Abstract
The longitudinal models were based on a well-fitting CFA with four latent factors, Individual Factors, Positive Workplace Factors, Overall Well-Being and Work Engagement. Composite variables, derived from factor score weights, were used at Times 1, 2 and 3 (n=198). Non-nested longitudinal models were compared as Stability, Causality, Reverse Causality and Reciprocal models. Whilst the Reciprocal was the best fitting model, removing trivial pathways improved fit and parsimony; Trimmed Reciprocal model, X2(29) = 32.7, CFI = .999, RMSEA = .025. The remaining paths indicate strong influences of concurrent functioning, stability of constructs over time and weaker reciprocal relationships between constructs across time. Individuals with greater confidence in themselves and the future (as optimism and self-efficacy), experience better work conditions, have greater wellbeing and work engagement. These day-to-day influences are equalled by the long term strength and stability of Individual Factors, Positive Workplace Factors, and Overall Well-Being. Whilst the reciprocal paths had only weak to mild effects, there was mutual reinforcement of Individual Factors and Overall Well-Being, with Positive Workplace Factors and Work Engagement counterbalancing each other, indicating a more complex relationship. Well-being, particularly, is anchored in the immediate and distant past and provides a robust stability to functioning into the future.