than their more urban counterparts. The extent to which stigmatized individuals perceive stigma could be influenced, in part, by prevailing community attitudes. Differences between rural and more metropolitan community members attitudes toward people with HIV/AIDS, however, have rarely been examined. Purpose: This study investigated motivation to control prejudice toward people with HIV/AIDS among non-infected residents of metropolitan, micropolitan, and rural areas of rural New England. Methods: A total of 2,444 individuals were identified through a random digit dialing sampling scheme, and completed a telephone interview to determine attitudes and concerns about a variety of health issues. Internal or external motivation to control prejudice was examined using a general linear mixed model approach, with independent variables including age, gender, community size, and perceived indentifiability within one s community. Findings: Results showed that community size, by itself, was not related to motivation to control prejudice. However, there was a significant interaction between community size and community residents perceptions about the extent to which people in their communities know who they are. Conclusion: Our results indicate that residents of rural areas, in general, may not show a higher level of bias toward people with HIV/AIDS. The interaction between community size and perceived identifiability, however, suggests that motivation to control prejudice, and potentially the subsequent expression of that prejudice, is more complex than originally thought. HIV/AIDS appeared in the United States primarily in metropolitan areas, and most research has focused on HIV-infected individuals living in metropolitan epicenters. HIV/AIDS is now occurring with increasing frequency in small towns and rural areas. The Centers for Disease Control and Prevention estimate that as of 2005, over 50,000 individuals were living in non-metropolitan communities when diagnosed, accounting for over 5% of the individuals diagnosed with AIDS in the United States. 1 Because many people with HIV/AIDS live outside of major metropolitan areas, it is important to examine issues relevant to individuals from non-metropolitan areas, and to document the ways in which their experiences differ from their urban counterparts. People with HIV/AIDS often report being stigmatized, a process by which individuals having 1 Department of Medical Biostatistics, University of Vermont, Burlington, Vt. 2 Department of Psychology, University of Vermont, Burlington, Vt. We greatly appreciate the contributions to sample selection and data collection made by Tracy Nyerges, Joan Skelly, and Barbara Branch. This research was supported by a National Institute of Mental Health Grant, Rural Ecology and Coping with HIV Stigma, RO1 MH 066848 obtained by Sondra E. Solomon, PhD and Carol T. Miller, Ph.D. For further information, contact: Janice Yanushka Bunn, PhD, Department of Medical Biostatistics, University of Vermont, 24C Hills Science Building, Burlington, VT 05405; e-mail janice.bunn@uvm.edu. C 2008 National Rural Health Association 285 Summer 2008

attributes or characteristics that are devalued by others experience prejudice, discrimination, stereotyping, and exclusion. 2 Because stigma arises from an affected person s experiences with unaffected community members, the extent to which an affected person perceives stigma could be influenced by prevailing community attitudes. The more negative community attitudes are toward people with HIV/AIDS and groups known or assumed to be at risk for HIV/AIDS, the more likely it is that people with HIV/AIDS residing in these communities will perceive that they are stigmatized. Heckman and colleagues found that the perception of discrimination and fear of discovery was greater among HIV-infected individuals living in rural rather than urban communities. 3 Early studies suggested that these perceptions may be accurate: rural community members appear to have more negative attitudes toward people with HIV/AIDS than those in large cities. 4,5 These studies, however, were conducted during the first decade of the HIV epidemic, and overt expressions of AIDS-related stigma appear to have decreased over time. 6,7 While initial reactions to people with HIV/AIDS may be negative and relatively automatic, responses can be adjusted, based on the degree of one s motivation to control prejudice. 8 The ability and motivation to control automatic responses may be what is responsible for the difference between prejudiced and non-prejudiced individuals. 9 Many people are externally motivated by societal norms not to be (or not to appear to be) prejudiced toward stigmatized groups. Previous research has demonstrated that expressions of prejudice depend on whether people are externally motivated by social disapproval to appear non-prejudiced and whether their behavior is identifiable. 