TPC Journal-Vol 10- Issue 2-FULL ISSUE

226 The Professional Counselor | Volume 10, Issue 2 & Fidell, 2007). Table 2 also includes the means and standard deviations for the variables in this model. Various guidelines were reviewed as a means for determining the appropriate sample size for this investigation. Jackson (2003) and Kline (2005) stated that a 20:1 ratio of sample size to parameters is preferable, and our current study exceeded this recommendation. Table 2 Correlations, Means, and Standard Deviations for the Variables in the Path Analysis Variables 1 2 3 4 5 6 1. Self-Stigma of Mental Illness - 2. Self-Stigma of Help Seeking .54 ** - 3. Attitudes Toward Help Seeking -.28 ** -.62 - 4. Drug Use -.02 .11 -.16 * - 5. Alcohol Use -.01 .12 -.14 * .68 ** - 6. Age .08 -.01 .05 -.27 -.26 ** - M(SD) 3.23(.89) 2.74(.81) 1.71(.61) 4.31(5.07) 9.79(8.01) 34.39(9.99) Note . Measures used in this study include the Self-Stigma of Mental Illness Scale (SSOMI; Tucker et al., 2013), the Self- Stigma of Help Seeking Scale (SSOHS; Vogel et al., 2006), Attitudes Toward Seeking Professional Psychological Help – Short Form (ATSPPH-SF; Fischer & Farina, 1995), the Alcohol Use Disorders Identification Test (AUDIT; Saunders et al., 1993), and Drug Abuse Screening Test-20 (DAST-20; Skinner & Goldberg, 1986). * = p < .01, ** = p < .001. We examined the variables in this study (i.e., self-stigma of mental illness and help-seeking, attitudes toward help-seeking, and drug and alcohol use) to evaluate for potential control variables. Specifically, we conducted several correlations comparing the variables in this study with demographic characteristics. For dichotomous variables, we utilized point-biserial correlations. These analyses indicated that age had significant relationships with both drug and alcohol use; thus, we included age in the path analysis as a control variable. Model Specifications The a priori hypothesized model tested in this path analysis included a total of six observed variables that were placed in a causal directional structure that we developed from our understanding of the literature. The exogenous variables included self-stigma of mental illness (as measured by the SSOMI; Tucker et al., 2013) and self-stigma of help seeking (as measured by the SSOHS; Vogel et al., 2006). In addition, attitude toward help-seeking (as measured by the ATSPPH-SF; Fischer & Farina, 1995) was both an exogenous and endogenous variable. Lastly, alcohol use (as measured by the AUDIT; Saunders et al., 1993) and drug use (as measured by the DAST-20; Skinner & Goldberg, 1986) and were endogenous. We correlated self-stigma of mental illness and self-stigma of help seeking along with the error terms for alcohol and drug use. In the model, we examined the direct effect of self-stigma of mental illness and self-stigma of help seeking on attitudes toward help-seeking. Furthermore, we examined the direct effect of attitudes toward help seeking on drug and alcohol use. We included age in this model as a control