How Teacher-Student Interactions Shape Math Attitudes and Gender-Math Stereotypes: Highlighting Gender Disparities and Cross-Subject Dimensional Comparison

• In Event: Comparative Studies in Education (II)

Proposal

Framework and Research Questions

The persistent underrepresentation of women in science, technology, engineering, and mathematics (STEM) fields has been a longstanding concern for educators, policymakers, and researchers. Despite gender similarity in math ability (Hyde et al., 2008; Lindberg et al., 2010), studies show persistent gender differences in STEM-related attitudes and aspirations among adolescents (Charles et al. 2014; Mann, Legewie, & DiPrete, 2015; OECD, 2015), raising questions about the underlying factors contributing to these disparities. This paper focuses on one key aspect that merits closer examination: the role of teacher-student interaction (TSI) in shaping student educational experiences and learning outcomes.

Teacher-student interaction plays a pivotal role in shaping student learning motivation, efficacy, aspiration, and performance. A supportive and engaging learning environment with positive teacher-student interactions can make students feel valued and understood by their teachers and profoundly impact their academic achievement and attitudes (e.g., Cornelius-White, 2007; Gentrup et al., 2020). However, while research has revealed notable gender biases in teacher expectancies and perceptions, especially in math and science classes (e.g., Riegle-Crumb & Humphries, 2012; Robinson-Cimpian et al., 2014), evidence regarding potential gender disparities in TSIs remains limited. Moreover, much of the existing literature on TSI has primarily focused on the within-subject effects of TSI, such as how interactions with math teachers shape student outcomes in math, while overlooking the potential cross-subject TSI effects, such as how interactions with language teachers shape student outcomes in math.

To address these limitations, this study adopts the dimensional comparison theory, which compares student achievements across different academic domains and reveals cross-subject contrast effects (Marsh & Hau, 2004; Möller & Marsh, 2013). Extending these findings to TSI, we argue that students do not interact with teachers in an isolated subject-specific environment, and it is crucial for us to develop a comprehensive understanding of how TSI in different subject areas may collectively shape student attitudes and performance.

Incorporating the dimensional comparison theory and previous literature on teacher-student interactions, this paper examines the effect of teacher-student interaction, including positive interaction (such as praising) and neutral interaction (such as asking questions), on student math attitudes, performance, and STEM aspirations. In particular, we extend previous TSI research by focusing on the potential cross-subject effects, that is, the non-math TSI effect on student math outcomes net of the math TSI effect. Utilizing data from the nationally representative China Education Panel Survey (CEPS), we examine the following research questions:

RQ1: Are there gender disparities in math and non-math TSIs after controlling for student academic performance and other sociodemographic characteristics?

RQ2: Are math and non-math TSIs associated with student math-related attitudes and performance?

RQ3: Are there gender differences in the with-in and cross-subject effects of TSIs?

Data and Methods

We utilize CEPS, a nationally representative student survey project conducted by the National Survey Research Center at Renmin University, China. Both the CEPS baseline (Wave 1, 2013-14) and follow-up (Wave 2, 2014-15) surveys are used. The sample includes approximately 20,000 students in the seventh and ninth grades from 112 schools across mainland China.

The investigation employs two questions from the Wave 1 student questionnaire to measure TSIs. Students were asked about the frequency with which their math, Chinese, and English teachers prompted them to answer questions (neutral TSI) and offered them praise (positive TSI). Responses, on a four-level scale, were averaged to create separate positive and neutral TSI measures for math and non-math subjects. Additionally, the difference between the math and non-math TSI measures was calculated to create variables for math advantage in positive and neutral TSIs.

The study scrutinizes four outcome variables from Wave 1 and Wave 2, including STEM aspirations, math self-efficacy, math and non-math performance, instrumental motivation to learn math, and endorsement of gender-math stereotypes. Control variables include student academic performance, sociodemographic characteristics (parental education, family economic status, household living arrangement, rural/non-rural hukou, local/migrant, only child, and boarding student or not). All control variables are measured in Wave 1.

The analytical strategy incorporates linear regression models with classroom fixed effects, predicting TSI measures using student gender, academic performance, and sociodemographic characteristics to explore potential gender disparities in TSI. Further, Wave 1 and Wave 2 outcome variables are predicted with Wave 1 TSI measures, controlling for student Wave 1 performance and sociodemographic characteristics. Interaction terms are included to examine the differential effects of TSI between boys and girls. Classroom fixed effects are included to control for all classroom-level heterogeneities.

Findings and Conclusions

First, we find significant gender disparities in TSI patterns: after controlling for individual characteristics and classroom fixed effects, boys interact more with both math and non-math teachers than girls. Moreover, boys enjoy a larger math advantage in TSIs than girls.

Second, we show that frequent teacher-student interactions positively influence students' learning attitudes within the same subject. Higher levels of positive and neutral math TSIs are associated with increased math self-efficacy and motivation. Furthermore, a higher level of positive math TSI is linked to an increased likelihood of students developing STEM aspirations later on.

Third, we uncover a dimensional comparison effect of TSI, with its direction dependent on the nature of the interaction. Accounting for math TSI, students with higher non-math positive TSIs experience a decrease in math self-efficacy, aligning with the “cross-subject contrast effect” observed in Western contexts (Dietrich et al. 2015). Conversely, students with higher non-math neutral TSIs display increased math motivation, suggesting a possible “cross-subject reinforcement” effect.

Lastly, we reveal significant gendered effects of TSI on student attitudes toward gender-math stereotypes. Positive interactions with math teachers enable girls to challenge these stereotypes, while potentially reinforcing these stereotypes for boys. Intriguingly, and in line with dimensional comparison predictions, increased praise from non-math teachers may unintentionally reinforce gender-math stereotypes among girls while encouraging boys to question those stereotypes.

In conclusion, this study underscores the importance for teachers to raise awareness of potentially gendered TSI patterns and to create inclusive classroom environments with positive TSIs across academic subjects. Policymakers and schools should monitor and evaluate such classroom practices in maximizing the positive effects of TSIs and addressing potential gender disparities.

Author

• Ran Liu, University of Wisconsin-Madison