Abstract
The general unavailability of and difficulty associated with obtaining records-based data on absenteeism suggests the potential value of self-report data for those conducting research on absenteeism. This should not be recommended, however, until the validity of these self-report measures is assessed. In this paper, we compare records-based and self-report measures of absenteeism for the same employees for the same period of time. We find that although the univariate descriptive data for the two measures are similar, the correlation between the two is .299. Although this is in the expected direction, its magnitude is small enough for us to question the validity of self- report measures of absenteeism. Several suggestions are offered for additional study of this.
