CULTURAL FACTORS, STIGMA, STRESS, AND HELP-SEEKING ATTITUDES AMONG COLLEGE STUDENTS

The study is examining the role of cultural and psy chological factors that predict attitudes towards seeking mental health treatment. Given the role of treatment in reducing emotional distress, it is important to und erstand the issues people consider when exploring mental health treatment options. Spe cifically, constructs of stigma and acculturation were explored in relation to both att itudes toward and intentions to seek mental health treatment. Undergraduate and graduate stud nts (246 women, 82 men; 237 White, 89 Nonwhite) recruited from two state co lleges and a community college participated in an online survey assessing demograp hics as well as measures of stigma, acculturation, ethnic identity, racial stressors, m ental health, attitudes towards helpseeking, and actual help-seeking. Stigma was found to be a significant mediator of acculturation (measured as relationship to heritage culture and ethnic identity) and attitudes towards seeking mental health treatment, l ding support to the importance of not using just race/ethnicity as a proxy variabl e for culture when investigating barriers to seeking mental health treatment. In add ition, not all types of stigma were significant predictors; while public and personal s tigmas were significant, self-stigma was not a significant factor. Since public stigma a nd personal stigma are speculated to be the first part of the cognitive process associat ed with stigma, efforts should be made to develop interventions at the public stigma level before the stigma is internalized. Greater personal stigma for men and greater cultura l stigma for ethnic minority participants warrant further investigation into the different types of stigma. In addition, racial/ethnic minority participants reported greate r race-related stress than their White counterparts; race-related stress was associated wi h increased anxiety and stress symptoms. For the racial/ethnic minority participan ts, heritage acculturation and ethnic identity were protective factors, in that gr eater heritage acculturation was associated with lower levels of stress and stronger ethnic identity was associated with increased intentions towards seeking counseling. Fu rthermore, future research in this area can elucidate the complex measurement issues w ith the constructs of acculturation and stigma in college students.

Kessler et al. (1994) reported that when the need was recognized, participants faced the following barriers: situational barriers, such as inconvenience or being unsure of where to seek care (52%), financial barriers (46%), and believing that treatment is not effective (45%). Furthermore, 72% of those who fail to seek treatment report wanting to take care of the problem on their own. Similar results were found for adolescents; in one study only one third of adolescents suffering from a mental illness received treatment (Merikangas et al., 2011). In addition, non-Hispanic Black and Hispanic adolescents were less likely to receive treatment for mood and anxiety disorders compared to White adolescents.
There are many individual, familial, and societal costs associated with untreated mental illness, including lower educational attainment (Kessler, Foster, Saunders, & Stang, 1995), loss of productivity at work (Adler et al., 2006;Kessler & Frank, 1997), and the exacerbation of medical illness, such as heart disease, diabetes, and hypertension (Katon & Ciechanowski, 2002). Yet the National Comorbidity Study found that fewer than 40% of people with serious mental illness (defined by impairment across DSM-IV diagnoses) receive treatment for serious mental illness (Kessler et al., 1994). Although underutilization of mental health services may be a widespread problem, it is particularly salient for racial/ethnic minorities. Research indicates that mental health care disparities among ethnic minorities are prevalent and they are a pressing issue to address (Miranda, Nakamura, & Bernal, 2003).
Acculturation and enculturation, family, collectivism and individualism, stigma, cultural mistrust, help-seeking, religion and spirituality, and finally variability in manifestation and interpretation of distress are key cultural variables in understanding underutilization among ethnic minorities (Goldston et al., 2008). Scheppers, van Dongen, Dekker, Geertzen, and Dekker (2006) reviewed 54 articles to determine the barriers of using health services for ethnic minorities. Scheppers et al. examined barriers at the patient (e.g., sex, ethnicity, SES), provider (e.g., sex, skills, and attitudes), and system (e.g., policy, organizational factors) levels. The review includes quantitative studies (n = 28), qualitative studies (n = 10), combined studies (qualitative and quantitative; n = 6), and other studies (literature reviews and published essays; n = 8). Due to the heterogeneity in the review studies, the authors chose to identify barriers, but the barriers are not necessarily empirically tested.
Barriers at the individual level were identified as being young, unmarried, living in disadvantaged neighborhoods (i.e., high crime rates), life style factors, such as an unhealthy diet or substance use, and poor familial or social support. In addition, the authors identified that ethnic minorities' perception about symptoms and knowledge of disease may be based on cultural values and these perceptions may result in different presentations of symptoms, which may result in diagnosing difficulties.
Duration of stay in the United States had mixed results, with some studies indicating that recent immigrants face greater barriers, while others do not indicate so. Low level of acculturation (i.e., separation or marginalization) is viewed as a potential barrier to seeking health services, with associated barriers such as being uninsured. Language barriers represent difficulties potentially affecting patient confidence, ability to express symptoms, providers' ability to understand such symptoms, overreliance on the family member, etc. The authors also identified other potential patient level barriers, such as gender, ethnicity, education, SES, duration of stay (in the country), language skills, health beliefs and attitudes (i.e., a holistic view of health, supernatural beliefs, etc.), knowledge of health services, health insurance benefits, immigration status, time availability, stress, availability and access to services, traditional remedies (in delaying seeking health services), perceived illness and perceived cause both by the individual and the family.
The present study examines a structural model predicting attitudes towards mental health help-seeking, with acculturation and stress as predictors, and stigma as a mediator. Additional analyses will examine these factors and group comparisons by racial/ethnic group and gender.

