Self-efficacy Factor Invariance across Pre-service and In-service Teachers in Botswana

Abstract

The challenges facing measurement experts and researchers in the social sciences field is ensuring that measurement scales operate equally across different sub-populations. Measurement invariance across gender for example means that males and females conceptualize and interpret the construct being measured similarly and thus meaningful comparisons between the groups can be made. Measurement non-invariance on the other hand can be a threat to validity as it implies that the construct being assessed has different structures across groups. The current study investigates the invariance of self-efficacy construct across pre-service and in-service teachers. The teachers were administered a Teacher Self Efficacy Scale with 33 Likert scale items and the data were analyzed using maximum likelihood procedures with Varimax rotation techniques. Reliable dimensions were extracted on the basis of two factor extraction procedures. The final solutions for the two groups were interpreted with reference to the factor loadings and item content for items that had significant loading on a particular factor.