Teacher-Student Interactions and Educational Outcomes
The organization and quality of education are increasingly in the spotlight, both in academic and policy circles across the world. Specifically, some recent policy initiatives rely on measures of teachers’ quality calculated under the assumption that a given teacher has the same impact on any student. We will focus instead on the potential importance of "teacher-student interactions" relying on a novel research design to credibly ascertain how changes in the composition of students affect teacher behavior: Do teachers teach differently across students? Do they direct effort towards more able or more disruptive students? Does this vary with the class they are assigned, e.g. a high vs. a low achieving class? Does it vary with the school they teach at, e.g. one serving high or low socioeconomic status students? Thus, this project will tackle new research that has not yet gained sufficient scientific attention but is potentially of first–order importance for educational policy aiming at providing children with equal educational opportunities and in fostering economic growth. Exploring such questions entails developing novel survey instruments focused on teacher-student relationships and using large administrative data. We will focus on Romania which provides a unique setting to tackle methodological and data-related challenges. In addition, the project has the potential to provide evidence on the long term consequences of educational quality on student outcomes.
Final report
The organization and quality of education are fundamental to both individual and societal growth. Public education is intended to promote economic and social mobility for all students. However, in countries such as the US, Canada, and parts of Europe, schools often group students into classrooms based on academic ability—a practice known as tracking—which can unintentionally reinforce the very stratification that education aims to reduce.
Economists have focused on estimating the causal effect of tracking on cognitive skills as measured by academic achievement. Such studies often highlight a potential tradeoff: tracking may allow for more targeted teaching, but it can also deny lower-achieving students the positive peer effects that emerge from interactions with higher-achieving students. Additionally, tracking may stigmatize some students, hurting their self-perceptions (self-confidence, effort in school, and so on), which could have short and long-term consequences.
The main purpose of this project was to deepen our knowledge on the effects of how having access to a better class affect the teacher-student interactions and the (self-)perceptions of the students and their teachers on the students’ ability, academic performance and disruptive behavior. Understanding these interactions is important but remains understudied, largely because answering these questions in a credible, causal way is challenging due to data limitations and methodological difficulties.
In “Self-, Peer- and Teacher Perceptions under School Tracking” we have examined how assignment to a lower vs. higher track affects non-cognitive skills as captured by perceptions of self-confidence, academic effort, disruptive behavior, and self-esteem. We consider this question within a causal framework—we ask whether the assignment to different tracks causally impacts how students perceive themselves relative to their peers and how their peers and teachers perceive them.
Our project leverages a unique education system stratified by ability in Romania, offering an ideal framework for addressing methodological and data-related challenges in understanding the effects of tracking in education. In this system, the “top” class contains the students with the highest admission scores, while the “bottom” class contains those with the lowest. Classes always receive instruction separately in all subjects, although they cover the same curriculum, often delivered by the same teachers. These features enable us to isolate the role of student allocations from differences in curriculum and teacher quality.
We use two strategies to explore how student, peer, and teacher perceptions vary under school tracking. First, we use a regression discontinuity design to estimate the effect of being assigned to the top relative to the bottom class on student, peer, and teacher perceptions. Second, we use student fixed effects when directly comparing student perceptions to teacher perceptions, and when directly comparing peer perceptions of one’s own classmates to those of students in other classes.
An important part of our project development project is data collection. Besides using high-quality administrative data, we have developed a new original survey of teachers and students that allows us to tackle the methodologies and questions described above. This is an important tool that will also be used in several other (ongoing) projects.
To summarize, in this data collection we focused on a sample of schools who track students by ability and implemented an individualized survey asking students about their perceived relative standing across both top and bottom classes. The survey elicited students’ perceptions of themselves and their peers in the same class and in the other classes, along academic domains such as self-confidence in school, effort (e.g., does homework, pays attention in class), ability (e.g., understands hard concepts easily), and expected performance at the standardized exam, as well as on disruptiveness (e.g., harasses/disparages peers). We also surveyed all teachers who taught the same subject to both the top and bottom classes, asking them to rank the same students across the same domains as the students.
We show five findings:
1) being allocated to the top vs. the bottom class lowers a student’s perception of how she ranks relative to her peers in the same specialization; these negative impacts are in stark contrast to teachers’ perceptions;
2) along all dimensions, students rank their own classmates significantly higher than their peers from other classes; this aligns with work on in-group biases, which emerge when individuals perceive members from their own group differently from others and tend to assume that in-group members share positive values and characteristics.
