IntroductionIt has been suggested that students with growth mindsets are more likely to achieve better mathematics learning results than their counterparts with fixed mindsets. However, inconsistent and some even contradictory results have been reported in recent studies which examined the associations between growth mindset and mathematics achievements, suggesting the complexity regarding the effects of growth mindset on academic achievements.MethodsThis study aims to examine students' growth mindsets, failure attributions, intrinsic motivation, mathematics self-efficacy, mathematics anxiety and mathematics achievements in one model to capture the sophisticated functioning processes of growth mindset. A total number of 266 middle school students in China participated in this study. Students' mindset and related variables (i.e.,motivations to learn mathematics, attributions of failure in mathematics, mathematics anxiety, mathematics self-efficacy) were measured at year 7, the first year of junior middle school in China. These students' mathematics learning outcomes were tracked from year 7 to year 9, the end of junior middle school. Structural equation modeling (SEM) was used to investigate the relations among students' growth mindsets, failure attributions, intrinsic motivation, mathematics self-efficacy, mathematics anxiety and mathematics achievements.ResultsThe results show that: (1) growth mindset doesn't directly predict mathematics achievements; (2) growth mindset indirectly influences mathematics achievements through intrinsic motivation; (3) failure attributions and mathematics self-efficacy sequentially mediate the association between growth mindset mathematics achievements; (4) failure attributions and mathematics anxiety sequentially mediate the relationship between growth mindset mathematics achievements.DiscussionThe results of this study contribute a better understanding about how growth mindsets make impacts on middle school students' mathematics achievements. These findings have important implications for mathematics education in that we could not simply cultivate students' growth mindset in schools with expectations of higher mathematics learning outcomes. Instead, along with the growth mindset intervention, it is fundamental to make interventions on students' intrinsic motivation, failure attribution, mathematics self-efficacy, and mathematics anxiety in mathematics teaching and learning.