Influence of cumulation strategies on the long-range prediction of absenteeism

The ability of three different types of self-report variables-biographical, attitudinal, and stress-to predict 70 months of absenteeism data was evaluated. Absence scores were aggregated in two different ways, as annualized segments and as longitudinal data cumulations. Biographical variables and attitudinal measures significantly predicted long-term absenteeism. In addition, results suggested that dependency relationships between absence metrics and cumulation periods may affect validational outcomes.