3) the magnitude of the in-group bias is significantly stronger in the bottom classes than in the top classes; in other words, students in the bottom classes believe that their classmates perform relatively better compared to those in the top class than vice versa. This is consistent with evidence that in-group favoritism can be larger among low-status groups.
4) students rank themselves higher than the corresponding rankings coming from their peers and their teachers; this finding aligns with a cognitive bias also known as superiority bias, that arises when individuals overestimate their abilities.
5) the gap between a student’s self-assessment and that provided by her peers and teachers is larger among students with lower admissions scores. This finding is consistent with the Krueger-Dunning bias, where low performers are typically more overconfident, while high performers assess their skills more accurately.
To summarize, we observe five empirical patterns that seem to mitigate the adverse effects that being assigned to a lower-achieving class has on self-perceptions. Finally, in our analysis, we describe some suggestive evidence of small but positive effects of attending a top class on a high-stakes exam taken at the end of high school.
At least two other important questions have emerged while working on this project: (i) comparing these findings—both cognitive and non-cognitive outcomes—with other types of student allocation, particularly random class assignment; and (ii) further providing new insights into the nature and the determinants of ability tracking and the impacts on their achievements.
To address these questions, we have already complemented the existing data with additional survey data which, combined with rich administrative data, provide deeper insights into how teachers and principals make class allocation decisions—whether through tracking, randomization, or other methods.
The already published paper was written in collaboration with the initial team: Ofer Malamud (Northwestern University), Cristian Pop-Eleches, and Miguel Urquiola (both from Columbia University). In the new research project, we are also working with Andrei Munteanu from Université du Québec à Montréal's École des Sciences de la Gestion (preliminary title: The Determinants of Ability Tracking and Its Effects).
The paper "Self-, Peer-, and Teacher Perceptions under School Tracking" has been presented at numerous seminars and conferences, including the CESifo Conference on Economics of Education, McGill University, the Institute for Policy Research at Northwestern University, the University of Maryland, UCLA, and the University of Gothenburg. It is also available as an NBER working paper (NBER Working Paper No. w32892).
Both the PI (Andreea Mitrut) and the co-authors have the latest versions accessible on their webpages. Additionally, The Review of Economics and Statistics journal offers open access.
Project PI: Andreea Mitrut: https://sites.google.com/site/andreeamitrut/home
Publications:
Self-, Peer-, and Teacher Perceptions under School Tracking, forthcoming Review of Economics and Statistics Malamud, O., Mitrut, A., Pop-Eleches, C., Urquiola, M.
Work in progress:
The Determinants of Ability Tracking and Its Effects, Malamud, O., Mitrut, A., Munteanu, A., Pop-Eleches, C., Urquiola, M.
Teacher-Student Interactions and Child Outcomes, Malamud, O., Mitrut, A., Pop-Eleches, C., Urquiola, M.
Economists have focused on estimating the causal effect of tracking on cognitive skills as measured by academic achievement. Such studies often highlight a potential tradeoff: tracking may allow for more targeted teaching, but it can also deny lower-achieving students the positive peer effects that emerge from interactions with higher-achieving students. Additionally, tracking may stigmatize some students, hurting their self-perceptions (self-confidence, effort in school, and so on), which could have short and long-term consequences.
The main purpose of this project was to deepen our knowledge on the effects of how having access to a better class affect the teacher-student interactions and the (self-)perceptions of the students and their teachers on the students’ ability, academic performance and disruptive behavior. Understanding these interactions is important but remains understudied, largely because answering these questions in a credible, causal way is challenging due to data limitations and methodological difficulties.
In “Self-, Peer- and Teacher Perceptions under School Tracking” we have examined how assignment to a lower vs. higher track affects non-cognitive skills as captured by perceptions of self-confidence, academic effort, disruptive behavior, and self-esteem. We consider this question within a causal framework—we ask whether the assignment to different tracks causally impacts how students perceive themselves relative to their peers and how their peers and teachers perceive them.
Our project leverages a unique education system stratified by ability in Romania, offering an ideal framework for addressing methodological and data-related challenges in understanding the effects of tracking in education. In this system, the “top” class contains the students with the highest admission scores, while the “bottom” class contains those with the lowest. Classes always receive instruction separately in all subjects, although they cover the same curriculum, often delivered by the same teachers. These features enable us to isolate the role of student allocations from differences in curriculum and teacher quality.
