A factor analysis study of a test of human relations ability

This study was conducted as the initial undertaking in determining the construct validity of forms. XL and XM of Supervisors' Problems, a test of human relations ability. Seventy-three industrial-organizational psychology students completed both forms of Supervisors' Problems and a battery of tests including several factorially pure reference tests, the Eysenck Personality Inventory and the Leadership Opinion Questionnaire. The scores on the instruments were intercorrelated and factor analyzed. The initial results indicated the two forms of Supervisors' Problems measured different constructs. The variance in form XL was accounted for by three factors-reasoning, leadership, and lie. Variance in form XM was accounted for by only one factor-leadership. These results suggested that the items in forms XL and XM were measuring different constructs. Forms XL and XM were rescored to yield a reasoning and leadership subscore. These scores and the reference test scores were intercorrelated and factor analyzed. The correlations among the reasoning and leadership subscores indicated that the reasoning subscores marginally met the criterion for convergent validity and failed to meet the criteria for discriminant validity, while the leadership subscores met the criteria for both convergent and discriminant validity. The factor analysis revealed that variance in the reasoning subscore was accounted for by a Supervisors' Problems specific factor, and that variance in the leadership subscore was accounted for by a Supervisors' Problems specific factor, a reasoning factor, and a laissez-faire leadership factor. Suggestions for future study of Supervisors' Problems were made.Psychology, Department o