Stigma
One of the key variables investigated in mental health utilization is stigma. In a World Mental Health Survey, perceived stigma about mental illness was found to be a worldwide phenomenon among individuals with common mental illness, particularly comorbid depression and anxiety, suggesting that stigma of mental illness is not limited to a particular cultural group and/or only to severe mental illness (Alonso et al., 2008). Mental health stigma is described as a cognitive-behavioral process manifested in three different ways: public stigma, personal stigma, and label avoidance (Corrigan & Wassel, 2008). Public stigma is defined as negative stereotypes about people with mental illness, such as perceiving them as dangerous and weak and blaming them for their problems or perceiving them as childlike and in need of others to take care of them. Self-stigma is defined as the internalization of public stigma (Link, 1987). Self-stigma can have an effect on self-esteem, self-efficacy and result in underachieving or in avoiding growth and independence altogether (Corrigan & Wassel, 2008). Finally, label avoidance is when individuals who are suffering from psychiatric conditions do not seek treatment so they are not labeled as mentally ill, thereby escaping the prejudice and discrimination associated with public and selfstigma (Corrigan & Wassel, 2008). Label avoidance is not a variable of interest in the current study since this study does not focus on a psychiatric sample. Another form of stigma is personal stigma or personal attitudes towards mental illness (Griffiths, Christensen, Jorm, Evans, & Groves, 2004). Personal stigma applies to those individuals who do not have mental illness or who are not yet aware of it, unlike selfstigma (Eisenberg, Downs, Golberstein, & Kivin, 2009). The first step in the stigma cycle is awareness of public stigma, followed by the formation of personal stigma, and then an evaluation of one's own status of mental illness (Corrigan, Watson, & Barr, 2006).
Personal stigma has been negatively associated with help seeking indicators, such as perceived need of medication or therapy and actual use of medication or therapy, and also with nonprofessional support (Eisenberg, Marilyn, Ezra, & Zivin, 2009). However, Eisenberg et al. did not find a significant association between public stigma and help seeking. In addition, higher self-stigma was found among students who were male, younger, Asian, international, more religious, or from a low socioeconomic status. Barney, Griffiths, Jorm, and Christensen (2006) did find a significant and negative relationship between public stigma and help-seeking attitudes, and between self-stigma and help-seeking attitudes. Furthermore, participants who endorsed greater self-stigma towards seeking professional services were less likely to actually seek services when followed up over a two month period, while participants who had sought help reported significantly less self-stigma before seeking help (Vogel, Wade, & Haake, 2006), indicating that stigma has predictive value in actual use of professional services. Many studies have found empirical support for a relationship between stigma and different aspects of treatment. In a clinical sample, greater public stigma has been found to be associated with lower treatment adherence (Sirey, et al., 2001a) and with premature termination (Sirey, et al., 2001b). In addition, individuals with stigma of mental illness were less likely to seek or use mental health services (Kessler et al., 2001;Vogel, Wester, Wei, & Boysen, 2005 (Breslau, Kendler, Su, Gaxiola-Aguilar, & Kessler, 2005). Some potential reasons for disparities include socio-economic status, perceived discrimination, and stress (Wang et al., 2005;Williams, Yu, Jackson, & Anderson, 1997). A factor in disparities could be that individuals from ethnic minorities may perceive symptom severity differently (Okazaki, Kallivayalil, & Sue, 2002). Another key variable in persistence of disorders may be the lower rate of treatment among racial-ethnic minorities (Wang et al., 2005). In addition, ethnic minorities have consistently been found to either delay seeking treatment until a condition is chronic, being reluctant to using mental health services, and having poorer access to services (USDHHS, 2001).
In a study assessing stigma in an urban community college sample, African Americans and Asians were found to have greater stigma towards mental illness than Whites; Latinos were found to have less stigma than Whites (Rao, Feinglass, & Corrigan, 2007). Rao et al. proposed that these disparities were due to African Americans having more negative attitudes toward mental illness compared to other groups due to stressors, such as racism and other social inequalities; they did not speculate on results for Asians or Latinos. In a qualitative study, acculturation, difficulty accessing services, and stigma played a major role in being deterrents of help-seeking behaviors among Chinese immigrants in New York City (Chung, 2010).
Among a clinical population of depressed older African Americans, high levels of public and self-stigma were associated with not being in treatment and not having positive attitudes towards treatment (Conner, 2010). Stigma was analyzed as a mediator in a study assessing attitudes towards psychological counseling in college aged South Asian and White students (Loya, Reddy, & Hinshaw, 2010). White students were found to have more positive attitudes towards counseling than South Asian students. In addition, increased self-stigma in South Asians partially mediated and accounted for 32% of the difference in attitudes towards psychological services.
The authors indicate the need to study cultural variables such as acculturation and its relationship to stigma and help-seeking attitudes.
In a qualitative study investigating panic disorder treatment of low-income African Americans, it was found that concerns about social stigma and concerns about confiding in others in the community, as well a dearth of information of panic disorder, were indicated as personal reasons to not seek treatment. Similarly, at the community level, participants demonstrated limited information about mental illness and stigma against those who seek treatment or have symptoms as "weak" or "spiritually flawed," and these perceptions were identified as deterrents in seeking treatment (Johnson, Mills, DeLeon, Hartzema, & Haddad, 2009). Similarly, among low-income depressed women, White participants were more likely to perceive a need for care than black or Latina women, which was also found to decrease the likelihood of expressing a need for help across all ethnic groups (Nadeem, Lange, & Miranda, 2009). An examination of the same dataset revealed that compared to US-born Whites, Black and Latina women (both immigrant and US-born) were more likely to report stigma concerns and were less likely to want treatment, with the exception of immigrant Latinas who were more likely to want treatment compared to US-born Whites. However, only perceived stigma was examined, and not self-stigma. Stigma was found to be a treatment barrier for African Americans compared to Caucasians in other research as well (Alvidrez, Snowden, & Kaiser, 2008;Snowden, 1998). Overall, there appear to be differences in the stigma against seeking mental health services across racial/ethnic groups.
Another factor explored within the context of stigma is gender. Stigma has been theorized to explain the systemic differences, with males having a lower rate of mental healthcare utilization (Wang et al., 2005). Male gender was positively associated with stigma avoidance and mistrust/fear of the system (Ojeda & Bergstresser, 2008). In the same study, it was found that White male status was specifically significantly associated with stigma avoidance, mistrust/fear of the system, and negative attitudes towards care compared to other racial/ethnic groups, which is in contrast to the literature that racial/ethnic minorities display greater stigma.
These conflicting results stress the importance of examining the intersection of race/ethnicity and gender and its relationship to stigma of seeking mental health treatment.
Together these different forms of stigma toward mental illness results in delayed seeking of mental health services and treatment compliance, which can result in negative long-term outcomes (Corrigan, 2004;Corrigan & Wassel, 2007). Although stigma against seeking mental health services is a widespread problem among all Americans, it is a particularly salient issue for racial/ethnic minorities where factors such as language barriers, acculturation, and access to services may make stigma even more salient (Gary, 2005). It has recently been theorized that ethnic minorities face "double stigma," where negative attitudes towards mental health treatment are compounded by the negative attitudes (stereotypes) and discrimination that ethnic minorities already face, thereby creating a delay in treatment seeking or aborted treatment, resulting in greater morbidity and mortality (Gary, 2005). The present study will assess three forms of stigma and examine it in group comparisons by gender and ethnicity/race, as a predictor of several mental health variables, and as a mediating variable in models of mental health help-seeking.

Acculturation
In psychology, acculturation is often studied at the individual level.
Acculturation occurs when groups of individuals from different culture are in contact, with changes resulting in one or both of the cultures (Redfield, Linton, & Herskovits, 1936). A popular framework used to organize these changes is a bilinear model proposed by Berry (1989) (Ryder, Alden, & Paulhus, 2000). The unilinear perspective has been critiqued for not accurately allowing for the many possibilities of acculturation, specifically biculturalism (For review, Miller, 2007;Rudmin, 2003Rudmin, , 2009). In contrast, Berry et al. (1989) categorized acculturation into four different types: integration (high heritage, high mainstream), assimilation (low heritage, high mainstream), separation (high heritage, low mainstream), and marginalization (low heritage, low mainstream).
Integration is associated with lower psychopathology and marginalization is associated with higher psychopathology (Berry, 2007).
In addition, the literature on acculturation has found that acculturation is significantly associated with the many different aspects of the therapy process. In Asian American college students, a significantly positive correlation was found between acculturation and students' ratings of the counseling process (Wang & Kim, 2010). Among Chinese Americans, acculturation, among other cultural factors, was found to be a significant predictor of attitudes towards seeking professional psychological help (Tata & Leong, 1994). Furthermore, acculturation was related to willingness to seek counseling in both Asian Americans (Atkinson & Gim, 1989) and recent Greek and Italian immigrants (Ponterotto et al., 2001). In all these studies, higher acculturation to US culture was positively related to attitudes towards counseling or counselors.
Recently, acculturation and stigma have been investigated to help elucidate their relationship to the underutilization of mental health treatment. Ting and Hwang (2009) examined acculturation, stigma tolerance, and help-seeking attitudes. Stigma tolerance is defined as being aware of the cultural group's stigma. Low stigma tolerance is indicated by being aware of and worried about the cultural group's stigma towards help-seeking. In contrast, high stigma tolerance is being aware of, but not worried about the group's stigma, if seeking help. These results indicate that the culture of origin may influence one's stigma towards help-seeking. The authors found that acculturation (measured bilinearly) did not have a relationship with help-seeking attitudes; however, stigma tolerance was predictive of help-seeking behaviors.
However, they note that the college sample in this study did not reflect low levels of acculturation and there was not a varied distribution of acculturation. Given the relationship between acculturation and stigma, the current study will include a measure of cultural stigma, defined as examining existing stigma items, from the perspective of the participant on how he/she perceives his/her own community perspective of mental health stigma.

Stress
Perceived discrimination and stress have been implicated in mental health treatment seeking by racial/ethnic minorities (Wang et al., 2005;Williams, Yu, Jackson, & Anderson, 1997). In addition, perceived discrimination has been associated with mental and physical health problems (Lee et al., 2009;Mossakowski, 2003).
Specifically, perceived racism significantly predicted number of poor mental and physical health days (Anderson, 2013). Gary (2005) posits that perceived discrimination/stressors associated with race and the stigma associated with mental health treatment contributes to decreased mental health seeking in ethnic minorities. In the current study, perceived discrimination is measured by a questionnaire on racerelated stressors.