10-14 Research on the effects of public versus private expressions of prejudice is important for understanding differences in how motivation to control prejudice might affect expressions of prejudice in rural and urban areas. One aspect of living in small communities is that people are more likely to know each other than are people in urban areas. This makes people more identifiable (less anonymous) in small communities, which increases the social pressure people experience. Highly identifiable people may conform more to social pressures with respect to prejudice than people who are less identifiable. Thus, both community size and perceived identifiability could affect how externally motivated community residents are to avoid acting prejudiced toward people with HIV/AIDS. Higher perceived stigmatization among people with HIV/AIDS living in rural areas could indicate that rural community norms are more accepting of behaviors that allow the enactment of HIV/AIDS stigmatization than more urban communities. Thus, there may be lower levels of internal motivation to control prejudice in rural areas than urban areas. The goal of this study was to compare motivation to control prejudice toward people with HIV/AIDS among residents of rural, metropolitan, and micropolitan communities. We hypothesized that community size and the degree to which people perceived themselves to be identifiable within their communities would be related to whether they were externally motivated to control prejudice toward people with HIV/AIDS. In addition, we hypothesized that people in rural areas might be less internally motivated to control prejudice toward people with HIV/AIDS than their urban counterparts. Methods Participants. This was part of a larger study in which we interviewed 203 HIV-infected individuals residing in Vermont and northern New England. They were recruited from Comprehensive Care Clinics providing medical services to individuals with HIV/AIDS, community-based organizations, including AIDS Service Organizations, word of mouth, and advertisements in local media. Eleven to 13 community members, ages 18-75, were selected for each participant with HIV/AIDS. The goal was 12 community members interviewed per participant, providing 95% confidence to estimate community attitudes within 16%. This report is limited to data collected from the community sample. Procedures. Interviews were conducted with the CI-3 Computer Aided Telephone Interviewing system 15 using a random digit dialing sampling scheme to identify community members within the 3-digit telephone exchange of the participant s telephone number or the telephone exchange for the participant s town of residence. The interview lasted approximately 30 minutes and consisted of a range of questions regarding attitudes and concerns about various health issues affecting their communities. Questions about motivation to control prejudice toward several target groups, including people with HIV/AIDS, were also asked. Interviews were carried out between 2004 and 2006. Respondents were not compensated for their participation in this study, which was approved by the University of Vermont Institutional Review Board. The Journal of Rural Health 286 Vol. 24, No. 3

Measures Motivation to Control Prejudice. Internal and external motivation to control prejudice scales were adapted from those devised by Pryor, Reeder and Landau 8 to measure motivation to control prejudice toward people with HIV/AIDS. We included 3 items from the original internal scale (eg, I attempt to act in non-prejudiced ways toward people with HIV/AIDS because it is personally important to me. ) and 3 from the external scale (eg, Because of today s politically correct standards, I try to appear non-prejudiced toward people with HIV/AIDS. ). Responses were recorded on a 4-point scale, ranging from 1 (strongly disagree) to 4 (strongly agree). Scores were calculated as the mean of the responses to the questions on each scale, and were computed such that higher values indicated increased motivation. The alpha coefficients for the internal and external motivation to control prejudice scales were 0.82 and 0.74, respectively. Community Size. The size of the respondents residential community was based on that of the matched participant with HIV/AIDS; thus, each community member was clustered within the community of residence. Communities were classified as metropolitan, micropolitan, or rural using guidelines set forth by the Office of Management and Budget for core based statistical areas. 16 Metropolitan counties have at least 1 urbanized area with a population of 50,000 or greater. Micropolitan counties have at least 1 urbanized area with a population between 10,000 and 50,000. All others are classified as rural. Nine metropolitan (42.4% of respondents), 10 micropolitan (38.4% of respondents), and 6 rural communities (19.2% of respondents) were represented in this survey. Identifiability. The interview included 2 questions regarding respondents perception of their identifiability within their community. Respondents were asked to estimate (1) what percentage of people in their town knew them personally and (2) what percentage of people in their town knew who they were, but did not know them personally. Responses, which ranged from 0% to 100%, were dichotomized at the median response for each question (10% for the estimate of those who knew the respondent personally, and 20% for those who knew the respondent, but not personally). Other Demographic Characteristics. Standard survey measures assessed age, gender, marital status, and race/ethnicity. Statistical Analyses. The analyses were based on the General Linear Mixed Model, incorporating adjustments for clustered design. 