Teacher Attention as a Self-Fulfilling Prophecy

Policy Context

Despite substantial global progress, disparities in educational attainment and labor market outcomes persist across many low- and middle-income countries, including Uganda. These disparities are shaped in part by entrenched social norms and beliefs about individuals’ abilities and future opportunities. When households and institutions face resource constraints, investments may be directed toward individuals perceived to yield higher returns, and thus reinforcing existing inequalities.

Teachers play a critical role in shaping educational outcomes (Bacher-Hicks and Koedel, 2023), yet they may unintentionally reinforce gaps due to internalized beliefs shaped by broader societal norms. These beliefs can subtly influence classroom interactions and attention allocation, indirectly perpetuating inequalities. Increasingly, evidence suggests that discrimination is often subtle, manifesting through differential attention allocation rather than overt hostility. For example, Bartos et al. (2016) document how discrimination leads to differential information acquisition, while Glover et al. (2017) show how biased attention allocation can generate self-fulfilling prophecies in performance. Related work suggests that limited attention to less salient cues can compound dynamically over time, leading to divergent outcomes (Ho and Huang, 2025).

In educational contexts, beliefs and expectations about students’ comparative strengths may inadvertently shape instructional choices and access to opportunities, steering students into different academic trajectories. Despite growing recognition of these mechanisms, the empirical link between teachers’ implicit beliefs, classroom attention allocation, and student learning outcomes in low-income settings remains underexplored.

Study Design

This project implements a field experiment in partnership with Uganda’s Ministry of Education and Sports, focusing on Primary 7 students. Primary 7 is a critical transition year, as students prepare for the Primary Leaving Exam (PLE), which determines secondary school placement. The intervention uses additional after-school tutoring sessions (which are highly valued but often prohibitively expensive) as an ethical way to manipulate teacher attention allocation without reallocating existing classroom resources.

Teachers first nominate students they believe would benefit most from tutoring, allowing measurement of baseline beliefs and attention allocation. Tutoring recipients are then randomly selected from these nominations, stratified by teachers’ initial selections. This design enables causal identification of how teacher attention allocation affects student learning outcomes and whether teachers update their beliefs based on observed student progress.

Results and Policy Lessons

As part of the scoping phase, test score data were collected for approximately 2,600 students in Primary 1 through Primary 7 across six government-aided primary schools in Jinja District. While boys and girls performed similarly in earlier grades, statistically significant gender gaps emerged by Primary 7, with boys outperforming girls across subjects.

To assess whether these gaps were correlated with implicit gender bias, mobile-phone based Implicit Association Tests (IATs) were piloted among teachers, and interviews were conducted with school-children about salient forms of discrimination. Across both exercises, little evidence of systematic gender bias was detected. Teachers did not consistently associate female names with household tasks or non-STEM subjects, and children rarely cited gender as a salient dimension of discrimination.

Taken together, these scoping results suggest that gender-based discrimination in schools may be less salient in this context than previously hypothesized. As a result, the project is expanding to explore settings where discrimination may be more salient, including refugee-host community interactions in schools in northern Uganda, where social identity differences may more strongly shape teacher attention allocation and student outcomes.