Introduction

Among individuals who are homeless or vulnerably housed (HVH), most research has focused on efforts to make health and housing resources more available and accessible, with only recent interest in the role of self-efficacy (one’s sense of personal competence) for this group. The 8-item Perceived Health Competence Scale (PHCS) and 7-item Housing SE Scale (HSES) measure health and housing specific SE, respectively. Each scale uses a total score, implying the scales are at least essentially unidimensional. The factor structures of these scales, however, have yet to be verified with a HVH sample.

Objectives

The study purpose was to examine the factor structure and reliability of scores of the PHCS and HSES with a sample of 333 HVH men and women.

Design/Methodology

Data were obtained from the Health and Housing in Transition (HHiT) study. Because the scales use 5-point Likert-type response formats, a one-factor confirmatory factor analysis (CFA) model was tested with each scale using a robust ordinal estimator and fit was determined using CFI and RMSEA (with 90% confidence intervals). With a lack of fit, exploratory factor analysis (EFA) and parallel analysis, based on the polychoric matrix, were employed to verify the number of factors. Internal consistency was examined using ordinal alpha.

Results

Results showed that a one-factor CFA model for strict unidimensionality did not fit the data for either scale. For each scale, EFA showed extra covariance that suggested a possible minor (negative) wording factor, but parallel analysis clearly supported an essentially unidimensional factor structure. Ordinal alpha was .87 for the PHCS and .85 for the HSES.

Conclusions

This study found, that despite some minor effects of (negative) wording, both the PHCS and HSES were essentially unidimensional. This supports using a total score for each scale. Moreover, reliability estimates for both scales were satisfactory for this HVH sample.