Adaptation and validation of the teachers’ self-efficacy scale in Ethiopian secondary schools: The case of the Amhara region

Teacher self-efficacy (TSE) research has garnered increased attention in recent years due to its significant relationship with important outcomes, including teachers' effectiveness in instruction, well-being, burnout prevention, and professional dedication. Although teacher self-efficacy has been extensively researched, there are no psychometrically sound measures specifically tailored to the Ethiopian secondary school educational system, hindering accurate evaluation and effective intervention. Furthermore, earlier research in other contexts found variations in the dimensional structure of the TSE Scale. To fill these gaps, this research adapts and validates the TSE scale for use in Ethiopian secondary schools. A deductive approach was utilized. The adaptation of the TSE Scale began with a thorough review of theories and indicators. Tschannen-Moran and Woolfolk Hoy's (2001) scale was selected due to its strong psychometric properties and relevance. To adapt the scale for the Ethiopian context, it was translated into Amharic, reviewed for clarity, and validated by subject experts. Initially, eight items from a total of 24 items were removed through expert judgment. This resulted in the refinement of 16 out of 24 items with strong content validity. Then, KMO and Bartlett's χ2 test were used to assess the likelihood that the data would factor well prior to the EFA analysis. KMO index was 904 and χ2 (df = 66, n = 714) = 5018.942, p < 001, above the suggested minimum value of .6. A pilot study with 92 teachers confirmed high reliability across subscales, and EFA with 714 participants verified the scale's structure. A substantial three-factor solution was found using the varimax rotation principal component analysis extraction approach. The three dimensions (each with 4 items) explained around 72.05 percent of the variation. Four items were removed due to overlapping problem and low factor loading during EFA results. Confirmatory factor analysis (CFA) further confirmed the model's fitness by removing one item, ensuring its robustness for Ethiopian secondary schools. The values of the re-specified model indices indicated that the model yielded reasonably good fit indices, with χ2 = 182.829, df = 38, χ2/df = 4.8111, CFI = .97, TLI = .97, RMSEA = .07, and RMR = .035, SRMR = .0355, NFI = .97, GFI = .91, NNFI = .96, and AGFI = .90 were used to accept the CFA model fit. The TSE and distributed leadership practice scales showed associations consistent with the theoretically expected connections between variables. Thus, with the initial evidence of a valid and reliable teachers' self-efficacy scale, a pool of 11 items under three factors (efficacy for student engagement, efficacy for classroom management, efficacy for instructional strategies) was developed. According to the study's findings, the self-efficacy measure is a useful instrument for measuring Ethiopian teachers' self-efficacy.