Teacher Work Engagement: A Psychometric Investigation

Year: 2017

Author: Perera, Harsha, Granziera, Helena

Type of paper: Abstract refereed

Background. Recent years have witnessed an increasing interest in teachers' work engagement. This interest has been attributed to accumulating evidence demonstrating the relations of teachers' beliefs, feelings, and behaviors with students' academic and non-academic outcomes. At least another reason for increased attention is the increasing turnover of teachers in several education systems. As well as placing pressure on the teaching pipeline to meet expected growth in student numbers, teacher attrition has educational and social costs, not least the loss of expertise in, and destabilization of, schools that may undermine student outcomes. It is in this climate that researchers and policy-makers have turned their attention to teacher engagement as an aspect of teacher effectiveness that may inform understandings of why and when teachers leave the profession. Notwithstanding accruing evidence supporting the role of teachers' work engagement in both teacher and student outcomes, extant work is limited by a dearth of clarity about the theoretical and empirical structure of the teacher engagement construct, which complicates investigations of the construct and its relation to substantively meaningful antecedents and outcomes.

Aims. The primary purpose of this study was to examine the dimensionality of teacher work engagement data using a novel framework designed to account for multiple sources of construct-relevant multidimensionality. In addition, measurement and structural invariance across teaching-level, latent mean differences across gender, longitudinal invariance, stability, and cross-lagged relations among the engagement dimensions, and criterion-validity with respect to work satisfaction were examined.

Sample. The sample comprised 594 school teachers from multiple Australian states and territories.

Method. Teachers completed a short, online battery of psychometric measures assessing their engagement with and job satisfaction. General structural equation modelling techniques, including exploratory structural equation modelling (ESEM), were used to analyse the data.

Results. Results supported a bifactor-ESEM representation and its full measurement invariance over teaching level; however, meaningful differences in the latent engagement means were obtained, such that primary teachers reported higher general engagement and specific engagement with students than secondary teachers. Evidence was also obtained for meaningful gender differences; female teachers across teaching levels exhibited lower emotional engagement. Evidence of longitudinal measurement invariance, theoretically meaningful covariance stability, and cross-lagged effects of engagement dimensions, and criterion validity of the engagement scores with respect to job satisfaction was obtained.

Conclusions. The study yields crucial evidence clarifying the structure of engagement and provides a tentative answer to the unresolved dimensionality question raised in the broader work engagement literature.