Construct validities and the empirical relationships between optimism, hope, self-efficacy, and locus of control.

This study assessed the construct validities and the relationships between Optimism [21], Hope [27], Self-Efficacy [18], and Locus of Control [12]. The validities and relationships between these scales were examined not only because they are to some degree rival constructs but also because there is an outstanding theoretical question in the literature of whether each of these variables is a state or trait variable or both. Synder' Hope scale was the "know marker" trait-state scale in this study. Data were obtained on all scales from 78 at-risk university students and 22 regular (or normal) students, as all scales either make claims about or have existing data on these two different types of subjects. Other background data were also collected. A Manova on the 13 variables for which data were obtained found significant profile differences between the two groups of subjects as theory and the literature predicted at the 0.01 level. The trait-state two factor structure of Synder's Hope scale was found except the structure was orthogonal in at-risk students as opposed to the oblique structure Synder found in normals. The same strong results were obtained for Schiever and Carver's Optimism scale which additionally resolved an outstanding interpetability issue with this scale. Hypothesized factor structures were not obtained for the self-efficacy or the locus of control scales and both scales best fit the trait-sate model, which contradicts Bandura's and other prevailing view of these two variables, but supports the results found by Bandalos et al. [2].