Biological sex, adherence to traditional gender roles, and attitudes toward persons with mental illness: an exploratory investigation.

Undergraduate students (n = 86) responded to the Community Attitudes Toward the Mentally Ill (CAMI) questionnaire and The Hypergender Ideology Scale, which measures the degree to which they adhered to traditional gender roles. Results indicated that males were significantly more likely than females to endorse intolerant attitudes toward persons with mental illnesses. However, when adherence to hypergender ideology was controlled for, no significant differences emerged between the genders. It was determined that strict gender-role adherence, rather than biological sex, accounted for the variance in CAMI scores. Implications for mental health counselors and for selecting predictor variables for future research are discussed.

**********

Negative stigma associated with mental illness has been an area of concern for mental health professionals, family members of, and individuals with, mental illness. Media portrayals of individuals with mental illness often include psychotic killers or violent individuals who sustain little or no basis in reality. People often use slang words such as crazy, psycho, or schizo to refer to others whose behavior deviates somewhat from an expected norm. In addition, misinformation regarding mental illness continues to perpetuate the stereotypes and prejudices that many people have developed with respect to this issue. The climate for acceptance of people with mental illness has been an area of ongoing advocacy for mental health professionals (Wahl, 1995).

Researchers have sought to better understand the variables that contribute to individuals' understanding and acceptance of, or conversely, their lack of empathy and societal rejection of, those with mental illness. The variables of age, race, socioeconomic status, and sex, all have been examined for potential relationships between each of these variables and the perceptions that people hold toward individuals with mental illness (e.g., Esses & Beaufoy, 1994; Granello & Pauley, 2000; Granello, Pauley, & Carmichael, 1999; Morrison, deMan, & Drumheller, 1993; Wolff, Pathare, Craig, & Leff, 1996). For instance, in examining the attitudes and beliefs that men and women each possess regarding mental illness, Leong and Zachar (1999) found that female college students had less restrictive and more benevolent attitudes toward people with mental illness than their male peers. They also found that the female students had more positive attitudes toward seeking psychological services than male students. Morrison et al. found similar results in their examination of university students' attitudes toward persons with mental illness. Specifically, they found that men score higher than women on measures of authoritarianism and social restrictiveness toward individuals with mental illness. More recently, Granello and Wheaton (2001) found that female college students were less socially restrictive in their attitudes toward persons with mental illness. However, other studies with college populations have not found support for differences in tolerance toward persons with mental illness based on the sex of the participant (Cormack & Furnham, 1998; Granello & Granello, 2000; Granello & Pauley; Granello et al.).

All of the previous research has examined differences in attitudes based on biological sex. Thus, the comparisons were simply between males and females, assuming that all individuals hold a sex-specific viewpoint on the matter. In emphasizing this seemingly simple dichotomy, previous research has failed to investigate the embodied construct of gender and its influence on attitudes and perceptions held by men and women. Therefore, previous studies that examined differences based on biological sex alone may have overlooked the intervening variable of gender-role adherence, a variable that has not been studied in relation to tolerance toward persons with mental illness. Kaschak (1992) asserted that it is not the physical sex of an individual that determines one's gender identity. Rather, it is the way in which others interact with and teach the individual how to be appropriately male or female. According to this view, an individual's gender identity is socially, rather than biologically constructed (Levant, 1996). Thus, the significance of biological differences between men and women depends on the situational context in which they live their lives, not on the physical components themselves (Bem, 1996).Therefore, it can be asserted that the gender roles to which individuals subscribe, rather than biological sex, may be the factor that influences their attitudes, opinions, and worldview.

Research on social tolerance has found that individuals who adhere to the gender norms of the society are viewed as normal or typical, whereas, those individuals who deviate from this subscribed norm are often labeled and judged (Schnittker, 2000). A woman who possesses characteristics that are viewed as traditionally male in nature (such as one who is aggressive, independent, in control of her emotions, and competitive) may be viewed as being mentally unhealthy. Likewise, a man who violates his expected gender role may be condemned and may be subjected to negative psychological labels (Chesler, 1989). Cormack & Furnham (1998) found that both men and women whose behavior deviated from traditional sex role norms were viewed as more pathological and received more severe negative societal reactions than their gender-norm-adhering counterparts.

