Abstract
K-12 mathematics education in the United States is failing too many students, disproportionately students of color. Furthermore, math anxiety is common, leading many students to feel, “I am just not a math person.” Yet little is known about how positive student attitudes toward math might relate to the negative feelings of anxiety. This study tests a mathematical resilience scale, previously developed for college students, in the first published study of American high school students (N = 1,057). Specifically, this research tested whether mathematical resilience can be measured by how students value math, how students view struggle in learning math, and how students view their capacity to grow in mathematical understanding. The study also interrogates whether mathematical resilience addresses negative feelings toward math. Confirmatory Factor Analysis (CFA) measured the validity and reliability of the theorized mathematical resilience structure for the sample, and Structural Equation Modeling (SEM) methods tested the relationships among mathematical resilience, mathematics anxiety, and student outcomes. Results show that value, struggle, and growth in mathematics measure acceptably but moderately in the sample. Mathematical resilience factors negatively correlate with mathematics anxiety, but positively with student academic success, intent to engage with math, and proclivity to pursue math-related careers. One key finding is the relative importance of how students value mathematics compared to the growth and struggle concepts, providing an intriguing differentiation from the growth mindset narrative, but requiring further research. Educators and leaders in K-12 mathematics education may apply these findings to design interventions and foster positive student engagement. Keywords: Mathematical resilience, mathematics anxiety, high school, self-efficacy theory, agency theory, implicit theories of intelligence, value of mathematics, struggle in mathematics, growth mindset, confirmatory factor analysis, structural equation modelling