| Abstract: | Recent reviews of the absenteeism literature indicate that absenteeism as a phenomenon is still neither well understood nor accurately predicted. The present study incorporated many of the suggestions for improving absenteeism research that were made in the reviews. More specifically, a longitudinal design was used to assess the accuracy with which the 1977 absenteeism (i.e., total days absent, absence frequency, and supervisory absenteeism rating) of 112 research scientists could be predicted from their previous absenteeism (1974–1976). In addition, the relationships between 1977 absenteeism and 3 work attitudes (job satisfaction, job involvement, and supervisory satisfaction) were assessed. Findings show that past absenteeism was a better predictor of 1977 absenteeism than were the 3 work attitudes. Data relevant to the psychometric equivalence and stability of the absenteeism measures are also reported. (14 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved) |
