Abstract
<title>Abstract</title> <p>Students’ attitudes toward mathematics—interest, confidence, and perceived value—are important predictors of educational outcomes, yet assessing all three dimensions simultaneously often imposes substantial survey burden. Using TIMSS 2023 data from 232,970 Grade 8 students across 44 countries, this study examined whether cross-construct information among mathematics affective scales can be used to develop a compact screening instrument. All 26 items from the TIMSS Students Like Learning Mathematics (SLM), Students Confident in Mathematics (SCM), and Students Value Mathematics (SVM) scales were pooled and analyzed using a Gaussian Graphical Model. Items were ranked by betweenness centrality and sequentially added to predict TIMSS official three-level classifications (High, Intermediate, Low) through multinomial logistic regression. Results showed that SVM items occupied the most central bridge positions in the affective network. The item “I would like a job that involves using mathematics” exhibited the highest betweenness centrality and was selected first across all three dimensions. Predictive performance plateaued with seven items for SLM, six items for SCM, and two items for SVM. The SLM and SCM solutions shared a common six-item core, yielding a seven-item battery capable of simultaneously screening all three dimensions with high accuracy (AUROC = .889–.939), representing a 73% reduction from the original 26-item instrument. Cross-national analyses demonstrated generally stable predictive performance across 44 countries despite moderate variation in bridge-item structure. The findings identify mathematics value beliefs as the structural hub of the mathematics affective network and provide a practical seven-item instrument for efficient affective monitoring in research and educational settings.</p>