| Utility analysis is a technique developed and incorporated by I/O psychologists to assess the effectiveness of organizational interventions. Most of the previous utility models (Brogden, 1946, 1949; Cronbach & Gleser, 1965; Raju, Burke, & Normand, 1990) were developed based on the assumption that the criterion is a continuous variable. However, in practice, many organizations use categorical or dichotomous scales to assess employees' job performance (i.e., a criterion). Hence, the assumption of criterion continuity may not be robust at times. Raju, Cabrera, and Lezotte (1996) developed a dichotomous utility analysis model to address part of this concern. The dichotomous utility model has been empirically tested and found to be applicable for the situation where job performance is scaled into two categories.;Expanding Raju et al.'s (1996) dichotomous utility model, the author proposed a polytomous utility analysis model in this study based on a similar logistic regression approach. This proposed model can be tailored to fit the situation where the criterion is categorically scaled into three or more graded responses. It was found that the polytomous utility model was applicable for assessing the utility for a 4-point job performance criterion with no obvious error made and high degree of confidence. The utility assessed based on this model also was found to be comparable to the utilities assessed based on other models. |
