PhD thesis defence by Yusuf Aguş
This thesis consists of three essays.
In the first chapter, I study how peer ability shapes children’s skill development and the social relationships that evolve alongside it. Leveraging the random assignment of first-grade students to classrooms in Turkish public schools and longitudinal data covering four years of primary school, I find that a one standard deviation increase in assigned peer ability raises fluid intelligence by 0.03 standard deviations, cognitive empathy by 0.06, numeracy by 0.09, and literacy by 0.11 on average. The effects on numeracy and literacy increasingly favor students with higher initial ability. These differences in learning emerge alongside changes in classroom social relationships: higher peer ability strengthens within-group friendship among high-ability students, while low-ability students initially form relatively more ties with high-ability classmates. This cross-ability response weakens over time, and low-ability students’ nominations toward high-ability classmates become relatively less likely to be reciprocated as students progress through primary school.
In the second chapter, jointly with Sule Alan, we study how teacher ability, measured by fluid intelligence, shapes student learning and the development of higher-order thinking skills. Exploiting the random assignment of first-grade students to teachers in Turkish public schools, we follow approximately 10,000 children over four years of primary school. A one standard deviation increase in teacher fluid intelligence raises foundational literacy and numeracy by 0.047 standard deviations and higher-order thinking skills by 0.034 standard deviations on average. The latter include abstract, analogical, categorical, and critical reasoning, skills that are rarely directly taught in the classroom. Evidence on mechanisms points to the classroom peer environment as one channel: more able teachers generate stronger peer effects and foster classroom environments that facilitate learning among students.
In the third chapter, jointly with Betul Turkum, we study how classroom peer relationships differed between cohorts of children who experienced the COVID-19 pandemic and those who did not. Using social network data from third- and fourth-grade students in Turkish primary schools, we compare friendship, academic support, and emotional support networks across pre-pandemic and pandemic cohorts. We document opposing changes across network types: friendship isolation increased by 22%, while academic support nominations rose by 20%. The increase in academic support was driven entirely by native students, while refugee students experienced deterioration in both friendship and academic support networks. These findings show that large-scale disruptions can reshape different dimensions of children’s social relationships in markedly different ways.
The event will take place in hybrid modality.
Register