Additional Factors
Help-seeking behaviors have been linked to gender (Rickwood & Braithwaite, 1994;Galdas, Cheater, & Marshall, 2005), with men less likely to seek help than women.

Overview of the Present Study
In summary, both acculturation and stigma have been associated with help seeking attitudes, but they have been studied together only on a limited basis. Given the mental health disparities facing racial/ethnic minorities and the greater disease burden, it may be beneficial to explore both psychological factors, such as stigma and cultural factors, such as acculturation and race-related stressors that may play a role in barriers to care. The current study extends the literature by examining acculturation, race-related stressors, self-stigma, public stigma, and attitudes in a comprehensive structural model. The model will allow examination of mediators and moderators. In particular, while decreased stress and acculturation to mainstream culture is expected to predict more positive help-seeking attitudes, this relationship will be mediated by stigma about mental health treatment, which will be associated with more negative attitudes. All results should be interpreted in the context of systemic factors such as access to care.

Research Hypotheses
• Hypothesis 1: In a comparison of the three structural models (Figures 1, 2, and 3) for all participants, the full model ( Figure 3) will be the best-fitting model: the predictors of acculturation and stress will be directly related to the outcome variable of help-seeking attitudes, while stigma about mental health treatment will also serve to mediate this relationship between acculturation and help-seeking attitudes.
o Hypothesis 1a: Race-related stress will be a significant negative predictor of help-seeking attitudes and a significant positive predictor of stigma.
o Hypothesis 1b: Decreased identification with heritage culture and lower identification with ethnic identity will be significant and positive predictors of help-seeking attitudes.
o Hypothesis 1c: Decreased identification with heritage culture and lower identification with ethnic identity will both be associated with decreased stigma about mental health treatment; in all three models, the direct pathway between acculturation and stigma is expected to be significant and positive in direction.
o Hypothesis 1d: Stigma will be significantly and negatively related to help-seeking attitudes.
• Hypothesis 2: For White participants, heritage and mainstream acculturation, ethnic identity, and race-related stressors will not be significant predictors of depression, anxiety, and stress nor of attitudes towards help seeking and/or intentions towards seeking counseling. Cultural factors will not be significant predictors of stigma. In addition, increased stigma (all types) will significantly predict lower attitudes towards help seeking and lower intentions towards seeking counseling.
• Hypothesis 3: For Nonwhite participants, both decreased heritage and mainstream acculturation, decreased ethnic identity and increased race-related stressors will be significant predictors of increased depression, anxiety, and stress, increased stigma (all types), and lower attitudes of help seeking and lower intention of seeking counseling. In addition, increased stigma (all types) will significantly predict lower attitudes towards help seeking and lower intentions towards seeking counseling.
• Hypothesis 4: White male and female participants will differ from Nonwhite male and female participants in the following ways:

Measures
The full survey is found in Appendix A.
Demographics. 21 questions regarding age, ethnicity, sex, marital status, education level, generation level, and employment status were asked. In addition, participants were asked to indicate the language spoken at home and their history of mental health treatment/current intention.

General Help-Seeking Questionnaire (GHSQ; Deane, Wilson, &
Ciarrochi, 2001). This measure consists of 15 items in matrix format that assess a problem-type and the source of help on a 7-point Likert-scale rating of "extremely unlikely" to "extremely likely," with a higher score indicating higher intention to seek help. The scores range from 11 to 77. Both the type of problem and the source of help can be modified according to the research project; in the original version, emotional or personal problem and suicidal thoughts were evaluated. In the current study, suicidal thoughts were not evaluated due to potential ethical concerns. Two additional sources of help were added for this study: minister or religious leader and chat rooms. help (e.g., minister or religious leader) were made to the AHSQ and the score ranges from 11 to 77. For the current study, participants were not asked to indicate the problem, due to potential confidentiality/ethical concerns; however, participants were asked to indicate if they had sought help from these sources over the course of their lifetime as well as in the past two weeks. , 1995). This is a 21-item scale assessing depression, anxiety, and stress, revised from the 42-item version (Lovibond & Lovibond, 1993 Beck & Steer, 1990). In the current study, the internal consistency of this measure is as follows: Depression (α = .92), Anxiety (α=.85), and Stress (α=.87), which is consistent with previous literature in a nonclinical student sample (Lovibond & Lovibond, 1995). In addition, the measure met the cut-off for good internal consistency of .8 (Carmines & Zeller, 1979), across subsamples of gender and race/ethnicity.

Self-Stigma of Seeking Help (SSOSH, Vogel, Wade, & Haake, 2006).
Based on Corrigan's model, a 10-item self-stigma scale (Self-Stigma of Seeking Help; SSOSH) was developed and conceptualized as "concerns about the loss in self-esteem a person would feel if they decided to seek help from a psychologist or other mental health professional" (Vogel, Wade, & Haake, 2006, p.137). The measure is rated on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree); 5 of the items are reverse scored. The score ranges from 10 to 50, with higher score indicating greater self-stigma. The Cronbach's alpha reliability of the scale ranged from .86 to .90, indicating good reliability. In addition, test-retest reliability was .72.
Furthermore, the SSOSH was negatively associated with attitudes towards and with intent to seek psychological services. However, it has only been tested on a predominantly White, college sample, in the past. In the current study, the Cronbach's alpha of this measure was .74, which is an acceptable reliability coefficient value (Loewenthal, 2001). In addition, the Cronbach's alpha was .7 or higher for female participants and across subsamples of race/ethnicity; however, for male participants it was .67, not meeting the criteria for acceptable internal consistency. Since there are not any self-stigma scales developed for or tested with ethnic minorities, using this scale on a sample of ethnic minorities provides data for at least some preliminary psychometric assessment. internal reliability for the 12-item scale was .89, with 6 reverse-scored items. The score ranges from 0 to 60, with a higher score indicating greater stigma. There is another subscale in the D-D, which consists of 3 items measuring personal stigma, with 1 reverse-scored item and the internal reliability for these items was .78. The score ranges from 0 to 15, with a higher score indicating greater stigma. In addition, 22.7% of the students were ethnic minorities and an additional 10% rated either as

Discrimination-Devaluation
Other or in multiple racial/ethnic categories. Although Eisenberg et al. did not report the psychometric properties for the ethnic minority students in particular, there were some racial/ethnic differences and gender found using this measure, in terms of stigma. Black students had the highest mean public stigma; Asian students had the highest mean personal stigma. Personal stigma was higher among Hispanic male participants compared to their female counterparts. However, a comprehensive psychometric analysis for this measure on racial/ethnic minorities is currently lacking.
In the current study, the Cronbach's alpha of the public stigma items was .83 and the Cronbach's alpha of the personal stigma items was .76. In addition, the Cronbach's alpha was .7 or higher across the subsamples of gender and race/ethnicity.

Cultural Stigma (Revised items from D-D scale).
These items are designed to assess the relationship between one's cultural group's stigma of mental health treatment and one's own personal stigma. The three items from the personal stigma subscale are revised for this study to include the term "cultural/heritage group." Items are: "Someone from my cultural/heritage group would willingly accept someone who has received mental health treatment as a close friend;" "Someone from my cultural/heritage group would think less of a person who has received mental health treatment (reverse scored);" and "Someone from my cultural/heritage group would believe that someone who has received mental health treatment is just as trustworthy as the average person." The total score ranges from 0 to 15, with a higher score indicating greater cultural stigma. These items did not demonstrate good internal consistency in the sample (α = .62). The only subsample that good internal consistency was demonstrated for, was the Nonwhite participants (α = .75), which may be due to the reference to one's cultural/heritage group.

