Mathematical engagement, self-efficacy, metacognition, and social support as predictors of mathematical problem-solving

Mathematical problem-solving skills are among the priority outcomes in teaching programs. Identifying the cognitive, affective, and social factors that influence the development of these skills is critically important in mathematics education, both theoretically and practically. The aim of this study is to test a theoretically specified model examining the relationships among teacher and peer support, mathematics self-efficacy, mathematical engagement, and metacognitive experience, and to investigate their direct and indirect effects on students’ mathematical problem-solving achievement. The sample consisted of 465 middle school students (250 female [53.8%], 215 male [46.2%]) from two state schools in Istanbul. Data were collected by a mathematical problem-solving test, along with scales measuring mathematics self-efficacy, metacognitive experience, mathematical engagement, and teacher and peer support. The findings support a path model in which social support indirectly predicts mathematical problem-solving performance through self-efficacy and metacognition, and self-efficacy also indirectly predicts problem-solving performance through metacognition. The fact that metacognition directly and significantly predicted problem-solving performance, while the mathematics engagement variable did not demonstrate a significant effect and was therefore removed from the model, is considered among the noteworthy findings of the study.