Poster #11 - Parental math relational language, math attitudes, and child math achievement: An investigation of low-SES families

• In Event: Poster Session 19

Abstract

Math learning, like many other cognitive abilities, is intertwined with one’s everyday learning experiences (e.g., Frank et al., 2008). Past research has shown the facilitative role of relational language in children’s STEM learning (Levine et al., 2010; Valle & Callanan, 2006). However, it is unclear how the learning practice is carried within lower-SES families given that much of the existing research has taken place in higher SES families. The present study, underscored by its goal of supporting math learning in underserved populations in the U.S., identifies the strengths of learning practices in low-SES families. We examine how parent use of math and non-math relational language during discussions of everyday scenarios with 1st graders relate to parent math attitudes and children’s math achievement.
Seventy low-SES families of first graders (predominantly African American) were given three scenarios (i.e., a weather report, a train station, a produce stand) to discuss together via Zoom (Figure 1). Parent-child dyads were given one scenario at a time in a fixed order and were given up to 5 minutes to discuss each.
The parent-child interaction was video recorded, transcribed, parsed, and coded at the utterance level. A comprehensive coding scheme was developed to analyze parents’ relational and non-relational language related to math/non-math topics. Math relational language was coded when the parent talked about math-related concepts using comparisons, definitions, self-connections, calculations, and the like, whereas math non-relational language focused on surface-level labeling of temperature, numbers, time, and other terms without making comparisons or connections. Parallel categories were coded for non-math topics (i.e., non-math relational, non-math non-relational). Parent math attitudes measures included math expectations and value and teaching math self-efficacy. Child math achievement was assessed at the beginning of the school year using the WJ-IV Math Facts Fluency subtest.
Preliminary results showed that parents used more math non-relational language (M=42.06, SD=21.58) than math relational language (M=29.96, SD=19.42; t(69)=4.47, p<.001) in general. A Repeated Measures ANOVA showed an interaction between scenario and language type (F(6, 65)=8.89, p<.001) such that parents used more math relational language in the weather report (M=11.61, SD =7.81) than the train station scenario (M=7.80, SD=7.07, M difference=3.78, SE=1.07, p=.002, 95% CI[1.15, 6.40]). They also used more math non-relational language in the produce stand (M=16.76, SD=9.68) than the weather report scenario (M=11.10, SD=8.65, M difference=5.66, SE=1.25, p<.001, 95% CI[2.59, 8.73]). Parents’ self-efficacy for teaching math predicted their use of math relational language (β=.28, p=.04), but not their use of math non-relational language (β=.12, p=.39). A multiple regression analysis showed that parent math expectancy and value of their child’s math learning (β=.50, p<.001) and their math relational language (β=.35, p=.004) positively related to children’s math achievement (Table 1).
Our results show wide variations in parents’ use of math relational language that is linked to their child-focused math attitudes and their children’s math achievement. These findings highlight potentially fruitful ways to support math relational talk in the home environment, which could be an effective lever in raising children’s math achievement and interest and reducing inequalities in math outcomes.

Authors

• Yu Zhang, University of Chicago
  Presenting Author

• Dania Smithstein, University of Chicago

• Susan Cohen Levine, University of Chicago