Pygmalion in the Chinese High School: Mixed-Methods Research

Abstract

Although studies on teacher expectations and teacher expectation effects have flourished over the past 52 years, little is known about how teacher expectations might influence student outcomes in a context different from the Western context. This doctoral research focussed on the teacher expectation-confirming process at senior high school level in an under-researched context: China. Specifically, the research provided perspectives of the teacher expectation-confirming process, differences in teacher characteristics across high-, medium-, and low-expectation teachers, the roles played by teacher characteristics in the expectation-confirming process, and the stability of teacher expectation bias across two school years. The findings of the doctoral research thus contribute importantly to the literature by revealing the crucial moderating role of teacher characteristics in the expectation-confirming process, the considerable class-level teacher expectation effects, the stability and change trajectories of teacher expectation bias over time, as well as contributing some unique findings (some of which might be culturally-specific) about the process. Study One, a qualitative study, brought together both teachers’ and students’ voices to inform the teacher expectation-confirming process in the Chinese context. Thematic analysis and the constant comparative method were employed to analyse interview data from 14 teachers and 14 students. Results suggested the expectation-confirming process existed in the new context and was similar to the process found in the Western context. Further, some unique findings from this study indicated the context-specific feature of teacher expectation theory. Study Two, characterised by a cross-sectional design, employed mixed methods (confirmatory factor analysis, multi-group invariance testing, multivariate analysis of variance, and structural equation modelling) to analyse data from 1,819 students and their 119 teachers. Results revealed that there were statistically significant differences in teacher characteristics across high-, medium-, and low-expectation teachers, and both teacher efficacy and the degree of passion for teaching predicted teacher expectations positively. Further, a moderate-to-strong teacher expectation effect was found at the class level, and teachers’ passion for teaching was found to be a significant predictor of students’ class-level achievement. Study Three incorporated a longitudinal design using Pearson correlation techniques to evaluate stability in teacher expectation bias across all teachers and hierarchical linear mod techniques to investigate the change trajectories of teacher expectation bias for high-, medium-, and low-expectation teachers. In this study, 567 students and their 49 teachers were followed for two school years. Results revealed that teachers seemed to adjust their expectation bias in the initial six months and maintain their adjusted expectation bias afterwards. However, contextual factors might induce teachers to adjust their expectation bias again. Further, some high- and low-expectation teachers tended to be always high- and low-expectation teachers across two school years. Collectively, this thesis concludes with the suggestion that the theory of teacher expectations is both cross-culturally universal and context-specific, and that teachers matter in the expectation-confirming process.