We use two strategies to explore how student, peer, and teacher perceptions vary under school tracking. First, we use a regression discontinuity design to estimate the effect of being assigned to the top relative to the bottom class on student, peer, and teacher perceptions. Second, we use student fixed effects when directly comparing student perceptions to teacher perceptions, and when directly comparing peer perceptions of one’s own classmates to those of students in other classes.
An important part of our project development project is data collection. Besides using high-quality administrative data, we have developed a new original survey of teachers and students that allows us to tackle the methodologies and questions described above. This is an important tool that will also be used in several other (ongoing) projects.
To summarize, in this data collection we focused on a sample of schools who track students by ability and implemented an individualized survey asking students about their perceived relative standing across both top and bottom classes. The survey elicited students’ perceptions of themselves and their peers in the same class and in the other classes, along academic domains such as self-confidence in school, effort (e.g., does homework, pays attention in class), ability (e.g., understands hard concepts easily), and expected performance at the standardized exam, as well as on disruptiveness (e.g., harasses/disparages peers). We also surveyed all teachers who taught the same subject to both the top and bottom classes, asking them to rank the same students across the same domains as the students.
We show five findings:
1) being allocated to the top vs. the bottom class lowers a student’s perception of how she ranks relative to her peers in the same specialization; these negative impacts are in stark contrast to teachers’ perceptions;
2) along all dimensions, students rank their own classmates significantly higher than their peers from other classes; this aligns with work on in-group biases, which emerge when individuals perceive members from their own group differently from others and tend to assume that in-group members share positive values and characteristics.
3) the magnitude of the in-group bias is significantly stronger in the bottom classes than in the top classes; in other words, students in the bottom classes believe that their classmates perform relatively better compared to those in the top class than vice versa. This is consistent with evidence that in-group favoritism can be larger among low-status groups.
4) students rank themselves higher than the corresponding rankings coming from their peers and their teachers; this finding aligns with a cognitive bias also known as superiority bias, that arises when individuals overestimate their abilities.
5) the gap between a student’s self-assessment and that provided by her peers and teachers is larger among students with lower admissions scores. This finding is consistent with the Krueger-Dunning bias, where low performers are typically more overconfident, while high performers assess their skills more accurately.
To summarize, we observe five empirical patterns that seem to mitigate the adverse effects that being assigned to a lower-achieving class has on self-perceptions. Finally, in our analysis, we describe some suggestive evidence of small but positive effects of attending a top class on a high-stakes exam taken at the end of high school.
At least two other important questions have emerged while working on this project: (i) comparing these findings—both cognitive and non-cognitive outcomes—with other types of student allocation, particularly random class assignment; and (ii) further providing new insights into the nature and the determinants of ability tracking and the impacts on their achievements.
To address these questions, we have already complemented the existing data with additional survey data which, combined with rich administrative data, provide deeper insights into how teachers and principals make class allocation decisions—whether through tracking, randomization, or other methods.
The already published paper was written in collaboration with the initial team: Ofer Malamud (Northwestern University), Cristian Pop-Eleches, and Miguel Urquiola (both from Columbia University). In the new research project, we are also working with Andrei Munteanu from Université du Québec à Montréal's École des Sciences de la Gestion (preliminary title: The Determinants of Ability Tracking and Its Effects).
The paper "Self-, Peer-, and Teacher Perceptions under School Tracking" has been presented at numerous seminars and conferences, including the CESifo Conference on Economics of Education, McGill University, the Institute for Policy Research at Northwestern University, the University of Maryland, UCLA, and the University of Gothenburg. It is also available as an NBER working paper (NBER Working Paper No. w32892).
Both the PI (Andreea Mitrut) and the co-authors have the latest versions accessible on their webpages. Additionally, The Review of Economics and Statistics journal offers open access.
Project PI: Andreea Mitrut: https://sites.google.com/site/andreeamitrut/home
Publications:
Self-, Peer-, and Teacher Perceptions under School Tracking, forthcoming Review of Economics and Statistics Malamud, O., Mitrut, A., Pop-Eleches, C., Urquiola, M.
Work in progress:
The Determinants of Ability Tracking and Its Effects, Malamud, O., Mitrut, A., Munteanu, A., Pop-Eleches, C., Urquiola, M.
Teacher-Student Interactions and Child Outcomes, Malamud, O., Mitrut, A., Pop-Eleches, C., Urquiola, M.