Attitudes Toward Seeking Professional Psychological Help Scale (ATSPPHS; Fischer & Farina, 1995).
The ATSPPHS is a 10 item scale which is a revision of the original 29-item ATSPPHS (Fischer & Turner, 1970). Items are answered on a 4-point Likert scale ranging from "disagree" to "agree;" five of the items are reverse scored and higher scores indicate positive attitudes towards help seeking. The total score ranges from 0 to 30, with higher scores indicating more given problem on a 6-point Likert scale, ranging from "very unlikely" to "very likely." The score ranges from 17 to 102, with higher score indicating higher intention of seeking counseling. Three subscales and internal consistencies were established as follows: Interpersonal Problems (.90), Academic Problems (.71), and Drug/Alcohol Problems (.86) were identified as subscales (Cepeda-Benito & Short, 1998).

Vancouver Index of Acculturation (VIA; Ryder, Alden & Paulhus, 2000).
The VIA is a 20-item measure assessing acculturation from a bilinear framework.
There are two subscales in the measure: heritage (10 items) and mainstream  Deane, Wilson, & Ciarrochi, 2001). In the AHSQ, participants reported who they sought help from in the past 2 weeks and lifetime, with the sources categorized similarly to the GHSQ as informal, formal, and electronic. There were no significant racial differences in the use of informal, formal, and electronic sources of help generally and actually used. Similarly, there were no significant gender differences in the use of informal, formal, and electronic sources of help generally and actually used.
Approximately half the sample (45.7%) indicated that they have sought help from a mental health professional (e.g., counselor, social worker, psychologist, psychiatrist) to get help for personal problems, with 11.3% rating the visit as extremely helpful and 80.8% rating it as helpful (rating of 3) or higher (rating of 4 or 5). Additional demographic information is presented in Tables 1-3. Pearson correlations for independent and dependent variables are listed in Table 4. Depression was significantly correlated with the Stress subscale (r = 0.70) Table 4 Pearson Correlations for Independent and Dependent Variables and Anxiety (r = 0.69); Anxiety subscale was significantly correlated with Stress (r = 0.73), suggesting some overlap of symptoms between these subscales. Heritage culture was significantly correlated with depression (r = -.11) and anxiety (r = -.14), which will be further explored in subsequent analyses. Mainstream and heritage culture were significantly correlated (r = .71); this strong correlation suggests multicollinearity, violating the orthogonal relationship proposed by Ryder, Alden, & Paulhus (2000).
Even for the racial/ethnic minority sample, heritage and mainstream culture were significantly correlated (p = .74). Race-related stress was significantly correlated with depression (r = .14), anxiety (r = .18), heritage culture (r = -.14), and mainstream culture (r = -.26), which will be further explored in subsequent analyses. Ethnic identity was significantly correlated with heritage culture (r = .38).

Path Analysis
In order to test hypothesis 1, direct, mediational, and full models were analyzed and compared for goodness of fit (See Figures 1, 2, and 3). Skewness and Kurtosis indices were examined and were found to be within normal limits (Curran, West, & Finch, 1996;DeCarlo & Finch, 1997). Due to the high correlation between the heritage and mainstream culture variables, the mainstream culture variable was dropped from the model. Mainstream culture was excluded as opposed as heritage culture because it was not significantly correlated to any of the other study variables.
Therefore, the predictor variable of Acculturation was measured by heritage culture and ethnic identity. The second predictor, race-related stress, was measured by the Race-related Events Scale. The mediator variable, Stigma, was measured by self, public, and personal stigma. The outcome variable was attitudes towards seeking mental health treatment. Path analysis results are summarized in Table 5.
First, the fit of each model was evaluated, with consideration of several indices. A χ 2 that was close to non-significant and ideally close to the degrees of freedom, which was interpreted as indicating good fit between the model and the population (Bollen, 1989). In addition, the Comparative Fit Index value (CFI) >.90 and a Root Mean-Square Error of Approximation (RMSEA) <.05 was considered a good fit (Bentler, 1992;Browne & Cudeck, 1993). Then the three models ( Figure 1,   Figure 2, and Figure 3) were compared. The indices of fit were evaluated separately for each model and then a χ 2 difference test was conducted to evaluate the best fitting model.

Direct
Model. In this model, acculturation and race-related stressors were direct predictors of attitudes towards seeking mental health treatment. The χ 2 was significant and not close to the degrees of freedom (Refer to Table 5). The CFI was .93, which meets the criteria for good fit. However, RMSEA was .10, which is not a good indication of fit. Overall, the direct model is considered to be a poor fit.  In the mediational model, the path from acculturation to stigma was statistically significant (β = -.30, p < .05), indicating that participants with lower level of identification with heritage culture and ethnic identity had a higher level of stigma.
There was a medium effect size, with R 2 = 0.13 (Cohen, 1992). However, the path from race-related stress to stigma was not significant (β = .15, p > .05), indicating that participants' race-related stress was not a direct predictor of stigma. The path from stigma to attitudes towards seeking mental health treatment was significant (β = .40, p < .05), indicating that higher levels of stigma were related to more positive attitudes towards seeking mental health treatment. There was a medium effect size, with R 2 = 0.16. In addition, both heritage culture and ethnic identity loaded on significantly onto acculturation and public and personal stigma loaded on significantly onto stigma; however, self-stigma was not significant.  Table 5). The CFI was .99 and RMSEA was .02, which is a good indication of fit.
In the full model, the direct paths from race-related stress to attitudes towards seeking mental health treatment and from acculturation to attitudes towards seeking mental health treatment were not significant. In addition, the path from race-related stress to stigma was not significant. However, the path from acculturation to stigma was significant (β = -.29, p < .05), indicating that participants with lower level of identification with heritage culture and ethnic identity had a higher level of stigma.
There was a small effect size, with R 2 = 0.11. Furthermore, the path between stigma and attitudes towards seeking mental health treatment was also significant (β = .35, p < .05), indicating that high levels of stigma were related to high levels of attitudes towards seeking mental health treatment, with a medium effect size (R 2 =.15). Similar to the mediational model, both heritage culture and ethnic identity loaded on significantly onto acculturation and public and personal stigma loaded on significantly onto stigma; however, self-stigma was not significant.

Multiple Regressions within White and Nonwhite Samples
To test hypotheses 2 and 3, a series of standard multiple regressions were conducted to on both the White and Nonwhite participants to examine whether acculturation, race-related stressors, ethnic identity, and different types of stigma were predictors of attitudes towards seeking mental health treatment. The hypotheses were based on the proposed models (Figures 1, 2, and 3). The analyses were conducted in three different steps to map onto these models, within the Nonwhite and White participants. First, multiple regressions were conducted to examine the relationship between cultural factors and attitudes towards seeking mental health treatment; intentions towards seeking counseling was additionally examined as well, although it was not a part of the original model. Second, multiple regressions were conducted to examine the relationship between cultural factors (i.e., acculturation and ethnic identity) and various types of stigma. Third, multiple regressions were conducted to examine the relationship between various types of stigma and attitudes towards seeking mental health treatment; intentions towards seeking counseling was examined as well. All these analyses were conducted for the White participants and then repeated for Nonwhite participants, as the available literature suggests that ethnic minorities may be even more susceptible to stigma about mental health.  Table 7.  .03 .08 * p < .05. ** p < .01. *** p < .001.

