Evaluating the factor structure of the General Self-Efficacy Scale

Abstract

Although the Sherer General Self-Efficacy Scale has often been used in clinical, personality, and organisational research, one of the major issues surrounding this instrument is that of an appropriate factor structure. In an endeavour to address the criticism levelled against this scale, this article draws on a study aimed at investigating the psychometric properties of the Sherer General Self-Efficacy Scale with respect to both its factor structure and estimates of reliability. A quantitative, cross-sectional research design with convenience sampling was used. A total of 295 aspiring chartered accountants who had sat one of their compulsory examinations completed the Sherer General Self-Efficacy Scale. Both the minimum average partial test and parallel analysis suggested that a unidimensional structure be investigated. Confirmatory factor analysis was used to compare two competing measurement models representing a three-dimensional factor structure and a unidimensional factor structure. Both models exhibited fairly similar levels of fit. To break this impasse, the Schmid–Leiman solution provided evidence that the Sherer General Self-Efficacy Scale consisted of a strong general factor which explained 76% of the variance. This study therefore concluded that general self-efficacy, as measured by the Sherer General Self-Efficacy Scale, may be treated as a unidimensional construct.