17 Preliminary results suggested that the pattern of factors influencing internal and external motivation differed; therefore, separate analyses were performed. The fixed effects included size of the community of residence, age, gender, and identifiability in the community, with model building to incorporate possible interactions. Random effects allowed adjustment for the intra-class correlation among community respondents within the same community. Reported results are those in which the covariance parameters are constrained to be positive. Since effect size (ES) estimates such as partial eta cannot be calculated with a mixed model, an ad hoc estimation of effect was calculated as the difference between the means divided by the standard deviation, incorporating the adjustment for clustering of individuals within the community. All analyses were performed using SAS, version 9.1. 18 Results The response rate was 68%, resulting in a total of 2,444 completed interviews. Forty-eight percent of the sample were 40 to 59 years old, 26% were less than 40 years old, and the remaining 26% were 60 years of age or older. A majority of respondents were female (64.5%), and, consistent with the racial/ethnic distribution of Vermont, New Hampshire, Maine and northern New York, over 95% were white. Over 95% were high school graduates, with almost 39% having completed college. Identifiability in the Community. Respondents estimates of the percentage of people in their community who knew them personally, and the percentage who knew who they were, but not personally, differed by community size (F[2, 197] = 24.92, P <.01, F[2, 197] = 37.47, P <.01, respectively), with metropolitan residents estimating lower percentage than micropolitan and rural residents. Rural and micropolitan residents did not differ in identifiability on either measure. Residents of metropolitan communities estimated that they were known personally by 14.5% ± 20.5 (mean ± standard deviation) of the people in their town, compared to micropolitan residents, who estimated being known personally by 22.3% ± 24.3 of town members, and rural residents who estimated being known personally by 23.5% ± 24.6 of their community. Respondents estimates of the percentage of the community who knew who they were, but did not necessarily know them personally, were higher at 21.1% ± 24.7, Bunn, Solomon, Varni, Miller, Forehand and Ashikaga 287 Summer 2008

Table 1. Gender and Age Differences in Motivation to Control Prejudice Toward People With HIV/AIDS Internal Motivation to External Motivation to Control Prejudice Control Prejudice Mean (SD) Mean (SD) Gender Females 3.26 (0.49) 2.65 (0.62) Males 3.05 (0.48) 2.51 (0.56) Age Group 18-39 years of age 3.29 (0.50) 2.60 (0.64) 40-59 years of age 3.19 (0.51) 2.58 (0.63) 60 year of age and 3.06 (0.44) 2.62 (0.52) older Responses on a 4-point scale where higher values indicate greater motivation. 33.4% ± 28.0, and 36.2% ± 29.9 for metropolitan, micropolitan, and rural residents, respectively. Motivation to Control Prejudice. The mean scores on the internal and external motivation to control prejudice scales were 3.18 ± 0.50 and 2.60 ± 0.61, respectively, indicating that the respondents were both internally and externally motivated to control prejudice toward people with HIV/AIDS (Table 1). While scores on the 2 scales were significantly correlated, the size of the correlation was moderate (r = 0.24, P <.01). Respondents reported higher internal than external motivation to control prejudice (F[1, 197] = 1,407.00, P <.01). Internal motivation varied by both age and gender of the respondent (Table 1). Females reported higher internal motivation scores than males (F[1, 2,375] = 102.50, P <.01) and scores decreased with increasing age (F[2, 2,375] = 36.03, P <.01). External motivation to control prejudice was higher for women than men (F[1, 2,377] = 27.24, P <.01), but these scores did not vary with age (F[2, 2,377] = 0.85, P =.43). Impact of Identifiability in the Community on Motivation to Control Prejudice. Being Known Personally. When corrected for age and gender, neither size of community of residence nor personal identifiability (being known personally by 10% or more of the community compared to being known personally by less than 10% of the community) was related to internal motivation to control prejudice (Table 2; F[2, 197] = 1.54, P =.22 for community size; F[1, 2,296] = 0.53, P =.46 for being known personally). Similarly, external motivation to control prejudice did not differ with size of the community of residence (F[2, 197] = 0.18, P =.83). On the other hand, external motivation to control prejudice was influenced by the percentage of the community to whom the respondent was known personally (F[1, 2,292] = 22.83, P <.01, ES = 0.20), with those known by 10% or more of the community reporting higher external motivation scores (2.66 ± 0.61) than the less well-known community members (2.53 ± 0.60). Being Known, But