Cultural factors, stigma, attitudes towards and intentions to seek help, and mental health (White participants
A standard multiple regression was conducted with the different types of stigma (public, self, personal, and cultural) predicting attitudes towards seeking mental health treatment. Regression coefficients are presented in Table 8. Public stigma, selfstigma, and personal stigma were examined as predictors of attitudes towards seeking mental health treatment. Only personal stigma was a significant predictor [F(3, 233) = 8.73, p = .000], with β =.276, indicating that a higher level of personal stigma was associated with more positive attitudes towards seeking mental health treatment. There was a small multivariate effect size, with R 2 = 0.101 and adjusted R 2 = 0.090. In addition, a standard multiple regression was conducted with the different types of stigma (public, self, personal, and cultural) predicting intentions towards seeking counseling. None of the stigma variables were significant predictors of intentions to seek counseling. .12 .28 *** * p < .05. ** p < .01. *** p < .001.
Additional analyses examined the relationship between cultural factors and mental health. A series of standard multiple regressions were conducted with heritage, mainstream, and race-related stressors predicting the DASS subscales of depression, anxiety, and stress. The acculturation predictors were not significant. Race-related stressor was a significant predictor of anxiety [F(1, 235) = 7.05, p = .008], with β =.171, R 2 = 0.029 and adjusted R 2 = 0.025, indicating a small multivariate effect size. This result indicates the greater level of race-related stress was associated with higher levels of anxious symptoms. .05 .08 * p < .05. ** p < .01. *** p < .001.

Cultural factors, stigma, attitudes towards and intentions to seek help, and mental health (Nonwhite participants).
To test hypothesis 3, the same multiple regressions were conducted with the further explored with racial/ethnic minority participants. Similar to the White sample, identification with mainstream culture, identification with heritage culture, and ethnic identity were not significant predictors of attitudes towards seeking mental health treatment. Mainstream and Heritage culture were also not significantly related to intentions towards seeking counseling.
Ethnic identity was a significant predictor of intentions to seek counseling [F(3, 87) = 3.139, p = .029], β = .343, indicating that greater identification with ethnic identity was associated with higher intention of seeking counseling. There was a small multivariate effect size, with R 2 = 0.098 and adjusted R 2 = 0.067. Regression coefficients are presented in Table 10. None of the cultural variables were significant predictors of stigma and none of the stigma variables were significant predictors of attitudes towards seeking mental health treatment or intentions of seeking counseling.
Acculturation and ethnic identity were not significant predictors of anxiety or depression. However, both mainstream and heritage cultural affiliation were significant predictors of stress [F(3, 87) = 3.751, p = .014], β = .448 and β = -.556, respectively. This result indicates that greater identification with mainstream culture was associated with higher levels of stress and greater identification with heritage culture was associated with lowers levels of stress. There was a small multivariate effect size, with R 2 = 0.115 and adjusted R 2 = 0.084. Regression coefficients are presented in Table 11. Race-related stressors were not significantly related to any of stigma variables or to attitudes and intentions towards seeking help. Race-related stressors were a significant predictor of stress [F(1, 89) = 5.063, p = .027], with β = .232, R 2 = 0.054 and adjusted R 2 = 0.043, indicating a small multivariate effect size and that greater levels of race-related stress was associated with higher levels of stress.
In addition, race-related stressors were a significant predictor of anxiety [F(1, 89) = 5.117, p = .026], R 2 = 0.054 and adjusted R 2 = 0.044, indicating a small multivariate effect size. This result indicates the greater level of race-related stress was associated with higher levels of anxious symptoms. Race-related stressors were not a significant predictor of depression. .09 .11 * p < .05. ** p < .01. *** p < .001.

Gender and Racial/ethnic Comparisons
To test hypotheses 4a and 5a, a 2 x 2 multivariate analysis of variance (MANOVA) was conducted to examine racial (White and Nonwhite) and gender group differences in heritage acculturation, mainstream acculturation, ethnic identity, and race-related stress. Preliminary testing was conducted to examine for normality, linearity, and homogeneity of variance. Homogeneity of variance was violated for ethnic identity and race-related stress (p = .008 and p = .000, respectively); similar criteria for interpretation as described previously will be used. As a result, Pillai's trace as opposed to the commonly used Wilks' Lambda will be used to interpret the overall group differences, with a larger value interpreted accounting for more of the variance in the linear combination DVs explained by IVs. Pillai's trace is the most robust index of group difference and is preferred when there are violations of homogeneity of variance or unequal sample sizes, both of which occur with the current model (Harlow, 2005). In addition, a conservative alpha level of .025 or .01 is recommended in examining the univariate F-test, rather than .05 alpha level (Tabachnick & Fidell, 2007); all results for this MANOVA will be interpreted using the above criteria. There were significant overall group differences between White and Nonwhite participants on the DVs, F(4, 319) = 6.30, p = .000; Pillai's Trace = .073; η 2 = .073, indicating a small effect size. However, there were no significant group differences between male and female participants on the DVs of heritage acculturation, mainstream acculturation, ethnic identity, and race-related stress.
Additionally, there was no interaction effect between racial group and gender.
However, given the more stringent criteria being used for alpha level, this result will not be further interpreted.
To test hypotheses 4b and 5b, a 2 x 2 MANOVA was conducted to examine racial (White and Nonwhite) and gender group differences in self, public, personal, and cultural stigma. Preliminary testing was conducted to examine for normality, linearity, and homogeneity of variance. Homogeneity of variance was violated for personal and cultural stigma (p = .004 and p = .005, respectively). There were significant overall group differences between White and Nonwhite participants on the stigma variables, F(4, 318) = 2.49, p = .043; Pillai's Trace = .03, with a small effect size. In addition, there were significant group differences between male and female participants on public, self, personal, and cultural stigma variables, F(4, 318) = 4.70, p = .001; Pillai's Trace = .056, indicating a small effect size.
However, there was no interaction effect between racial group and gender. Univariate between-subject tests showed that Nonwhite participants were significantly different than White participants on cultural stigma, (F(3, 321) = 5.96, p =. 015). An examination of the mean scores (Table 3)  To test hypotheses 4c and 5c, a 2 x 2 MANOVA was conducted to examine racial (White and Nonwhite) and gender group differences in attitudes towards help seeking and intentions of seeking counseling. Preliminary testing was conducted to examine for normality, linearity, and homogeneity of variance. Homogeneity of variance was violated for attitudes towards help seeking (p = .020); similar criteria for interpretation as described previously will be used. There were no significant overall group differences between White and Nonwhite participants on attitudes towards seeking help and intentions to seek counseling. However, there were significant group differences between male and female participants on the DVs of attitudes towards seeking help and intentions to seek counseling, F(2, 321) = 11.74, p = .000; Pillai's Trace= .068), indicating a small effect size. Additionally, there was no interaction effect between racial group and gender. Univariate between-subject tests showed that female and male participants differed significantly on attitudes towards seeking help, F(3, 322) = 23.25, p = .000). An examination of the mean scores (Table 2)  To test hypotheses 4d and 5d, a 2 x 2 MANOVA was conducted to examine racial (White and Nonwhite) and gender group differences in the mental health variables of depression, anxiety, and stress. Preliminary testing was conducted to examine for normality, linearity, and homogeneity of variance, with no violations observed. There were no significant overall group differences between White and Nonwhite participants on depression, anxiety, and stress. However, there were significant group differences between male and female participants on the DVs of depression, anxiety, and stress, F(3, 320) = 6.26, p = .000; Wilks' λ= .945; η 2 = .055), indicating a small effect size. Additionally, there was no interaction effect between racial group and gender. Although there were significant macro-level differences, univariate between-subject tests showed that female and male participants did not differ significantly on depression, anxiety, and stress.

