Mathematics Anxiety in Kindergarten Students: Concurrent and Longitudinal Relations with Mathematics Performance

• In Event: 2-023 - Identifying Important Concurrent and Longitudinal Correlates of Math Anxiety in Elementary School Children

Integrative Statement

Over the past decade, researchers have become more interested in the early development of mathematics anxiety and have designed mathematics anxiety scales for use with young children (Harari et al., 2013; Ramirez et al., 2013). This research has found small-to-moderate relations between mathematics anxiety and mathematics performance (e.g., Ganley & McGraw, 2016). However, none of this work has studied children in kindergarten. In this study, we examine whether we can feasibly measure mathematics anxiety in kindergarten students (ages 5-6) using an adapted version of the Math Anxiety Scale for Young Children-Revised (MASYC-R; Ganley & McGraw, 2016; Harari et al., 2013). We also examine how well mathematics anxiety predicts mathematics performance for kindergarten students, and we compare these relations to those found with a sample of grade 1-3 students.

Participants were 105 kindergarten students and 130 students in grades 1-3 tested in fall and spring of the school year. For kindergarten students, we adapted the likert-type MASYC-R response options from “No, Not really, Kind of, Yes” to “No, No, Yes, Yes”, where the text size of the leftmost “No” and rightmost “Yes” were larger than the text of the middle options. We also measured mathematics confidence and interest to use as covariates (Fredricks & Eccles, 2002). Mathematics performance was measured using an adapted assessment from Schoen et al. (2017), and scores were standardized within grade.

Table 1 displays descriptive statistics, coefficient alphas, and Pearson correlation coefficients for measures in both kindergarten and grades 1-3. Coefficient alphas for the MASYC-R were similar for kindergarten and grades 1-3 students (.88
Using multiple regression, accounting for clustering within teachers, we predicted spring mathematics performance from grade level (kindergarten or grades 1-3), fall mathematics anxiety, confidence, interest, performance, and the interaction between grade level and fall mathematics anxiety (Table 2). Results showed mathematics anxiety predicted spring mathematics performance only when fall mathematics performance was not included in the model (Models 1 and 2). The nonsignificant interaction terms indicated that the strength of this relation did not differ across grade level.

Thus, we find some initial evidence that the adapted MASYC-R can be used with children in kindergarten and that it performs similarly to how it performs with students in grades 1-3. The higher ratings at fall of kindergarten could indicate some difficulty with the scale for children this young, which we will investigate further using cognitive interviews. For all students, mathematics anxiety predicted later mathematics performance, but only when past mathematics performance was not included. These results suggest that mathematics anxiety is measurable in kindergarten and has some relations with mathematics performance. Interventions for mathematics anxiety may need to start earlier in development than initially thought.

Authors

• Colleen M Ganley, Florida State University
  Presenting Author

• Connie Barroso, Florida State University
  Non-Presenting Author

• Elyssa A Geer, Florida State University
  Non-Presenting Author

• Rachel Conlon, Florida State University
  Non-Presenting Author

• Amanda L McGraw, Florida State University
  Non-Presenting Author

• Robert C Schoen, Florida State University
  Non-Presenting Author

• Christopher Schatschneider, Florida State University
  Non-Presenting Author