Longitudinal relationship between teacher self-efficacy and work engagement: Testing the random-intercept cross-lagged panel model

Even though a longitudinal relationship between self-efficacy and work engagement has been investigated rather extensively, previous studies rarely considered their dynamic nature by separating stable trait-like effects from fluctuating state-like ones. In the present three-wave longitudinal study involving 3010 teachers (82% women), the random-intercept cross-lagged panel modeling (RI-CLPM) approach was used to explore whether: (1) stable parts of TSE and work engagement are related to each other, (2) higher than expected levels of TSE are associated with higher than expected levels of work engagement at a given time point, and (3) higher than expected levels of TSE at a given time point are associated with future higher than expected levels of work engagement. Results showed a positive correlation between stable parts of TSE and work engagement, a positive time-specific correlation between TSE and work engagement, and spill-over effects from work engagement to TSE implying that teachers with higher than usual levels of work engagement, also experience higher than usual levels of TSE at subsequent assessment. Unlike previous findings and theoretical expectations, reciprocal relationship was not confirmed as only work engagement predicted TSE and not vice versa.