Not Personally. The results differed when the measure of identifiability was the percentage of the community to whom one is known, but not necessarily known personally (Table 2). When corrected for age and gender, community size interacted with perceived identifiability to influence internal motivation (F[1, 197] = 3.57, P =.03). Examination of the simple effects suggests that among those who perceive themselves to be more well known, internal motivation to control prejudice is lower for those living in rural communities than those living in Table 2. Effect of Community Size and Identifiability on Motivation to Control Prejudice Known Personally Known Personally Known, but not Personally, Known, but not Locus of Motivation Size of to Less Than 10% to 10% or More to Less Than 20% Personally, to 20% or to Control Prejudice Residence of the Community of the Community of the Community More of the Community Mean (SD) Mean (SD) Mean (SD) Mean (SD) Internal Metropolitan 3.22 (0.51) 3.20 (0.50) 3.21 (0.49) 3.22 (0.53) Micropolitan 3.18 (0.47) 3.17 (0.50) 3.15 (0.47) 3.19 (0.51) Rural 3.18 (0.51) 3.14 (0.49) 3.23 (0.49) 3.11 (0.50) External Metropolitan 2.55 (0.61) 2.66 (0.64) 2.56 (0.60) 2.64 (0.65) Micropolitan 2.50 (0.59) 2.68 (0.61) 2.52 (0.59) 2.67 (0.62) Rural 2.57 (0.57) 2.62 (0.57) 2.64 (0.58) 2.58 (0.57) Responses on a 4-point scale where higher values indicate greater motivation. The Journal of Rural Health 288 Vol. 24, No. 3

metropolitan (t[197] = 2.49, P =.01, ES = 0.19) or micropolitan communities (t[197] = 2.23, P =.03, ES = 0.16), with no difference between the metropolitan and micropolitan residents (t[197] = 0.45, P =.65). Size of community had no effect among those who perceived themselves to be less well known (F[2, 197] = 1.32, P =.27). Examined within community size, the effect of being known, but not personally, appears to be important only in rural communities, with those who perceive themselves to be more identifiable reporting lower internal motivation to control prejudice than those who perceive themselves to be known by less than 20% of their community (F[1, 197] = 5.62, P =.02, ES = 0.23). Community size and perceived identifiability also showed a significant interaction when the outcome variable was external motivation to control prejudice (F[2, 197] = 4.03, P =.02). In contrast to internal motivation, however, external motivation in rural respondents did not differ with the percentage of the community by whom the respondent was known (F[1, 197] = 1.40, P =.24). Similarly, this did not affect external motivation in those residing in metropolitan communities (F[1, 197] = 3.27, P =.07). The estimated percentage of people to whom one is known did, however, affect external motivation scores for those living in micropolitan communities, with those known, but not personally, by 20% or more of their community reporting higher external motivation than those who were known by less than 20% (F[1, 197] = 10.46, P <.01, ES = 0.22). Discussion Internal motivation to control prejudice arises from individuals personal endorsement of egalitarian and non-prejudiced beliefs. Although those who are highly internally motivated to control prejudice may harbor automatic or unconscious prejudiced reactions to members of a stigmatized group, they are motivated to bring their beliefs, attitudes, and behavior in line with their personal non-prejudiced values. In contrast, external motivation to control prejudice arises from fear of social disapproval. Although they may not personally endorse egalitarianism, people high in this domain strive to present themselves as non-prejudiced. This study examined motivation to control prejudice rather than examining prejudiced attitudes directly. Motivation to control prejudice, however, has been shown to be correlated with prejudiced attitudes, and indeed may be what distinguishes prejudiced individuals from those who attempt to act in an unbiased fashion. 9 Research on African Americans as targets of prejudice indicates that higher internal motivation to control prejudice is associated with more favorable attitudes, while higher external motivation is associated with more explicit bias. 11,14 Similar results were seen when the target of prejudice were people with HIV/AIDS. 19 To our knowledge, this is the first study to compare urban-rural differences in community members motivation to control prejudice toward people with HIV/AIDS. Overall, respondents tended to report that they were motivated to control prejudice toward people living with HIV/AIDS. This supports the findings of the national samples studied by Herek and colleagues, 6,20 who suggest that overt expressions of prejudice toward people with HIV/AIDS are not endorsed by the majority of non-infected community members. Interestingly, community size alone did not predict differences in internal or external motivation to control prejudice. The lack of a main effect for community size suggests that community members expression of prejudice is influenced by more than the size of the town in which they live. Indeed, our results suggest that internal and external motivation to control prejudice toward people with HIV/AIDS are a function both of