Individuals who have extremely strict adherence to traditional gender ideals have been labeled as hypergender (Hamburger, Hogben, McGowan, & Dawson, 1996). Hypergender individuals possess extremely stereotypic views about sex-related variables and gender beliefs. Thus, the ideals of men being tough and masculine, and women being dependent, care taking, and passive are viewed as the desired characteristics. Hypergender individuals often have similar expectations for the people around them. Thus, women who subscribe to strict models of femininity seek out male partners who, likewise, subscribe to the constructs of hypermasculinity. Inherent in the hypergender construct is a belief that others should behave in certain socially prescribed ways. This belief results in lowered acceptance of individuals whose behavior deviates from social norms. For example, individuals who have strict traditional gender-role adherence have been found, in general, to be less tolerant toward persons who are gay, lesbian or bisexual (Hoover & Fishbein, 1999). Because hypergender individuals are less tolerant toward persons whose behavior deviates from socially sanctioned gender norms, it may be that they also are less tolerant of persons who deviate from any social norms. Thus, hypergender individuals may be less tolerant toward persons who deviate from socially sanctioned expectations of mental health.

The current study investigated the attitudes that undergraduate students hold toward individuals with mental illness to identify if biological sex is a sufficient variable in the prediction of attitudes toward mental illness or whether the students' views differed according to their gender role adherence. The first hypothesis was that there would be differences in self-reported tolerance toward persons with mental illness based on the biological sex of the participant. A second hypothesis was that when adherence to hypergender ideology is used as a covariate, there will be differences in self-reported tolerance toward persons with mental illness, based on the biological sex of the participant. However, we expect that the null will fail to be rejected (or this form of the hypothesis will not be supported), which would be consistent with the notion that sex is not sufficient for examining these attitudes. This hypothesis has not previously been investigated in the literature.

METHOD

Participants

Participants were 86 undergraduate students at a large, Midwestern university. Four students left several items missing and, therefore, were excluded from the analysis, for a total n of 82. The sample was primarily female (66%, n = 57) and Caucasian (84%, n = 72), with 12% (n = 10) identifying as African American, and 4% identifying as members of other races. The majority (84%) were between ages 18 and 21, with an overall range from 18 to 30. Students were divided between classes, with 16% identifying as first year undergraduates, 39% as sophomores, 25% as juniors, and 20% as seniors.

Instruments

The Community Attitudes toward the Mentally Ill (CAMI). The CAMI was developed by Dear and Taylor (1979) to measure the public's attitudes toward persons with mental illnesses on several subscales. It is a 40-item self-report instrument scored on a 5-point Likert-type scale. Respondents are asked to indicate the degree to which they agree with a particular item, with responses ranging from SA (strongly agree) to SD (strongly disagree). Half of the items are reverse-scored to minimize the possibilities of response set bias. The CAMI instrument contains an operational definition of mental illness. Participants are asked to respond to questions regarding their beliefs about mental illness and persons with mental illness, based on a statement in the instructions to the instrument that reads, "the mentally ill refers to people needing treatment for mental disorders but who are capable of independent living outside a hospital." Responses to individual items are added together to form a subscale score, with a possible range from 10 to 50 on each subscale. There are four separate subscales on the CAMI: Authoritarianism; Benevolence; Community Mental Health Ideology; and Social Restrictiveness. Authoritarianism reflects the view that persons with a mental illness are "an inferior class requiring coercive handling" (Taylor & Dear, 1981, p. 226). It includes topics such as the need to hospitalize those with mental illnesses and the importance of custodial care. Sample questions on this scale are: "As soon as a person shows signs of mental disturbance, he should be hospitalized" and "Mental illness is an illness like any other" (reverse scored). Benevolence was described as" a paternalistic, sympathetic view, based on humanistic and religious principles" (Taylor & Dear, p. 226). It addresses the beliefs that society should assume responsibility for those with mental illnesses, that people should be sympathetic and kind, and be willing to be personally involved (Thornton & Wahl, 1996). Sample questions include: "The mentally ill have for too long been the subject of ridicule" and "The mentally ill are a burden on society" (reverse scored). Social Restrictiveness was intended to measure "viewing the mentally ill as a threat to society" (Taylor & Dear, p. 226). It reflects sentiments involving the dangerousness of people with mental illness and the need to maintain social distance (Thornton & Wahl). Sample questions are: "The mentally ill should be isolated from the rest of the community" and "No one has the right to exclude the mentally ill from their neighborhood" (reverse scored). Community Mental Health Ideology has been conceptualized as traditional (mental hospital) versus non-traditional (community-based) mental health care. It addresses the belief in the therapeutic value of community and the acceptance of deinstitutionalized care (Thornton & Wahl). Sample questions include: "The best therapy for many mental patients is to be part of a normal community" and "Mental health facilities should be kept out of residential neighborhoods" (reverse scored). Higher scores indicate greater agreement with the factor concept for each subscale. Therefore, a tolerant person would be expected to have higher scores on the subscales of Benevolence and Community Mental Health Ideology and lower scores on Authoritarianism and Social Restrictiveness.

