Longitudinal and reciprocal links between metacognition, mathematical modeling competencies, and mathematics achievement in grades 7–8: A cross-lagged panel analysis

The relevance of metacognition and mathematical modeling competencies to the development of good mathematics achievement throughout schooling is well-documented. However, few studies have explored the longitudinal relationship among metacognition, mathematical modeling competencies, and mathematics achievement. More importantly, the existing research has mostly focused on unidirectional effects with metacognition typically modelled as antecedents of mathematical modeling competencies and mathematics achievement. Nevertheless, the relationships among metacognition, mathematical modeling competencies, and mathematics achievement may be dynamic, and variables might reciprocally influence each other. Hence, we conducted a longitudinal study examining the reciprocal associations between metacognition, mathematical modeling competencies, and mathematics achievement. To this end, we recruited 408 seventh-grade students to complete a metacognition-related questionnaire and a mathematical modeling competencies test concurrently. This procedure was repeated one year later. A cross-lagged panel analysis showed four main findings: (a) metacognition in Grade 7 longitudinally predicted mathematical modeling competencies in Grade 8; (b) mathematical modeling competencies in Grade 7 longitudinally predicted metacognition and mathematics achievement; (c) higher levels of mathematics achievement drive the subsequent shaping of metacognition and mathematical modeling competencies; (d) There were no gender differences among metacognition, mathematical modeling competencies, and mathematics achievement. Finally, theoretical and practical implications are discussed.