Abstract
The Internal Control Index (ICI) is a 28-item measure of locus of control. The present study utilized item response theory (IRT) to determine if the length of the instrument can be reduced to improve its psychometric properties. Students at CSU Sacramento (N = 631) completed the ICI for course credit. When the scale was reduced to 11 items, the graded response model demonstrated a good fit, M2(484) = 823.34, p < .001, RMSEA = .03. Because summed scores were strongly linearly related to IRT scale scores (θ), r(596) = .997, p < .001, they were considered to be a good approximation of locus of control. Additionally, summed scores on the original and new scoring strategies correlated highly, r(596) = .82, p < .001, and both versions correlated similarly with related constructs. It therefore appears that the ICI-R may provide an equally valid and more efficient measure of locus of control.