Considerable criticism of the narrowness of teacher efficacy operationalisations argues they fail to tap the complexity of the teacher’s role. At the same time, we need more knowledge about how self-efficacies change for different kinds of teachers, dependent on their individual qualities and early career experiences, and consequences for their professional engagement. In this seminar, I will review the development and validation of our multidimensional SET (self-efficacy for teaching) scale1 , explore different patterns of change from teacher education through early career, identify associated demographics and early career experiences, and examine impacts on professional engagement. The SET extends the widely-used Teachers’ Sense of Efficacy Scale (Tschannen-Moran & Woolfolk Hoy, 2001), to build a comprehensive and ecologically-valid measure beyond their classroom work by developing new items to tap 5 additional competences: ‘Critical reflection on practice’, ‘Value cultural diversity’, ‘Professional interactions with the community’, ‘Respect confidentiality’ and ‘Meet legislated ethical requirements’.
Beginning teachers (N = 339) from four Australian universities completed the SET at 2 timepoints: near the end of teacher education, and again after on average 3 years teaching. Early career self-report measures included experiences of excessive demands, mentoring, collective-efficacy and belonging. Professional engagement was measured at both timepoints (effort, planned professional development, persistence). CFAs established psychometric validity and scalar invariance across timepoints. Five different change patterns were distinguished using hierarchical cluster analysis (Ward’s method) of change scores: “High competence”, “Moderate growth”, “High growth—longest timeframe”, “Declining competence” and “Declining competence—shortest timeframe”. Repeated-measures MANOVA compared clusters on trajectories which associated with distinct demographic/school contextual characteristics. High competence started out and retained high self-efficacies; highest-rated collective-efficacy/belonging T2 may have supported this continuity. By contrast, ‘declining’ clusters reported least supports. Moderate growth/High growth—longest timeframe reported similar contextual supports as others. Declining self-efficacies mattered for teachers’ level of professional engagement. Although tempting to infer self-efficacy increases with length of experience it was unrelated for one ‘declining’ cluster, pointing to quality rather than quantity of experience as key. It is crucial for policymakers, educators and administrators to set in place contextual structures and practices to nurture beginning teachers’ professional engagement and development by supporting the range of expected competences outlined by national frameworks.
Acknowledgments: This work is conducted jointly with Professor Paul Richardson, Monash University, funded by Australian Research Council: DP0666253, DP0987614, DP140100402
1 Watt & Richardson, 2008 – presented at Oxford in 2009.
