AN EXAMINATION OF THE PSYCHOMETRIC PROPERTIES AND NORMOLOGICAL VALIDITY OF SOME REVISED AND REDUCED SUBSTITUTES FOR LEADERSHIP SCALES

Confirmatory factor analysis in two large samples (N = 411 and N = 1,235) was conducted to examine the convergent and discriminant validity of the 74-item (revised) and 41-item (reduced) versions of the substitutes for leadership scales recently developed by Podsakoff, Niehoff, MacKenzie, and Williams (1993), and Podsakoff, MacKenzie, and Fetter (1993). Following this, the reliabilities and subscale intercorrelations of the two versions of the scale were compared in order to determine how faithfully the 41-item version represents the 74-item scale. Next, the reliabilities of both versions were compared with the reliability of Kerr and Jermier's (1978) original scale, and their nomological validity was evaluated. The overall pattern of results indicated that both versions were reasonably reliable and valid, and both were better than Kerr and Jermier's original scale.