THE LEVENSON LOCUS OF CONTROL SCALES: CONFIRMATORY FACTOR ANALYSES AND EVALUATION

Abstract

Confirmatory factor analyses were conducted to test the factor structure of several versions of Levenson's (1973) locus of control scales. Two- and three-factor models based on all 24 of Levenson's items and on 20 of her items were tested. The 3-factor models provided a good fit. Models proposed by R. M. Shewchuk, G. A. Foelker Jr., and G. Niederehe (1990) and R. M. Shewchuk, G. A Foelker Jr., C. J. Camp, and F. Blanchard-Fields (1992) also provided a good fit of the data. In concurrent and prospective tests of the predictive ability of the various models, the 24 and 20 item versions of Levenson's models accounted for a significant amount of variance In depressive symptomatology. The three-factor models revealed that only scores on the chance scale reliably predicted time 2 depressive symptomatology. Neither of the models proposed by Shewchuk and colleagues accounted for a significant amount of variance.