The reliability of the CAMI has been measured through test-retest, with alpha coefficients on the subscales ranging from .68 to .88 (Taylor & Dear, 1981). In the current study, alpha coefficients ranged from .64 to .83 (see Table 1). Factor analysis was used to assess construct validity. There is evidence of some overlap between the subscales, with interscale correlations ranging from .63 to .77 in previous studies (Taylor & Dear), indicating that the distinctions between the subscales may be more of an academic distinction than a practical one. In the current study, interscale correlations ranged from .64 to .76 (see Table 1).

Hypergender Ideology Scale. The Hypergender Ideology Scale (HGIS; Hamburger et al., 1996) was designed to measure traditional gender role adherence in both men and women. The HGIS is a 57-item instrument that uses a 6-point Likert-type scale with responses ranging from (1) strongly disagree to (6) strongly agree. Twenty of the items are reversed scored in order to reduce response set bias. The results are summed. Higher scores indicate increased traditional gender role adherence in both men and women. Sample questions are "Most women need a man in their lives," "Men should be in charge during sex," and "I like relationships in which both partners are equals" (reverse scored).

The HGIS is based on a blend of The Hypermasculinity Inventory (HMI; Mosher & Sirkin, 1984) and The Hyperfemininity Scale (HFS, Murnen & Byrne, 1991). The HMI was designed to measure macho personality characteristics, attitudes, and beliefs in men. Such personality constellations were found to include aggressive beliefs about entitlement to sex, negative attitudes toward women, an increased propensity toward violence, particularly sexual violence, and the perception that danger is exciting (Mosher & Sirkin). The HFS was designed to measure a woman's belief regarding traditional and stereotypic feminine gender roles. Hyperfeminine women believe that their ability to be successful is based on their relationships with men, they view their physical appearance and sexuality as of paramount importance in romantic relationships, and they hold the expectation that men will adhere to traditional male gender roles (Murnen & Byrne). The research on both the HMI and the HFS has indicated that hypermasculine men and hyperfeminine women share many similar attitudes and beliefs regarding sex and gender roles. Hamburger et al. (1996) proposed that the acceptance of these hypertraditional roles represents a group of attitudes that encompasses both measures.

Reliability for the HGIS was calculated using a sample of 297 participants (150 men, 147 women) resulting in a coefficient alpha of .96 (.94 for the men and .92 for the women). Additionally, a 21-day test-retest reliability of r(119) = .95, was calculated in an independent sample of undergraduate students (Hamburger, Hogben, McGowan, & Dawson, 1998). In the current study, the alpha coefficient was .93 (see Table 1).The validity of the HGIS was supported by the finding that the HGIS is significantly and positively correlated with the HMI in independent samples of men, r(148) = .76 and r(106) = .60. Additionally, independent samples of women indicate a correlation between the HGIS and the HFS with r(145) = .54 and r(125) = .60.

Procedure

All participants were given the CAMI and HGIS questionnaires along with demographic questions during one of their regularly scheduled undergraduate courses in the College of Education. Selected courses were electives and included teacher preparation (primarily education students), study skills (university-wide enrollment), and career development (university-wide enrollment). The demographic questions were included after the CAMI and HGIS questions so as not to bias the opinions of study participants. Completion took approximately 20 minutes.

RESULTS

A preliminary Pearson correlation analysis indicated that the HGIS score had small but significant correlations with the CAMI subscale scores, which were also significantly related to one another, in the anticipated direction (see Table 1). Therefore, to analyze hypothesis one, that differences in tolerance would be based on the biological sex of the participant, a MANOVA was run with biological sex as the independent variable and the CAMI subscales and HGIS total score as the dependent variables. Prior to running the MANOVA, the data were tested using SPSS's (1996) EXPLORE program, and all assumptions of the MANOVA were met. Effect sizes were generated from the MANOVA, PARAMETER subcommand, within SPSS. Effect size was estimated by eta, part of the r family of effect sizes (Rosenthal, 1994). Post-hoc power was estimated from the MANOVA program. MANOVA results revealed that males endorsed significantly less tolerant beliefs on the CAMI subscales of Benevolence and Social Restrictiveness. Additionally, males were significantly more likely to have higher scores on the Hypergender Ideology Scale than females (see Table 2).