DISCUSSION AND CONCLUSIONS
The main objective of this study was to examine potential cultural, social, and psychological predictors of attitudes towards mental health seeking. Seeking help for mental healthcare is a complex issue, particularly due to stigma against people with mental health concerns. There have been mixed results in the literature as to what type of stigma (public vs. self) plays a more prominent role in predicting attitudes, intentions, and behaviors related to mental help seeking. In addition, race/ethnicity is often used as a proxy measure of culture in the mental help seeking literature. The current study extends the literature by more directly measuring cultural experiences, through examining the constructs of acculturation, ethnic identity, and race-related stress in White and Nonwhite college students and using these constructs of culture as predictors of attitudes and intentions towards help-seeking.
The sample was predominantly White and female. Most of the participants were not first generation college students (87.4%). Participants generally reported being mentally and physically healthy; approximately 19.6% of the participants indicated being physically ill for 6 or more days in the past 30 days. Additionally, 92.9% of the participants described their general physical health as excellent, very good, or good, indicating that this sample is overall physically healthy. The measurement of physical health as part of the assessment of mental health is important, due to overlap of somatic symptoms in common mental health concerns, such as anxiety and depression (e.g., shortness of breath, insomnia, weight changes).
In addition, literature indicates that there are cultural variations in the expression and experience of mental health symptoms, with a particular emphasis on somatic symptoms for racial/ethnic minorities (e.g., Lin & Cheung, 1999). However, physical health was not utilized in subsequent analyses due to the homogeneity of the group.
A small segment (13.1%) of the sample is taking psychiatric medications and less than a quarter of the sample (18.9%) indicated feeling the need to seek mental health treatment. However, quarter of the sample indicated that they would seek treatment in the next six months, if needed. Overall, most of the sample (81.4%) feels that they had the resources to seek mental health treatment, if needed. This confidence about having available resources may be indicative of the sample being college students, all on campuses with access to college and/or university counseling centers.
However, it may also be a result of college students being more aware of mental health concerns and resources in general due to the increased educational efforts targeting this generational cohort (Eisenberg et al., 2009).

Psychometrics and Correlations
The subscales of the DASS are significantly correlated, particularly depression with stress and anxiety and the anxiety subscale being significantly correlated with the stress subscale. These results suggest overlap of symptoms and are consistent with the correlations in the literature (Lovibond & Lovibond, 1995). In addition, the overlap may also be representative of the comorbidity between mood and anxiety disorders (DSM-IV-TR, 2000). Since the measure is a self-report and due to the symptom overlap, the results of this questionnaire are not interpreted diagnostically; they represent more of a reflection of some level of mental health distress.
The VIA was designed to measure an individual's relationship to heritage culture and an individual's relationship to mainstream culture independently and as a result the correlations between the subscales are expected to be low (Ryder, Alden, & Paulhus, 2000). For this particular sample, the mainstream and heritage culture are significantly correlated (p = .71). The high and significant correlation in the sample indicates multicollinearity, where both subscales are measuring the same construct.
This high correlation reflects the fact that the sample is predominantly White and therefore the terms "heritage" and "mainstream" may represent the same aspect of their cultural experience. In addition, the sample was predominantly second generation or higher (92.3%), in terms of being born in the United States. Within the racial/ethnic minority sample, where some variation may be expected in terms of the relationship to heritage and mainstream culture, the subscales are similarly significantly correlated (p = .74). Similarly, most of this subsample was also of second generation or higher (71%), which may explain the highly correlated relationship between the heritage and mainstream culture subscales.
Race-related stress is positively correlated with depression and anxiety for the overall sample, consistent with the literature that those who endorse experiencing a race-related stress exhibit symptoms of anxiety, specifically posttraumatic stress disorder . When this significant relationship was further explored with White participants, race-related stress was a significant predictor of anxiety, which does not support hypothesis 2. Within the racial/ethnic minority sample, racerelated stress was a significant predictor of anxiety and stress, but not depression, which is in partial support of hypothesis 3. In addition, race-related stress indicated a negative correlation with both heritage and mainstream cultures in the overall sample, indicating that strong identification with a cultural group may be a potential buffer for race-related stress. However, race-related stress was not significant in terms of stigma and/or attitudes towards treatment for both Whites and Nonwhites. Overall, there is some support that race-related stress is an important aspect of one's cultural experience that warrants further exploration within the context of mental health.
Although there was significant negative correlation between heritage and mainstream culture, and attitudes towards help seeking, neither heritage nor mainstream culture were significant predictors of attitudes towards help seeking for White participants. Similarly ethnic identity was also not a significant predictor of attitudes towards help seeking for the White participants. These results not being significant is in support of hypothesis 2 and may be due to the homogeneity of the sample. Within the racial/ethnic minority sample, these cultural factors were not significant predictors of attitudes towards help seeking, which is not in support of hypothesis 3. However, ethnic identity was a significant predicator of intention to seek counseling, indicating that stronger ethnic identity is related to a greater intention to seek counseling. In contrast, race-related stress was a significant predictor of attitudes towards seeking mental health treatment and intention to seek counseling within the White sample; however the effect size for both predictors did not meet criteria for even a small effect size (i.e., negligible) and will not be further interpreted.
Public stigma is positively correlated with stress and depression, indicating that higher levels of stigma are related to higher levels of stress and depression. Since the study was not designed to establish causal relationships, it is unclear if public stigma results in increased symptoms, due to not seeking treatment or if having symptoms of mental health increases public stigma. In addition, public stigma is negatively correlated with heritage culture and ethnic identity, suggesting that a closer affiliation with heritage culture and/or greater level of ethnic identity may indicate lower level of public stigma. However, neither one of these cultural factors were significant predictors of public stigma, in the White sample, in support of Hypothesis 2, or the racial/ethnic minority sample, which does not support hypothesis 3. Public stigma was not a significant predictor of attitudes towards seeking mental health treatment in both of the subsamples. This is in contrast to the hypotheses 2 and 3 and the literature (Barney, Griffiths, Jorm, and Christensen, 2006), where a negative correlation was expected between the two variables.
Personal stigma was significantly correlated with multiple measures; in particular it was a significant predictor of attitudes towards help seeking for White participants. The results indicate that higher personal stigma meant more positive attitudes towards seeking help. However, self-stigma was not a significant predictor.
Additionally, self-stigma was not significantly correlated with any of the study measures, which explains the very low loading onto the stigma latent variable in the path analysis.
Mainstream culture was a significant predictor of personal stigma, with a negative correlation between the two variables. In addition, mainstream culture was a significant predictor of cultural stigma, with a negative correlation between the two variables. For the White participants, lower affiliation with mainstream culture was affiliated with higher personal stigma, indicating that affiliation with a cultural group is important, in contrast to the prediction in hypothesis 2 where null findings were expected. In addition, higher affiliation with heritage culture was related to lower cultural stigma, potentially implying that for participants where the heritage culture experience is distinctly different from mainstream culture, experience with cultural stigma is important to explore. These results were not significant for racial/ethnic minorities and not in support of hypothesis 3. For both subsamples, mainstream culture was not a significant predictor of any other forms of stigma; heritage culture was also not significant predictor of any of public, personal, and self-stigma. Heritage culture was negatively correlated with public, personal, and cultural stigma; heritage culture was a significant predictor of cultural stigma in the overall sample. Ethnic identity was not a significant predictor of any forms of stigma.

