Abstract

Despite the intended aims of PD programs for teachers, critical questions remain about their effectiveness in general, rather than the number of PD programs teachers participated in and completed. This study aims to evaluate the effectiveness of a content-specific knowledge (CSK)-focused PD program for teachers in low-performing schools in Kazakhstan, delivered online over 2 months. Using secondary data, it estimates teachers’ CSK gains and identifies patterns of variation in outcomes that can inform the design of future large-scale interventions. Overall, the PD program involved 43,724 teachers across 943 schools and required substantial financial and human-capital investment. Evaluating the effectiveness of this 80-hour online PD under these conditions is essential for accountability and for generating evidence that can guide future PD policy and program improvement. To answer the study’s RQs, a pre-experimental one-group pretest-posttest design was employed. However, this study focuses on 7037 Math, Physics, Biology, and Chemistry teachers out of 43,724 from 943 targeted schools. These schools were selected from urban and rural areas based on their results in international and national exams, and their teachers were mandated to complete PD programs specifically designed for them. Overall, the average CSK gain was very small, and heterogeneity in the average gains was detected across demographic and categorical variables. Pretest results consistently predict posttest outcomes, with some adjusted subgroup differences remaining after controlling for pretest as the baseline CSK. In multilevel analysis models, the main share in posttest variances was attributed to individual characteristics, with a small but important share attributable to schools and regions. This means that contextual conditions also shape teachers’ CSK results. As a result, these insights indicate that the well-known “one-size-fits-all” approach to improving teachers’ CSK through online PD programs will cause uneven gains and suggest that improving the effectiveness of these PDs will require targeted mechanisms.