The Professional Counselor | Volume 12, Issue 3 253 counseling due to feelings of embarrassment or shame (e.g., “It would damage my reputation”). Scores on the Value subscale range from 4 to 20, with higher scores indicating a greater disinclination to seek counseling because they believe the effort required would not be worth the potential benefits (e.g., “Counseling is unnecessary because my problems will resolve naturally”). The Global Barriers to Counseling scale is composed of test takers’ average composite score across the three Fit, Stigma, and Value subscales and produces an overall estimation of a test taker’s sensitivity to barriers toward seeking counseling. Scores on the Global Barriers to Counseling scale range from 13 to 65, with higher scores indicating a greater reticence to seek counseling. The collective findings of past investigators demonstrated evidence for the internal structure validity (confirmatory factor analysis) and internal consistency reliability (α = .70 to α = .91) of scores on the RFSV Scale with a number of noncollege populations (Kalkbrenner, Goodman-Scott, & Neukrug, 2020; Kalkbrenner & Neukrug, 2018, 2019; Kalkbrenner et al., 2019). Data Analysis A confirmatory factor analysis (CFA) based on structural equation modeling was computed in IBM SPSS AMOS version 26 to answer the first RQ about the dimensionality of STEM students’ RFSV scores. We used the joint suggestions from Dimitrov (2012) and Schreiber et al. (2006) for acceptable model fit in CFA: chi-square absolute fit index (CMIN; non-significant p-value or χ2 to df < 3), comparative fit index (CFI; .90 to .95 = acceptable fit and > .95 = close fit), root mean square error of approximation (RMSEA; ≤ .08), and the standardized root mean square residual (SRMR; ≤ .08). Internal consistency reliability evidence of test scores is another important step in testing a scale’s psychometric properties. Cronbach’s coefficient alpha (α) is the most popular internal consistency reliability estimate; however, its proper use is dependent on the data meeting several statistical assumptions (McNeish, 2018). Composite internal consistency reliability estimates, such as McDonald’s coefficient omega (ω), tend to produce more stable reliability estimates of scores. Accordingly, the present investigators computed both α and ω. College students are more likely to discuss mental health concerns with their peers than with faculty, staff, or other university personnel (Wawrzynski et al., 2011; Woodhead et al., 2021). Accordingly, college counseling researchers and practitioners are devoting more time to peer-to-peer mental health support initiatives with the goal of increasing peer-to-peer referrals to the counseling center (Kalkbrenner, Sink, & Smith, 2020; Olson et al., 2016). Past investigators (e.g., Kalkbrenner, Neukrug, & Esquivel, 2022) found that the RFSV barriers were significant predictors of peer-to-peer referrals to the counseling center with non-STEM students. To test the generalizability of this finding with STEM students, we conducted a logistic regression analysis to answer the second RQ regarding the capacity of STEM students’ RFSV scores to predict at least one peer referral to the counseling center. STEM students’ interval-level composite scores on the Fit, Stigma, and Value subscales were entered into the model as predictor variables. The criterion variable was quantified on a categorical scale. On the demographic questionnaire, students responded to the following question: “Have you ever referred (recommended) another student to counseling services?” and selected either “0 = never referred a peer to the counseling center” or “1 = referred one or more peers to the counseling center.” A 2(gender) X 3(race/ethnicity) X 2(help-seeking history) multivariate analysis of variance (MANOVA) was computed to investigate the third RQ regarding demographic differences in RFSV barriers among STEM students. The three categorical-level independent variables included gender (male or female), race/ethnicity (Latinx, White, or other ethnicity), and help-seeking history (never attended counseling or attended at least one counseling session). The three interval-level dependent variables included STEM students’ composite scores on the Fit, Stigma, and Value subscales. Discriminant analysis was employed as a post hoc test for MANOVA (Warne, 2014).
RkJQdWJsaXNoZXIy NDU5MTM1