Assessing Organizational Behavior Models: A Comparison of Linear and Nonlinear Methods1

This study highlights the use of nonlinear statistic techniques in a behavioral model that incorporates dispositional and situational variables. When ordinary least squares and log-linear/logit results were compared, various unique associations were found between job satisfaction, job complexity, locus of control, and the number of available benefits. Overall, the findings prove that more nonlinear modeling is needed in the behavioral sciences.