Because males were far more likely to score on the HGIS in a manner that indicated hypergender ideology, it is difficult to determine whether the sex of the participant or the degree of adherence to hypergender caused differences in CAMI subscale scores. In other words, males tended to score in less tolerant directions on the subscales, but from these results, it is impossible to know whether their biological sex or their hypermasculinity was the relevant variable. Therefore, to analyze hypothesis number two, that when adherence to traditional gender roles was used as a covariate, differences in attitudes would emerge based on the biological sex of the participant, a MANCOVA was run to factor out the effects of HGIS score and determine whether sex would still impact the CAMI subscale scores. The fixed factor in this analysis was sex, and the covariate was HGIS score. Results indicated that when hypergender scores were statistically controlled for, there were no significant effects on any CAMI subscale score based on participant sex [Authoritarianism, F(1,79) = .18, p = .67; Benevolence, F(1,79) = 2.38, p = .13; Community Mental Health Ideology, F(1,79) = .04, p = .84; Social Restrictiveness, F(1,79) = .51, p = .48].

DISCUSSION

The results of the study supported our apriori beliefs about the hypotheses. The first hypothesis was supported. In general, males scored in a less tolerant direction on two of the four CAMI subscales (Benevolence and Social Restrictiveness). For the second hypothesis, we failed to reject the null hypothesis, which was the result that was anticipated. In other words, biological sex was not significantly related to tolerance when hypergender ideology was controlled for. Thus, adherence to hypergender ideology, rather than biological sex, was related to attitudes toward persons with mental illnesses. Correlations between the instruments demonstrate that persons with higher hypergender scores were more likely to be more authoritarian, more socially restrictive, and less benevolent toward persons with mental illnesses, as well as holding less tolerant beliefs about community mental health.

The current study has two major implications. The first relates to tolerance toward persons with mental illnesses. The second implication focuses attention on a larger issue, the use of biological sex as a predictor variable in mental health research.

Mental health counselors have recognized the importance of social and familial support in the treatment and progress of those with mental illnesses (Riebschleger, 1991). However, it has been only recently that research has focused on individual characteristics that may contribute to, or detract from, tolerant attitudes. The findings from the current study are the first to examine gender role adherence, rather than biological sex, and indicate that those individuals with extremely traditional gender roles are less likely to express attitudes that are supportive and caring of persons with mental illnesses. It is possible that individuals with rigid gender-role adherence may have limited abilities to offer social or familial support to persons with mental illness, and interactions between these persons may, in fact, be detrimental to the functioning of the person with the illness. Although the mental health counselor may be unable to have a tremendous effect on a life-long socialization process, he or she can be cognizant of the implications that inflexible and gender-conforming expectations and behaviors can have on relationships. To the extent that this finding generalizes to the population at large, mental health counselors may want to work to limit the interactions of their clients with persons who hold rigid hypergender beliefs or at least give their clients some skills to counter possible negative interactions with the hypergender people in their lives. The results also help mental health counselors understand that certain family members may have a more difficult adjustment to a family member's diagnosis with a mental illness. What may initially seem to be conflict or resistance to accepting the diagnosis of a presenting illness could, in fact, be a manifestation of the family member's intolerance and rigidity. This could be an important treatment consideration for family or individual counseling. The results also serve to remind mental health counselors of the importance of working to find ways to reduce intolerance toward persons with mental illnesses and the social stigma surrounding mental illness. As mental health professionals continue in their advocacy role, working toward the reduction of stigma toward people with mental illness is inherent in this process (Granello & Granello, 2000). There is some evidence to support that education about mental illness can increase participants' willingness to assume some responsibility for assisting persons with mental illnesses, to express higher levels of kindness and benevolence, and to be less willing to view persons with mental illnesses as a threat to society (Granello et al., 1999).

