Conceptualising and measuring executive functions in Africa: A mixed methods approach to developing an observational measure for young children in Nigeria.

Abstract

Executive function (EF) measures of children in Africa are heavily dependent on task-based assessments that are mostly normed in non-African contexts. This challenge limits the chances of understanding how children use their EF to solve everyday problems and engage in routine-based behaviours, especially in the classroom. Thus, exacerbating the transfer problem associated with EF measures. To fill this vacuum, this thesis is divided into three distinct studies and sets out to develop an observational measure that helps to capture how children display their EF behaviours in a natural classroom environment without disruption to their daily routines. The first study was a systematic of review of EF studies across Africa 54 countries and eight territories where 87 studies were qualitatively synthesised. Results indicated that most of the authors were affiliated with non-African institutions and task-based measures accounted for over ninety percent of the measures used. This sets the foundation for the development of an ecologically informed observation measure in the second study. In study two, I used interviews to obtain teachers’ contextual understanding of EF in their classrooms. The results of the thematic analysis from both inductive and deductive coding were used to develop a 10-item children's executive function observation scale (CEFOS). The face and content validity of CEFOS were assessed by classroom teachers and developmental scientists who were experts in EF studies and had experience working in the African context. In the third study, the psychometric properties of the 10-item CEFOS were evaluated after observation data was obtained from 102 participating children. Positive estimates were obtained from the inter-item correlation analysis. Reliability estimates using Cronbach Alpha per item ranged between 0.91 and 0.92. Results from the exploratory factor analysis (EFA), after accounting for the role of age, showed that the three-factor model is the best fit. I integrated the findings into a general discussion, highlighted the implications, and provided directions for further research.