Public school teachers' use of sick leave due to mental illness: Multilevel structural equation modeling using individual and prefecture-level data

Abstract

The present study examined a causal model of public school teachers' use of sick leave due to mental illness through multilevel structural equation modeling, using scores on teachers' perception of their work environment and stress responses from an attitude survey as within-level data, and government statistics on prefectural sickleave rates, educational administration, and mental health measures for teachers as between-level data. For both models, that is, models for elementary and for junior high school teachers, within-level models demonstrated that the teachers' perception of their work environment explained their stress response, according to a traditional occupational stress model. Between-level models showed that the teachers' attitudes mediated the relation between educational administration and mental health measures on the one hand, and the teachers' sick leave usage on the other. Common or similar factors in both the elementary and junior high school models included the extent of use of non-regular teachers, the number of teachers or number of boards of education per student, the unionization rate, and the number of schools. The results suggest that even over a wide area such as a prefecture, organizational adjustment of teachers' environment and other measures could assist in the prevention of stress responses and decrease the use of sick leave due to mental illness.