The second finding in this study has broader implications. Research in the field of mental health often uses biological sex as an independent variable. In the current study, had biological sex alone been used as a determinant of attitudes toward persons with mental illness, males would have emerged as generally less benevolent and more socially restrictive. However, further analyses revealed that the intervening variable of adherence to hypergender attitudes is what accounted for the differences, not biological sex alone. Therefore, the results of the current study support the hypothesis that biological sex alone is not a significant predictor of tolerance toward persons with mental illness and helps explain why previous research that has used biological sex as a predictor for attitudes toward persons with mental illness has yielded mixed results. This second finding has implications far beyond the current study. To the extent that previous research has relied on the biological construct of sex to predict behavior, this study calls this practice into question and adds to the growing body of literature that questions the over-reliance on biological sex as a predictor variable (Bem, 1996).

There are limitations that must be addressed regarding this study. First the instruments that the students completed were self-report, and it was not difficult to determine the socially desirable responses for most of the questions. A second limitation involves the construct validity of the CAMI. Interscale correlations are quite high for several subscales and raise the concern that the subscales are not discrete measures. The lower alpha reliability for two subscales, Authoritarianism and Social Restrictiveness, also may have affected the results. Additionally, the size of the sample, as well at its demographic composition (college students, primarily Caucasion, in a large Midwest city) limits the generalizability of the results. Finally, care must be taken when inferring actions based on expressed beliefs. A large body of literature demonstrates that attitudes are a weak predictor of actual behaviors (Rauwald & Moore, 2002), and it is possible that persons who express strong tolerant or intolerant beliefs may actually behave in ways that are quite incongruent with their expressed beliefs. Values and beliefs are based in the abstract whereas behaviors as tied to specific situations. For this reason, indirect methods (e.g., behavioral observations, disguised procedures) to measure attitudes toward persons with disabilities have been advocated (Antonak & Livneh, 1995). In spite of these limitations, the results have clear implications for mental health counselors, both in the ability to contribute to the understanding of tolerance and its relationship to the socialization process as well as informing the selection of predictor variables in future research.

Table 1. Intercorrelations and Alpha Coefficients of the CAMI Subscales
and the Hypergender Ideology Scale

                     CAMI Subscales                    Hypergender
              Auth.      Bene.     Comm.     Social
Auth.         .64 (a)
Bene.        -.69 **     .74 (a)
Comm.        -.64 **      64 **    .83 (a)
Social        .69 **    -.67 **   -.76 *     .67 (a)
Hypergender   .37 **    -.39 **   -.28 *     .28 *        .93 (a)

Note: Auth. = Authoritarian; Bene. = Benevolence; Comm. = Community
Mental Health Ideology; Social = Social Restrictiveness;
Hypergender = Hypergender Ideology Scale

(a) alpha reliabilities are written on the diagonal.

* p<.05. ** p<.01.

Table 2. CAMI Subscale and HGIS Means, Standard Deviations, F, Effect
Sizes, and Post-hoc Power Estimates

Subscale              Males           Females
                   M      (SD)      M    (SD)       F(5, 76)

Authoritarian      26.04 (4.47)     24.21 (4.17)     3.30
Benevolence        35.11 (4.94)     28.53 (3.68)    12.34 **

Community
Mental Health
Ideology          33.93 (5.87)     35.72 (5.24)     1.97

Social
Restrictiveness    27.44 (5.49)     25.22 (3.73)     4.69 *

Hypergender
Ideology
Total Score      163.15 (22.48)   122.47 (25.60)   49.39 **

Subscale          Effect Size   Post-hoc
                   (eta)       Power

Authoritarian         .20         .43
Benevolence           .37         .93

Community
Mental Health
Ideology              .15         .28

Social
Restrictiveness       .22         .57

Hypergender
Ideology
Total Score           .62        >.99

* p<.05. ** p<.01

REFERENCES

Antonak, R. F., & Livneh, H. (1995). Direct and indirect methods to measure attitudes toward persons with disabilities, with an exegesis of the error-choice test method. Rehabilitation Psychology, 40, 3-24.

Bem, S. L. (1996). Transforming the debate on sexual inequality: From biological difference to institutionalized androcentrism. In J. Chrisler, C. Golden, & P. Ruzee (Eds.), Lectures on the psychology of women (pp. 9-21). New York: McGraw Hill.

Chesler, P. (1989). Women and madness. San Diego, CA: Harcourt Brace.

Cormack, S., & Furnham, A. (1998). Psychiatric labeling, sex role stereotypes, and beliefs about the mentally ill. International Journal of Social Psychiatry, 44, 235-247.