Predicting Mental Health Help-seeking
Direct, mediational, and full models were compared with acculturation and race-related stressors predicting stigma, which in turn predicted attitudes towards mental health help-seeking. The direct model with acculturation and race-related stressors as a predictor of attitudes towards seeking mental health treatment was not significant. This result is important, given that cultural variables, particularly acculturation and race/ethnicity, have been found to play a key role in barriers to seeking help, but not on a consistent basis (Scheppers et al., 2006). The missing piece in these prior examinations may have been including mediating variables, such as stigma. Beliefs about health and perception of illness have also been implicated to be examined closer and the not significant direct model lends support to further investigate mediating variables in the decision making process of seeking treatment. Although the mediational model and the full model were a good fit for the data, the direct paths in the full model were not significant, which is to be expected given that the direct model was not significant. Therefore, the full model did not appear to add any additional improvement of fit, over the mediational model. As a result, the mediational model appeared to best fit the data and was chosen as the best fitting, due to parsimony (Bentler, 2004). These results were only in partial support of hypothesis 1; stigma about mental health treatment did mediate the relationship between acculturation and attitudes towards seeking mental health treatment.
Specifically, it was found that the less identification with heritage culture and lower ethnic identity, the higher the stigma is towards seeking mental health treatment.
Therefore, a strong relationship with one's heritage culture and strong ethnic identity may be a protective factor against having stigma, by functioning as some sort of buffer. However, race-related stress was not a significant factor. Acculturation has been found to have inconsistent relationships to attitudes towards seeking mental health treatment in the literature. The current data provides more support for examining the construct of stigma as crucial in the decision to seek mental health treatment.
Hypothesis 1a was not supported; race-related stress was not a direct predictor of attitudes towards seeking mental health treatment and it was not significantly related to stigma or attitudes towards help seeking. This hypothesis may not have been supported because of the predominantly White sample, who are not expected to have as many race-related stressors . Hypothesis 1b was not supported, since the direct path between acculturation and attitudes towards mental health seeking was not significant, further lending support to examining the mediating variable of stigma.
Hypothesis 1c was partially supported, with acculturation being a significant predictor of stigma about mental health treatment; however, the two constructs are negatively correlated, as opposed to the predicted positive correlation. Higher identification with heritage culture was found to be a potential barrier for seeking mental health treatment (Schepper et al., 2006), although the assumption was that higher identification with heritage culture meant lower identification with mainstream culture. In the current sample, participants identified highly with both heritage and mainstream culture. Given the predominance of White participants, for some participants both subscales may be measuring the same cultural relationship. Since formal mental health treatment was viewed as more of a Western concept, it was speculated that lower identification with heritage culture may imply lower levels of stigma. In contrast, the current result indicates that higher identification with heritage culture and higher identification with ethnic identity are associated with lower levels of stigma. This may be due the highly correlated relationship between heritage and mainstream culture; at any rate, higher identification with mainstream culture is associated with lower level of stigma. However, there is another component to the acculturation construct, which is ethnic identity. Therefore, the current result suggests that having a high identification with heritage culture and a high ethnic identity may be a protective factor against stigma about mental health treatment.
Hypothesis 1d was partially supported, with stigma being significantly, but positively related to attitudes as opposed to a negative correlation. It is possible that those who have a need to seek mental health treatment are both more aware of stigma and the need to seek the treatment. In addition, stigma was a significant mediator between acculturation and attitudes towards seeking mental health treatment in all the three models, direct, mediational, and full.

Race/Ethnicity and Gender: Stigma, Mental Health, and Help-seeking
Comparable to the results in the White sample, acculturation and ethnic identity were not significant predictors of attitudes towards seeking mental health treatment. Results were similar to another study by Ting & Hwang (2009), where acculturation was measured bidimensionally among college students. Similar to the current study, there was not much variability in acculturation in their sample as well.
Although heritage and mainstream culture were not significant predictors of intentions to seek counseling, ethnic identity was a significant predictor of intentions to seeking counseling. This result may be reflective of the high proportion (81.4%) of the overall current sample indicating that they would seek mental health treatment in the next six months, if needed. Cultural factors in the overall sample were not related to stigma among racial/ethnic minority participants. In the White sample, mainstream culture was a significant predictor of personal stigma and heritage culture was a significant predictor of cultural stigma. Since the racial/ethnic minority sample is smaller, both cultural and personal stigmas, which are 3-item subscales, may not have been significant in the multiple regressions. In addition, none of the stigma variables were significant predictors of attitudes and/or intentions towards seeking mental health treatment, in contrast to the White sample, where personal stigma was a significant predictor of attitudes towards seeking mental health treatment. It is possible that stigma towards mental health treatment is mediated by other potential cultural experiences within racial/ethnic minorities, such as distrust of the system. Increased affiliation with mainstream culture was associated with increased stress for the racial/ethnic minority sample, indicating that association with mainstream culture is not a protective factor against stress. However, association with heritage culture is protective factor against stress, with increased affiliation with heritage culture being associated with lower levels of stress. These results further lend support to the importance of exploring cultural relationship from a bidimensional perspective. In addition, the disparate results between mainstream acculturation and heritage acculturation with stress are particularly of note given the high positive correlation between heritage and mainstream acculturation. Overall, acculturation being significantly related to stress offers further support to investigate mental health variables in relation to cultural variables.
Race-related stress was a significant predictor of stress and anxiety as well. In the White sample, race-related stress was a significant predictor of anxiety. Given the high correlation among the subscales of anxiety, stress, and depression, the subscales may not be interpreted as indicative of Major Depressive Disorder or an anxiety disorder. However, generally, the results regarding race-related stress are consistent with greater stress being correlated with symptoms of anxiety and distress (Waelde et al., 2010).
In support of hypothesis 4a and 5a, racial/ethnic minorities had higher levels of race-related stress than Whites, as expected and consistent with the literature . However, there was no significant difference in levels of acculturation, which is not in support of hypothesis 4a and 5a. In partial support of hypothesis 4b and 5b, Nonwhites had higher levels of cultural stigma than Whites. Cultural stigma items, adapted from the personal stigma scale for this particular study, only showed good internal consistency of the items for the Nonwhite participants. This, as well as the significant group difference between Nonwhites and Whites for cultural stigma lends support to further exploring stigma that may be specific to one's cultural group when it pertains to mental health treatment. Past research examined public stigma, personal stigma, and self-stigma as perceived by the mainstream culture only.
Although there no significant difference between Whites and Nonwhites for public, personal, and self-stigma, as predicted in hypotheses 4b and 5b, past research has found such differences and cultural stigma may be a key variable in explaining these differences. A potential reason there may not be significant differences between the groups on other indicators of stigma may be because approximately half the racial/ethnic minority sample was Hispanic/Latino and in a particular study examining stigma from the same framework in an urban college sample, Latinos were found to have less stigma compared to their counterparts (Rao, Feinglass, & Corrigan, 2007).
Overall, given the group differences in cultural stigma and past literature suggesting racial/ethnic group differences regarding stigma, stigma should continue to be explored further as an important component of the decision making process to seek mental health treatment.
In support of hypotheses 4b and 5b, male participants endorsed higher levels of personal stigma compared to female participants, which is a factor to further examine in the differences of help-seeking between men and women, with men particularly delaying treatment seeking (Galdas, Cheater, & Marshall, 2005). In addition, personal stigma items had the greatest internal consistency for male participants (.79) compared to other measures of stigma. Self-stigma, in particular, did not have good internal consistency with male participants. In addition, attitudes towards seeking help had poor internal consistency for male participants. Future studies could examine psychometric differences regarding gender, stigma, and help seeking attitudes, Additional items that are particularly corresponding to masculinity and help seeking would further the research in developing interventions to address lower levels of helpseeking among men.
Despite the differences regarding some types of stigma across race/ethnicity and gender, there were no significant differences in the type of help sought, indicating that when help is sought male/female and White/Nonwhite are equal in seeking out informal, formal, and electronic sources of help. This result suggests that efforts to address help seeking behaviors may not be about the type of help available, but that intervention would occur earlier in the decision making cycle to increase the chance of seeking help.
Contrary to the predictions of hypotheses 4c and 5c, White and Nonwhite participants did not differ on attitudes towards seeking help and intentions of seeking counseling. However, there were significant differences between male and female participants on attitudes towards seeking help, but in the opposite direction, with males having slightly better attitudes towards seeking help than females. This result further supports the complex nature of stigma for male participants and their subsequent decision making process of seeking mental health treatment. Despite significant differences in race-related stress, which is related to higher levels of anxiety and stress, there are no significant group differences between Whites and Nonwhites on mental health variables, which is not in support of hypotheses 4d and 5d. The non-significance may be due to the sample being relatively healthy, both physically and mentally. Although there were significant group differences in gender for depression, anxiety, and stress, which supports hypotheses 4d and 5d, the lack of micro-level differences makes it difficult to interpret the nature of the difference.

