Key points are not available for this paper at this time.
Summary Two previous articles (Lawshe Lawshe, Bolda, Brune, & Auclair, 1958) in Personnel Psychology have discussed expectancy charts as methods of displaying data on predictor‐criterion relationships. The second of these articles (Lawshe, et al. , 1958) dealt with charts based upon theoretical in place of empirical expectancies. A method, based on chi square, of testing the discrepancies between empirical and theoretical expectancies was described. The present article applies this test to 15 independent sets of predictor‐criterion data. In every case, the use of theoretical expectancies based on an assumed bivariate normal distribution was found to be justified.
Tiffin et al. (Tue,) studied this question.