Despite the social and technological advancements of the 21st century, school bullying remains a significant problem in education systems. Beyond the psychological distress, school bullying affects the mind’s core cognitive appraisals and creates an unfair academic disadvantage to victims. The moral dilemma of random assignment impedes experimental research on bullying, thereby making it difficult to gain insights into the detrimental effects of bullying on victims’ cognitive appraisals and academic outcomes. Propensity score analysis can simulate a quasi-experimental design with balanced groups to overcome this challenge. This study seeks to analyse the prevalence of bullying across nations and to employ propensity score matching (PSM) on a nationally representative TIMSS 2019 dataset to mitigate selection bias in evaluating the effect of bullying on science performance and cognitive appraisals. Descriptive statistics were obtained from 162,365 eighth-grade students in 24 countries. PSM analyses were performed on the Türkiye dataset (4,077 students). Generally, the percentage of high-frequency bullying was associated with lower science achievement across countries. PSM results showed that bullying victimization significantly decreased students’ science achievement and perceived control in learning science. Additionally, a marginally significant positive effect of bullying on perceived value of learning science was found. Similar results from OLS regression and sensitivity analyses provided evidence for the robustness of the estimated effects. Overall, the findings show that bullied students experience cognitive disadvantages. The contributions of these domain-specific findings to control-value theory and its implications for anti-bullying initiatives were discussed.