Dear, M. J., & Taylor, S. M. (1979). Community attitudes toward the mentally ill. Hamilton, Ontario, Canada: McMaster University.

Esses, V. M., & Beaufoy, S. L. (1994). Determinants of attitudes toward people with disabilities. Journal of Social Behavior and Personality, 9(5) 43-64.

Granello, D. H., & Granello, P. F. (2000). Defining mental illness: The relationship between definition and tolerance. Journal of College Counseling, 3, 100-112.

Granello, D. H., & Pauley, P. (2000). Television viewing habits and their relationship to tolerance toward people with mental illness. Journal of Mental Health Counseling, 22, 162-175.

Granello, D. H., Pauley, P., & Carmichael, A. (1999). The relationship of the media to attitudes toward people with mental illness. Journal of Humanistic Counseling, Education and Development, 38, 98-110.

Granello, D. H., & Wheaton, J. E. (2001). Attitudes toward persons with physical disabilities and mental illnesses. Journal of Applied Rehabilitation Counseling, 32(3), 9-16.

Hamburger, M. E., Hogben, M., McGowan, S., & Dawson, L. J. (1996). Assessing hypergender ideologies: Development and initial validation of a gender-neutral measure of adherence to extreme gender-role beliefs. Journal of Research in Personality, 30, 157-178.

Hamburger, M. E., Hogben, M., McGowan, S., & Dawson, L. J. (1998). Hypergender Ideology Scale. In C. M. Davis (Ed.), Handbook of sexuality-related measures (pp. 287-289). Thousand Oaks, CA: Sage.

Hoover, R., & Fishbein, H. D. (1999). The development of prejudice and sex-role stereotyping in white adolescents and white young adults. Journal of Applied Developmental Psychology, 20, 431-448.

Kaschak, E. (1992). Engendered lives: A new psychology of women's experience. New York: Basic Books.

Leong, F. T. L., & Zachar, P. (1999). Gender and opinions about mental illness as predictors of attitudes toward seeking professional psychological help. British Journal of Guidance and Counselling, 27, 123-132.

Levant, R. (1996). A new psychology of men. Professional Psychology: Research and Practice, 27, 259-265.

Morrison, M., deMan, A. F., & Drumheller, A. (1993). Correlates of socially restrictive and authoritarian attitudes toward mental patients in university students. Social Behavior and Personality, 21, 333-338.

Mosher, D. L., & Sirkin, M. (1984). Measuring a mach personality constellation. Journal of Research in Personality, 18, 150-163.

Murnen, S. K., & Byrne, D. (1991). The Hyperfemininity Scale: Measurement and initial validation of the construct. Journal of Sex Research, 28, 479-489.

Rauwald, K. S., & Moore, C. F. (2002). Environmental attitudes as predictors of policy support across three countries. Environment and Behavior, 34, 709-739.

Riebschleger, J. L. (1991). Families of chronically mentally ill people: Siblings speak to social workers. Health and Social Work, 16, 94-103.

Rosenthal, R. (1994). Parametric measures of effect size. In H. Cooper & L. V. Hedges (Eds.), The handbook of research synthesis (pp. 231-244). New York: Russell Sage Foundation.

Schnittker, J. (2000). Gender and reactions to psychological problems: An examination of social tolerance and perceived dangerousness. Journal of Health and Social Behavior, 41, 224-240.

SPSS. (1996). SPSS[R] base 7.0 for Windows[TM] user's guide. Chicago: SPSS.

Taylor, S. M., & Dear, M. (1981). Scaling community attitudes toward the mentally ill. Schizophrenia Bulletin, 7, 225-240.

Thornton, J. A., & Wahl, O. F. (1996). Impact of a newspaper article on attitudes toward mental illness. Journal of Community Psychology, 24, 17-25.

Wahl, O. F. (1995). Media madness: Public images of mental illness. New Brunswick, NJ: Rutger University Press.

Wolff, G., Pathare, S., Craig, T., & Leff, J. (1996). Community attitudes to mental illness. British Journal of Psychiatry, 168, 183-190.

Lisa Hinkelman, LPC, is a doctoral candidate at The Ohio State University and a community agency counselor. Darcy Haag Granello, Ph.D., LPCC, is an associate professor of Counselor Education at The Ohio State University. E-mail: [e-mail]