community size and of how identifiable people believe they are in their community, a relationship that differs for internal and external motivation. The influence of perceived identifiability in the community was dependent upon whether identifiability was measured as being known personally by members of the community, or being known, but not personally. No attempt was made to examine the accuracy of the respondents estimates of perceived identifiability because our focus was on community residents perception of how well known they are, rather than the actual percent of the community that either knows them or knows who they are. People who reported being known personally by 10% or more of their community were more externally motivated to control prejudice. This is consistent with the conceptualization of external motivation, which posits that the evaluations and approval of others are important motivators. Thus, the more people with whom one interacts directly, the more important it may become to appear non-prejudiced. On the other hand, although both internal and external motivations to control prejudice were related to the perception that one is recognized, this influence was dependent upon the size of the community in which one lived. We had predicted that internal motivation to control prejudice, stemming from personal beliefs and values, would be lower in rural than in urban communities, due to the social conservativism of rural areas. Instead, we found that although community size by itself was not related to internal motivation to Bunn, Solomon, Varni, Miller, Forehand and Ashikaga 289 Summer 2008

control prejudice, rural community members who reported being relatively well known, but not personally, were less internally motivated to control prejudice toward people with HIV/AIDS than rural community members who reported being less well known. Being better or less well known by people in the community was not related to internal motivation to control prejudice in metropolitan or micropolitan communities. This finding suggests that in rural areas, people who are identifiable by greater numbers of community members may be less likely to control their expressions of prejudice. Because of their more identifiable role in the community, the attitudes of this subset of individuals may be more influential than the attitudes of those who are less known in the community. Thus, the reports of greater perceived stigma by people with HIV/AIDS living in these rural communities may be a consequence of encounters with a highly identifiable subset of their community. Among people living in micropolitan areas, being known, but not personally, is related to higher external motivation to control prejudice. This was not found in the rural communities, however, despite similar perceptions of community identifiability. Because micropolitan communities are intermediate in size between metropolitan and rural locales, people living in these communities may feel less certain about the congruence of their internal motivation with those of the surrounding community, yet more pressured to conform to prevailing standards. It is possible that social messages may be more mixed in these communities, leading residents, particularly those who perceive themselves to be more identifiable to both friends and strangers, to be more attuned to social cues. However, without asking respondents their perceptions of their towns, or determining social trends in these 3 types of communities (eg, political conservativism, beliefs about social norms not health related), this explanation is speculative. These findings suggest that urban-rural differences in internal and external motivation to control prejudice are subtle, with small effect sizes. Because correlations between motivation to control prejudice and both implicit and explicit attitudes toward outgroups are moderate, it is likely that differences in attitudes are similarly subtle. Ultimately, the interaction of community members with the targets of prejudice is what is of interest, but the link between community prejudice toward people with HIV/AIDS and the stigma perceived by them has not yet been demonstrated. Equally unclear is the magnitude of differences in prejudice that can be perceived by stigmatized individuals. Thus, it is entirely possible that small, subtle differences in community prejudice can result in large differences in the perception of stigma by people with HIV/AIDS. Further research in this area is essential to understand the environment in which people with HIV/AIDS reside. This is one of the few studies to examine motivation to control prejudice in a community sample. Because identification of the community sample was not proportional to the urban-rural distribution of the population in rural New England, it may not be representative of this region as a whole. Community members were, however, selected randomly within their community; thus, attitudes could be considered representative of those communities identified for examination. In addition, at least 1 resident of each community was HIV-infected, although that may not have been known to the community member being interviewed. We recognize that