Primary Outcomes (explanation)
The main objective of this study is to provide causal evidence on discrimination experienced by immigrant students and to evaluate the impact of two low-cost interventions that aim to reduce teachers’ bias in grading against this group. The first one is based on an informative video while the second one provides IAT feedback. We focus on Ecuador, a country that has received sudden and increasing inflows of Venezuelan immigrants over the past five years.
First, we measure whether there exists a de-facto bias against immigrant students by exploiting administrative records that allow us to compare native and immigrant students’ grades based on exams graded by their teachers, controlling by their performance on anonymous standardized national exams (Ser Estudiante, Ser Bachiller). Complementarily we design and administer a survey in order to construct a measure of teachers’ explicit bias towards immigrant students through self-reported attitudes. The survey is administered online before the interventions. It allows us to construct a psychometric measurement instrument based on standard questions from the Gallup's Migrant Acceptance Index and the World Values Survey. The survey collects information on teachers’ beliefs about the performance and aspirations of migrants compared to natives as well as from other minority or vulnerable groups. We expect that not focusing exclusively on migrants will prevent teachers from reinforcing their attitudes towards them before they receive treatment. Additionally, we gather teachers’ demographic and socioeconomic information that we use in our estimations as controls.
After the survey, we proceed with the first intervention. We present a 2-minute informative video to the teachers who were randomly allocated to treatment. Notice that randomization takes place at the school level rather than at the teacher level to avoid contamination between teachers. The treatment takes place before the semester ends which allows us to estimate its impact on the end-of-term grades of the first semester (end of September) by comparing treatment and control groups. We are also able to explore treatment effect heterogeneity by exploiting the information on beliefs gathered previously in the survey.
The second part of this study takes place at the beginning of the second semester and follows closely the study of Alesina et. al. 2018. We develop an Implicit Association Test (IAT) that allows us to measure teachers’ implicit bias. This computer-based test reduces the influence of social desirability bias in self-reported replies (Greenwald et al., 2009, Alesina et. al. 2018) by measuring the difference in reaction times when teachers are asked to associate positive and negative attributes with Venezuelan (immigrant) and Ecuadorian (native) representations. We then explore to what extent the difference in grading is associated with implicit stereotypes and compare how they relate to the explicit bias identified in the survey.
Finally, the second intervention consists of providing feedback via e-mail on the IAT results to a randomized group of teachers. Using the second-semester grades as an outcome, we will evaluate whether generating awareness of each one’s stereotypes can reduce discriminatory behavior. We randomize at the school-level again among the teachers that received the first treatment and the teachers that were in the first control group, which would allow us to i) measure the impact of both interventions combined and ii) to determine which one is more effective at reducing teachers’ bias in grading against immigrant students.
The final outcome to evaluate would be the grades at the end of the academic year (February 2023), in which we will able to evaluate both interventions.
References:
Alan, S., Duysak, E., Kubilay, E. & Mumcu, I., 2021. Social Exclusion and Ethnic Segregation in Schools: The Role of Teacher’s Ethnic Prejudice.
Alesina, A., Carlana, M., La Ferrara, E. & Pinotti, P., 2018. Revealing Stereotypes: Evidence from Immigrants in Schools. NBER Working Paper Series, Issue 25333.
Bertrand, M., & Duflo, E. (2017). Field experiments on discrimination. Handbook of economic field experiments, 1, 309-393.
Carlana, M., 2019. Implicit Stereotypes: Evidence from teachers´gender bias. The Quarterly Journal of Economics.
Castro, J. y otros, 2010. Desigualdad de Oportunidades educativas y segmentación laboral en la población de 15 a 29 años de Brasil y Colombia según autoclasificación racial, 2010: Universidad del Valle.
Gertler, P. y otros, 2016. Impact Evaluation in Practice. Washington D.C: World Bank Group and Inter-American Development Bank.
Khandker, S. R., Koowal, G. B. & Samad, H. A., 2010. Randomization. En: Handbook on Impact Evaluation. Quantitative Methods and Practices. Washington D.C.: The World Bank, pp. 33-52.
Greenwald, A. G., Nosek, B. A., and Banaji, M. R. (2003). Understanding and using the Implicit Association Test: I. An improved scoring algorithm. Journal of personality and social psychology, 85(2):197.
Nosek, B. A., Greenwald, A. G. & Banaji, M. R., 2007. The Implicit Association Test at age 7: A methodological and conceptual review. En: Automatic processes in social thinking and behavior. s.l.:Psychology Press, pp. 265-292.
Peña, Y., 2016. Evaluación del prejuicio racial implícito (IAT): Caso República Dominicana. Ciencia y Sociedad, pp. 77-111.
Rodríguez, C., Cárdenas, J. C., Oviedo, J. D. & Villamizar, S., 2013. La discriminación racial en el trabajo. Un estudio experimental en Bogotá, Bogotá: Centros de Estudios de Derecho, Justicia y Sociedad, Dejusticia.
Webber, S. & Prouse, C., 2018. The New Gold Standard: The Rise of Randomized Control Trials and Experimental Development. Economic Geography, pp. 166-187.