Dimensionality, cross-validation and reliability of Help-Seeking Dilemma Scale

Help-seeking dilemmas are potential barriers to use of treatment, social interventions, and social support systems. While previously existing scales provided a means of assessing help-seeking intentions, a fundamental gap is their limitation to mental health cases or gender lines. Hence, Help-Seeking Dilemma Scale (HSDS) was developed for assessing help-seeking dilemmas regardless of mental health conditions or gender lines. The study adopted cross-validation design with 660 participants across three samples. First sample comprised 15 participants (males = 7, females = 8) in an in-depth interview that facilitated the generation of HSDS items. Second sample comprised 446 participants (male = 49.6%, female = 50.0%, unspecified gender = .4%, mean age = 29 ± 11.31 years) randomly split into EFA (n = 227) and CFA (n = 219) subsamples. HSDS was cross-validated with Exploratory Factor Analysis (EFA) in the EFA subsample and Confirmatory Factor Analysis (CFA) in the CFA subsample. Third sample comprised 199 participants (male = 85%, female = 11.5%, unspecified gender = 3.5%, mean age = 31.29 ± 6.55 years) where predictive validity was tested. Data were collected using HSDS, General Help Seeking Questionnaire (GHSQ), Self-stigma of Seeking Help Scale (SSOSH) and Actual Help-Seeking Questionnaire. EFA revealed sampling adequacy and data appropriateness for factorization [KMO = .910, χ2(231) = 2241.50, p < .0001] and extracted five dimensions, accounting for 62.4% of total variance. CFA confirmed the five-dimensional model of the EFA, with acceptable model fit indices: χ2(160) = 321.041, p < 0.001; CMNI/DF = 2.01; RMSEA = .07 (95% CI: .06 – .08). All HSDS dimensions showed convergent validity. Fornell-Larcker Criterion and Heterotrait-Monotrait (HTMT) ratio confirmed dimensional discriminant validity. Predictive validity showed that HSDS significantly predicted actual help seeking [OR = 1.19, p < 0.001]. Pearson’s correlations confirmed discriminant validity between HSDS and SSOSH, r = .14, p < .01, and GHSQ, r = .05, p > .05. All dimensions demonstrated good internal consistency, with reliability score ranging from .79 to .88. Overall internal consistency for HSDS was .95. HSDS is valid and reliable for assessing help-seeking dilemmas in academic and general populations and without limitation to any gender.