A Confirmatory Factor Analysis of the Hopkins Symptom Checklist with Ghanaian Female Survivors of Trafficking

Background: This study examined the factor structure of the 58-item Hopkins Symptom Checklist (HSCL) scale (Derogatis et al., 1974) with a sample of Ghanaian female victims of trafficking. The scale is scored on five underlying dimensions: Somatization, Obsessive-Compulsive, Interpersonal Sensitivity, Depression and Anxiety. This scale validation and cultural adaptation is part of a large study whose goal is to create evidence-informed reintegration programs for female victims of trafficking. The scale has recently demonstrated high reliability among Ugandan female care-givers living with HIV (Familiar et al., 2016).

Methods: Cross-sectional data were collected in April 2016 from a sample of n=144 female victims of trafficking. Participants had received psychosocial and human capital development services in a local residential facility between 2010 and 2015. Research participants were recruited through the service providers contacts.

Descriptives:  The age range of the participants was 23-34, Mn=23 (SD=2.89). In terms of education, 86(60%) had secondary school education, 43(30%) elementary, 11(7.6%) no formal education, and just one participant had some post-secondary education. Seven (5%) participants already had one or more children at the time of admission into the facility and this number increased to 42(29%) at the time of the study. Twenty-five (17%) of the participants did not have a mother who was alive and of these, 16(70%) lost their mothers before joining the facility. Sixty-three 63(43%) had lost their father by the time of the interview and of these, 45 (74%) had lost their father before joining the facility. The average number of siblings for the participants was M=4 (SD=2.53).

Analysis: Three CFA models were assessed using maximum likelihood methods with AMOS 22.0. In keeping with past studies, each model had five structures. The first model evaluated the original 58-item version of the HSCL. The factor loadings for each item were examined and assigned to the factor on which they had the largest loading. Items that did not meet a cutoff point (≤.5) were not retained in the second structure which retained 35 items. Lastly, items that did not meet a cutoff point (≤.6) were not retained in the third and final structure, with 15 items being retained for this final model.

Results: The results were consistent with recent factor analysis by Familiar et al., (2016) and revealed that the reduced 15-item form of the HSCL scale with a five factor underlying structure was the most parsimonious model, [χ²= 145.693(df=84), CFI=.924, TLI=.891, RMSEA(90%CI) =.072(.052-.091)]. Cronbach’s α showed high reliability: Somatization = .865, Obsessive-Compulsive = .783, Interpersonal Sensitivity = .780, Depression = .771, and Anxiety = .777. 

Implications: The findings suggest that less items are needed to capture symptom configurations for this population. Given that human trafficking is a global challenge, social workers and others who want to assess the types and forms of symptoms among women who have experienced sexual and/or labor trafficking must take into account the cultural orientations in different countries. Assessing the types of symptoms for trafficking survivors is essential for social workers in both research and clinical contexts.