Paper Summary
Share...

Direct link:

Improving Math Attitudes with a Play-Based Intervention: Evidence from Kosova

Thu, April 9, 9:45 to 11:15am PDT (9:45 to 11:15am PDT), Westin Bonaventure, Floor: Level 2, Beverly

Abstract

Objectives and Theoretical Framework
The affective domain (beliefs, attitudes, and emotions) is an important yet understudied dimension of mathematics education (Lewis, 2013). Students’ attitudes toward math are closely linked to performance, particularly with challenging topics like rational numbers (Dowker et al., 2019). This highlights the need for math activities that improve both learning and affective outcomes, especially in low- and middle-income countries (LMICs) facing learning poverty and home to 90% of the global population (UNESCO, 2023).
A play-based math intervention co-designed with teachers and students has demonstrated positive cognitive and social-emotional outcomes in under-resourced U.S. communities (Authors, 2024). Its low-resource model shows promise for scalability in LMICs with limited infrastructure. Kosova, a post-conflict LMIC, where math education is still developing (Kingji-Kastrati et al., 2017), offers a unique setting to adapt this approach through co-design.
Situated Expectancy-Value Theory (Eccles & Wigfield, 2020) posits that students’ perceived competence influences the value they place on learning. While our prior research showed that the intervention improved rational number learning (Authors, 2024), some scholars have raised concerns that academically effective instruction may undermine students’ social-emotional experiences (Blazar & Pollard, 2023; Boaler, 2014). This study investigated whether, in a game-based context, affective outcomes were aligned with students’ learning gains. We also examined whether blocks of classrooms showing greater learning impacts would also exhibit stronger impacts on affective outcomes, offering insight into whether positive affect and achievement necessarily involve trade-offs.
Methods
The study involved three schools, 25 teachers, and 652 fourth- and fifth-grade students in Kosova. In collaboration with local teachers, we co-designed a 15-unit intervention delivered over three weeks (see Figure 1). Classrooms were assigned to treatment or control conditions using matched randomization based on pretest math scores. Students completed identical pre- and posttests covering 13 rational number subskills (e.g., fraction addition, fraction to decimal conversion, and number line). A composite rational number score was used.
Students’ attitudes were measured using 16 Likert-scale items adapted from Simpkins et al. (2016), covering importance, self-concept, and interest. Items were administered post-math section at both time points. We pre-registered construct-level outcomes (PRE-REGISTRATION WEBSITE) and created composite scores for overall fraction and decimal attitudes.
Results
The intervention yielded positive impacts on several attitude constructs, ranging from 0.07 to 0.42 (Table 1). We found a statistically significant impact on decimal self-concept (b = .42, p = .02) and a marginally significant impact on overall attitude towards decimal (b = .32, p = .02).
Additionally, fifth-graders benefited more across decimal self-concept, interest in fractions and decimals, perceptions of decimal importance, and overall attitude toward decimals, compared to fourth-graders. Female students showed greater gains in fraction self-concept than males.
Exploratory analysis showed that blocks of classrooms with greater academic impacts also exhibited larger attitude improvements (Figure 2), suggesting no trade-offs between cognitive and affective gains.
Significance
Play-based interventions can simultaneously support math learning and attitudes, even with challenging topics. These findings are especially relevant in LMICs like Kosova, where scalable, low-cost programs may offer promising approaches to supporting student learning and development.

Authors