Search

Examined the influence of deviant scale items on item parameter estimates of focal scale items and person parameter estimates through a comparison of item response theory (IRT) and classical test theory (CTT) models. Used Monte Carlo methods to explore results from a pilot investigation of job attitude data. Discusses implications for researchers using IRT and CCT methods in scale development. (SLD)

A decoy effect occurs when the addition of an inferior choice alternative changes the preference relations among the other alternatives in the set. Whereas many authors have suggested that the decoy effect has considerable relevance to applied decision-making contexts, others have suggested that the phenomenon may not be found in situations more realistic than those used in typical decoy experiments. The authors examined whether the effect would generalize to situations in which decision makers are required to infer attribute values. The results of the current study showed that even when no numerical attribute information is presented, decoy effects can be quite strong. These findings suggest that decoy effects may indeed be generalizable to realistic choice situations requiring attribute-level inferences.

Many authors have suggested that situational judgment tests (SJTs) are useful tools for assessing applicants because SJT items can be written to assess a number of job-related knowledges, skills, abilities and other characteristics (KSAOs). However, SJTs may not be appropriate for measuring certain KSAOs for some applicants. We posit that using SJTs to measure interpersonal skills may lead to invalid inferences about applicants with higher levels of angry hostility (AH), and thus, AH should moderate the relation between interpersonally oriented SJTs and job performance. Three studies, using samples of healthcare workers ( n = 225), police officers ( n = 54), and medical doctors ( n = 92), provided support for hypotheses in that that relations between SJT scores and performance criteria were significantly weaker among employees higher in AH compared to those lower in AH. In addition, none of the other facets of neuroticism tested (self-consciousness, anxiety, depression, immoderation, or vulnerability to stress) consistently moderated SJT validity, providing support for the uniqueness of AH. Implications for practice, and for future research studying the relations between interpersonal skills as measured by SJTs and job performance, are discussed. [ABSTRACT FROM AUTHOR]

The article describes the Internet marketplace Mechanical Turk (MT) from Amazon and discusses its potential role in data collection and diverse research initiatives in the field of industrial and organizational psychology. It provides a brief overview on the history of MT. It explains the characteristics of MT workers and their reasons for participation. It also explores the rules for setting the reward amount for human intelligence tasks.