Dataset for the Study \"What predicts changes in perceived teacher work ability?\"

The dataset accompanying the study “What Predicts Changes in Perceived Teacher Work Ability?” contains longitudinal survey data from 469 primary and lower secondary school teachers across 44 schools. It comprises a comprehensive set of 23 variables representing antecedents and outcomes of perceived teacher work ability (PTWA). The dataset includes indicators of job demands (quantitative, cognitive, and emotional), job resources (supervisor support, coworker support, and sense of community at work), and personal resources (teacher self-efficacy). In line with the JD-R model, it also incorporates proximal outcomes reflecting both the health-impairment and motivational pathways—namely burnout, self-rated physical and mental health, and work engagement. Additionally, a range of sociodemographic, occupational, and life-related characteristics is included, such as age, gender, teaching experience, educational level taught, class or mentor teacher role, caregiving responsibilities, and work–life conflict. The dataset was generated within a two-stage analytic framework that integrates machine learning and structural equation modeling (SEM) to examine predictors of change in PTWA over time.