Limitations
Despite some of the significant findings in this study, there are some limitations to consider. The sample in the study was not as diverse as anticipated in terms of race/ethnicity and generational status in the US. As a result, it is difficult to interpret some of the results as they relate to acculturation. In addition, the sample is a relatively healthy cohort, in terms of physical and mental health. Since this is not a psychiatric sample, the results related to stigma and attitudes and/or intentions of seeking mental health treatment may not be generalizable to a more need-based population. In addition, the study did not measure longitudinal assess if attitudes and/or intention to seek help predicted actual behavior to seek help. Previous literature demonstrates that there can be an established link between attitudes and behaviors.
The current sample is a young cohort and may not have been exposed as much to the society's attitudes about stigma towards mental health treatment. Furthermore, the current study examined mental health treatment factors in a closed access system, where college/university students had access to the school counseling center and other health services. However, attitudinal barriers have been found to be just as prevalent, if not more prevalent, than structural barriers (Jagdeo, Cox, Stein, & Sareen, 2009).

Conclusions
Addressing barriers to seeking mental health treatment is a complex area of study, with many variables to consider. Stigma and acculturation are key variables in predicting attitudes towards seeking mental health treatment. Particularly, stigma mediates the relationship between acculturation and attitudes towards seeking mental health treatment suggesting that researchers consider measuring stigma when studying cultural variables, within the context of treatment seeking. The model was significant even within a predominantly White, female, and non-first generation college sample, lending support to previous results that stigma is widespread problem. Although the nature of some of relationships between the different types of stigma and acculturation was not always clear, they were significantly associated with one and other, which warrants further exploration.
Another key issue to consider when examining mental health factors is the measurement of stigma, acculturation, ethnic identity, and attitudes towards treatment seeking. It is crucial to assess the psychometric properties of these measures along gender and race/ethnicity. There was heterogeneity with the relationship of acculturation with variables study variables. However, a stronger relationship to a cultural group (either mainstream or heritage) appears to be buffer in experiencing race-related stressors, which were associated with symptoms of distress, regardless of race/ethnicity. A strong relationship with heritage culture and a high ethnic identity may not necessarily be risk factors for lower attitudes of seeking mental health treatment; they may potentially serve as protective factors. The results also lend support to measuring cultural variables in various ways, including heritage acculturation, ethnic identity, and race-related stress, given the different relationships that the cultural variables had with stigma and attitudes and intentions towards seeking treatment. The literature indicates that ethnic identity is a component of acculturation.
The VIA-heritage subscale measures behavioral components of acculturation and the MEIM-R measures the strength of one's relationship and sense of belongingness to his/her ethnic group.
Furthermore, cultural stigma, examining an individual's cultural group's view of mental health treatment shows promise for further exploration, given the significance of this factor for Nonwhite participants, in particular. In contrast, selfstigma did not contribute significantly to attitudes towards helps seeking and was not significantly related to any study variables. Given the relatively healthy sample in this study, there may not yet have been an internalization of public stigma. In designing interventions to address stigma, public stigma and personal stigma should be directly focused on more prominently, particularly among healthy young adult samples, who may not yet have developed self-stigma.
Another area of further study is the measurement of help-seeking for males may in particular need to incorporate issues unique to men, such as masculinity, since the literature indicates that men are less likely to seek help. Overall, the results in this study support further exploration of cultural factors, in relation to stigma, as part of the decision making process to seek mental health treatment. 1st generation (born outside the United States and came as an adult) 1.5 generation (immigrated to the United States by age 13) 2nd generation (born in the United States, at least one parent was not born here) 3rd generation (born in the United States, at least one parent was born here) 4th+ generation (born in the United States, one or more grandparents were also born here)

Help-seeking History/Intention
Please answer the following questions as honestly as possible; you will not be judged for the answers.
14. Thinking about your physical health, which includes physical illness and injury, for how many days during the past 30 days was your physical health not Below is a list of people who you might seek help or advice from if you were experiencing a personal or emotional problem.
Please circle the number that shows how likely is it that you would seek help from each of these people for a personal or emotional problem during the next 4 weeks? 1 = Extremely Unlikely 3 = Unlikely 5 = Likely 7 = Extremely Likely a. Partner (e.g., boyfriend or girlfriend) 1 2 3 4 5 6 7 b. Friend (not related to you) 1 2 3 4 5 6 7 c. Parent 1 2 3 4 5 6 7 d. Other relative/family member 1 2 3 4 5 6 7 e. Mental health professional (e.g., psychologist, social worker, counselor) 1 2 3 4 5 6 7 f. Phone helpline 1 2 3 4 5 6 7 g. Doctor/GP 1 2 3 4 5 6 7 h. Professor/Academic Advisor i. Minister or religious leader (e.g., Priest, Rabbi, Imam, Chaplain) 1 2 3 4 5 6 7 j. Chat rooms 1 2 3 4 5 6 7 k. I would not seek help from anyone 1 2 3 4 5 6 7 l. I would seek help from another not listed above (please list in the space provided, (e.g., work colleague. If no, leave blank)_________________________1 2 3 4 5 6 7 Prior Below is a list of people who you might seek help or advice from if you were experiencing a personal or emotional problem. Tick any of these who you have gone to for advice or help in the past 2 weeks for a personal or emotional problem and if you have ever gone for advice or help (in the lifetime category).
Past 2  j. Chat rooms k. I would not seek help from anyone l. I would seek help from another not listed above (please list in the space provided, (e.g., work colleague. If no, leave blank)______________________________________

Depression Anxiety and Stress Scale-21 (DASS-21; Lovibond & Lovibond, 1995)
Please read each statement and mark the number that indicates how much the statement applied to you over the past week. There are no right or wrong answers. Do not spend too much time on any statement.
The rating scale is as follows: 0 Did not apply to me at all 1 Applied to me to some degree, or some of the time 2 Applied to me to a considerable degree, or a good part of the time 3 Applied to me very much, or most of the time 1 I found it hard to wind down 0 1 2 3 2 I was aware of dryness in my mouth 0 1 2 3 3 I couldn't seem to experience any positive feeling at all 0 1 2 3 4 I experienced breathing difficulty (eg, excessively rapid breathing, breathlessness in the absence of physical exertion) 0 1 2 3 5 I found it difficult to work up the initiative to do things 0 1 2 3 6 I tended to over-react to situations 0 1 2 3 7 I experienced trembling 0 1 2 3 8 I felt that I was using a lot of nervous energy 0 1 2 3 9 I was worried about situations in which I might panic and make a fool of myself 0 1 2 3 10 I felt that I had nothing to look forward to 0 1 2 3 11 I found myself getting agitated 0 1 2 3

Reminder of rating scale:
0 Did not apply to me at all 1 Applied to me to some degree, or some of the time 2 Applied to me to a considerable degree, or a good part of the time 3 Applied to me very much, or most of the time 12 I found it difficult to relax 0 1 2 3 13 I felt down-hearted and blue 0 1 2 3 14 I was intolerant of anything that kept me from getting on with what I was doing 0 1 2 3 15 I felt I was close to panic 0 1 2 3 16 I was unable to become enthusiastic about anything 0 1 2 3 17 I felt that I wasn't much as a person 0 1 2 3 18 I felt that I was rather touchy 0 1 2 3 19 I was aware of the action of my heart in the absence of physical exertion (e.g., sense of heart rate increase, heart missing a beat) 0 1 2 3 20 I felt scared without any good reason 0 1 2 3 21 I felt that life was meaningless 0 1 2 3

Perceived public stigma
Please indicate whether you agree or disagree with the following statements.
1. Most people would willingly accept someone who has received mental health treatment as a close friend. 11. Most young adults would be reluctant to date someone who has been hospitalized for a serious mental disorder.* 0 1 2 3 4 5 12. Once they know a person has received mental health treatment, most people will take that person's opinions less seriously.* 0 1 2 3 4 5

Personal stigma
Please indicate whether you agree or disagree with the following statements.
1. I would willingly accept someone who has received mental health treatment as a close friend. 0 1 2 3 4 5 2. I would think less of a person who has received mental health treatment.* 3. I believe that someone who has received mental health treatment is just as trustworthy as the average person. 0 1 2 3 4 5 * indicates reverse scored items