ABSTRACT
Expectancy for locus of control of reinforcement (LOC) was originally conceived as being a unidimensional construct. Results from recent studies employing samples from varied populations have generally shown LOC to be multidimensional, that there are two distinct types of externals: those who perceive reinforcement as determined by chance factors and those who perceive reinforcement as controlled by powerful others. The purpose of the present study was to investigate the dimensionality of the LOC construct in a prison sample. Given the nature of control in prisons, the dimensionality of LOC was expected to show variation. The Levenson (1972) multidimensional LOC measure was administered to a sample of volunteers (n = 174) from randomly selected inmates at a state prison for women in the Western United States. Unlike past factor analytic studies which showed items to load on three factors, factor analyses of data obtained from the present sample showed the presence of one factor, the internal-external dimension. Furthermore, the dimensionality of LOC was unrelated to length of imprisonment. Homogeneity of sex, race, and socioeconomic status perhaps best explain results. Implications for prisoner rehabilitation involving manipulation of inmate perception of control for reinforcement are discussed.
