TPC Journal-Vol 10- Issue 1

6 The Professional Counselor | Volume 10, Issue 1 = 1.25. In addition, the CFI = 0.98, GFI = 0.96, IFI = 0.98, TLI = 0.98, RMSEA = 0.03, 90% confidence interval [<.00, .06], and SRMR = 0.05 also demonstrated a strong model fit. Internal consistency reliability analyses (Cronbach’s coefficient alpha) revealed satisfactory reliability coefficients for an attitudinal measure, Diminish/Avoid (α = 0.73) and Approach/Encourage (α = 0.70). In addition, the path model coefficient (-0.04) between factors supported the structural validity of the scales (see Figure 1). Figure 1 Confirmatory Factor Analysis Path Diagram for the Mental Distress Response Scale Multivariate Analysis A 2 X 2 (gender X help-seeking history) MANOVA was computed to investigate demographic differences in faculty members’ responses to encountering a student in mental distress (research question #2). G*Power was used to conduct an a priori power analysis (Faul et al., 2007) and revealed that a minimum sample size of 151 would provide a 95% power estimate, α = .05, with a moderate effect size, F 2 (v) = 0.063. A significant main effect emerged for gender: F (3, 196) = 8.27, p < 0.001, Wilks’ λ = 0.92, 2 p η = 0.08. The MANOVA was followed up with a post hoc discriminant analysis based on the recommendations of Warne (2014). The discriminant function significantly discriminated between groups: Wilks’ λ = 0.91, X 2 = 18.85, df = 2, p < 0.001. The correlations between the latent factors and discriminant function showed that Diminish/Avoid loaded more strongly on the function ( r = 0.98)