Math Belonging and Anxiety: Longitudinal Associations Across High School

• In Event: Understanding How Math Anxiety Develops, Operates, and Is Better Treated Using Multi-Method Approaches

Abstract

Based on the Situated Expectancy-Value Theory and supporting studies, research suggests that the attitudes students hold about math have significant consequences for their performance and willingness to engage with the subject (Eccles & Wigfield, 2020; Herges et al., 2017). Math anxiety (MA) is one of the most well-researched attitudes about math and has been shown to have deleterious effects on students’ success in mathematics (Ashcraft, 2002). While the relationship between MA and educational outcomes is well documented, less is known about how MA relates to another important but under-researched contributor to success in mathematics: the extent to which students feel like they belong and are accepted in math classes. Considering that adolescents’ sense of belonging in math classroom is a noteworthy attitude that can be cultivated to facilitate students’ learning and engagement in mathematics (Barbieri & Miller-Cotto, 2021) and may be thwarted by the presence of math anxiety, it is critical to consider the co-development between math belonging (MB) and MA. Therefore, using a sample of high school students from Grades 9 to 12, the present study examined the longitudinal reciprocal relations between MB and MA at both the inter-individual and intra-individual levels.
Participants were students surveyed in ethnically diverse California high schools as part of a large-scale, longitudinal survey (N = 4400, 53.7% female). The sample was 33% Latinx, 24% White, 15% Asian, 11% African American, and 17% Other. MA was measured annually from Grades 9 to 12 with 2 items about students’ feelings of anxiety within their math class (αs = .75-78) (e.g. I feel stressed out during math class). Similarly, MB was measured annually by six items capturing students’ feelings of belonging and acceptance within their math class (αs = .79- .81) (e.g. I have good friends in my math class).
The Random Intercept Cross-Lagged Panel Model (RI-CLPM; Hamaker et al., 2015) was used to disaggregate the associations between MA and MB at the between-person level, as well as within-person year-to-year co-fluctuations and predictive effects. Results (as shown in Figure 3) suggested that at the between-person level, students with higher MA are likely to experience lower MB. At the within-person level, students with higher MA than usual in a given year co-occurred with lower MB than their typical level at the same grade (i.e., contemporaneous co-fluctuations). Findings did not show that higher MA in a given year prospectively predicted lower MB the following year (i.e., no within-person reciprocal effects).
Findings showed that the associations between MA and MB may likely operate at the stable, trait-like between-person level, rather than reciprocally state-like within-person level. These patterns suggest that targeting the structural and relatively stable but potentially modifiable contextual factors in students’ classroom settings may be a promising path towards not only reducing students’ MA, but also improving their sense of belonging in mathematics.

Authors

• Piper Stanger, California Department of Education

• Tianyu Li, Austin Peay State University

• Sandra Graham, University of California - Los Angeles