the metropolitan/micropolitan/rural classification may oversimplify the variability of the social ecology within counties. For example, a county identified as metropolitan may include pockets of rural areas. People living in these rural areas, however, may possess attitudes toward people with HIV/AIDS that are similar to residents of the nearby metropolitan area due to such things as work travel patterns and use of services in the urban center. It should be pointed out that, because this study was conducted in a largely rural section of New England, the population of the metropolitan communities is substantially smaller than the large centers that are normally considered metropolitan. While this study was limited to an examination of motivation to control prejudice, the actual attitudes toward people with HIV/AIDS were not assessed, thus limiting our ability to examine possible urban-rural differences in attitudes. While both internal and external motivation are independently correlated with attitudes, previous work examining the exact nature of the influence does not clarify the combined effect of both sources of motivation, 11,14 making a prediction of actual attitudes toward people with HIV/AIDS, and their potential to vary by community size, impossible in this study. The study was also limited to states with a low incidence of HIV infection. Thus, generalizability to larger metropolitan communities with a greater prevalence of HIV infection is limited. In summary, our findings indicate that rural areas may not necessarily have a higher level of bias toward people with HIV/AIDS, at least as estimated by community motivation to control prejudice. However, there may be a few key residents in rural areas with relatively little internal compunction about being prejudiced who have disproportionate effects on the overall level of prejudice in a community. This might The Journal of Rural Health 290 Vol. 24, No. 3

explain why people with HIV/AIDS in rural areas report more perceived stigma than those living in other areas. References 1. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report, 2005. Vol. 17. Atlanta, Ga: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention.; 2006. 2. Dovidio JF, Major B, Crocker J. Stigma: introduction and overview. In: Heatherton TF, Kleck RE, Hebl MR, Hull JG, eds. The Social Psychology of Stigma. New York: The Guilford Press; 2000:1-30. 3. Heckman TG, Somlai AM, Kalichman SC, Franzoi SL, Kelly JA. Psychosocial differences between urban and rural people living with HIV/AIDS. J Rural Health. 1998;14(2):138-145. 4. Mondragon D, Kirkman-Liff B, Schneller ES. Hostility to people with AIDS: risk perception and demographic factors. Soc Sci Med. 1991;10:1137-1142. 5. Dhooper SS, Royse DD. Rural attitudes about AIDS: a statewide survey. Hum Serv Rural Environ. 1989;13(1):17-22. 6. Herek GM, Capitanio JP, Widaman KF. HIV-related stigma and knowledge in the United States: prevalence and trends, 1991-1999. Am J Pub Health. 2002;92(3):371-377. 7. Herek GM, Capitanio JP, Widaman KF. Stigma, social risk, and health policy: public attitudes toward HIV surveillance policies and the social construction of illness. Health Psychol. 2003;22(5):533-540. 8. Pryor JB, Reeder GD, Landau S. A social-psychological analysis of HIV-related stigma: a two-factor theory. Am Behav Scient. 1999;42(7):1193-1211. 9. Devine PG. Stereotypes and prejudice: their automatic and controlled components. J Personal Soc Psychol. 1989;56(1): 5-18. 10. Dunton BC, Fazio RH. An individual difference measure of motivation to control prejudice reactions. Personal Soc Psychol Bull. 1997;23:316-326. 11. Devine PG, Plant EA, Amodio DM, Harmon-Jones E, Vance SL. The regulation of explicit and implicit race bias: the role of motivation to respond without prejudice. J Personal SocPsychol. 2002;82(5):835-848. 12. Klonis SC, Plant EA, Devine PG. Internal and external motivation to respond without sexism. Personal Soc Psychol Bull. 2005;31(9):1237-1249. 13. Plant EA, Devine PG. Internal and external motivation to respond without prejudice. J Personal Soc Psychol. 1998;75(3):811-832. 14. Hausmann LR, Ryan CS. Effects of external and internal motivation to control prejudice on implicit prejudice: the mediating role of efforts to control prejudiced responses. Basic Appl Soc Psychol. 2004;26(2/3):215-225. 15. WinCati, Version 4.2. Northbrook, Ill: Sawtooth Technologies; 2003. 16. Office of Management and Budget. Standards for Defining Metropolitan and Micropolitan Statistical Areas; Federal Register, 65 (249):82228-82238. Retrieved November 13, 2005 from www.census.gov/population/www/estimates/00-32997.pdf. 17. Murray DM. Design and Analysis of Group-Randomized Trials. New York: Oxford University Press; 1998. 18. SAS System for Windows, Version 9.1. Cary, N.Y.: SAS Institute, Inc.; 2007. 19. Weibust K. The Role of Death Anxiety in the Stigmatization of Persons With AIDS. Burlington, Vt: Social Psychology, University of Vermont; 2007. 20. Capitanio JP, Herek GM. AIDS-related stigma and attitudes towards injecting drug users among Black and White Americans. Am Behav Scient. 1999;42(7):1148-1161. Bunn, Solomon, Varni, Miller, Forehand and Ashikaga 291 Summer 2008