Abstract
This study examined the association between low-achieving secondary school students’ exposure to non-physical punitive discipline and their academic self-concept in Trans-Nzoia County, Kenya, where punishment remains a routine response to poor performance despite a thin local evidence base. Framed by self-determination theory and Shavelson’s model of academic self-concept, the study used a cross-sectional, convergent parallel mixed-methods design; because it delivered no treatment and no repeated measurement, it reports associations rather than effects. From a population of 25,075 low-achieving students, 964 class teachers and 241 deputy principals, 384 students completed a five-point Likert questionnaire and 68 teachers and 34 deputy principals were interviewed. Quantitative data were analysed using descriptive statistics and bivariate regression; interview data using reflexive thematic analysis. A statistically significant but weak positive association emerged between reported punishment and academic self-concept (β = 0.311, R² = 0.097), explaining under 10 per cent of the variance and running counter to the international literature; it is attributed mainly to a belief item carried in the predictor and to learners reading structured attention as care. The interview strand, in which teachers doubted the value of punishment, diverged from the survey and is treated as a substantive finding. The cross-sectional, unadjusted design and an unvalidated punishment measure preclude causal or directional claims. Schools should replace punitive responses to academic failure with private, behaviour-specific corrective feedback supported by school counselling.