Here you can find my previous and ongoing research projects, including publications, working papers, research in progress, and policy reports.
Brand, J.E., Xu, J., Koch, B., & Geraldo, P. (2021) “Uncovering Sociological Effect Heterogeneity Using Tree-Based Machine Learning." Sociological Methodology, 51(2):189-223
Geraldo, P. & Brand, J.E. (2020) “Causal Inference." Oxford Bibliographies in Sociology, Ed. Lynette Spillman. New York: Oxford University Press.
Gonzalez, K., Geraldo, P., Estay, M., & Franklin, C. (2019) “Solution-Focused Brief Therapy for Individuals With Alcohol Use Disorders in Chile." Research in Social Work Practice, 29(1):19-35
Geraldo, P. (2018) “Secondary Schools’ Vocational Track and the Reproduction of Social Inequality." National Thesis Award: Thinking the Youth 2015. Santiago, INJUV (in Spanish)
Maldonado L., & Geraldo P. (2018) “Fixed Effects Regression and Effect Heterogeneity." In: Giesselmann M., Golsch K., Lohmann H., Schmidt-Catran A. (eds) Lebensbedingungen in Deutschland in der Längsschnittperspektive. Springer VS, Wiesbaden
Geraldo, P. (2013) “The Crisis of Chilean Marxist Historiography." Proceeding of 5th International Congress in Philosophy and Theory of History. Viña del Mar, UAI (in Spanish)
Geraldo, P. (2012) “Catholic Church Members during Popular Unity Government." Simon Collier Seminar. Santiago, History Department, Universidad Catolica de Chile (in Spanish)
Abstract: Causal inference with observational data critically relies on untestable and extra-statistical assumptions that have (sometimes) testable implications. Well-known sets of assumptions that are sufficient to justify the causal interpretation of certain estimators are usually called identification strategies. However, empirical research do not perfectly map into identification strategies. In this paper, I contend that causal diagrams can help researchers to assess the quality of evidence without excessively relying on these research templates. First, I offer a concise introduction to causal inference frameworks. Then I describe the current practice of applied causal inference using graphical models, discussing some controversies around their usefulness. Third, I show a series of examples using Directed Acyclic Graphs to encode the substantive assumptions that researchers invoke in their empirical work, demonstrating that the distinction between identification strategies is not clear-cut. Finally, I provide some recommendations for practitioners on how to assess the credibility of a study causal argument to be more sensible to a study’s particular setting.
Geraldo, P., “Tracking before tracking? Academic performance and expectations in secondary vocational schools in Chile”
Stratified academic achievement and educational expectations are two major sources of inequality in educational attainment. Drawing from theories of horizontal stratification in education, I argue that exposure to a school that offers exclusively vocational education has a detrimental impact on students’ academic performance and their expectation of future academic attainment, even before the tracked curriculum begins. I test this hypothesis using administrative data and national standardized tests in Chile. Relying on conditional ignorability and parallel trends assumptions, I found evidence of a detrimental effect of attending vocational schools during 9th and 10th grades on students’ mathematics performance and their educational expectations, even though the differentiated curriculum officially starts in 11th grade. Furthermore, the effect on academic performance seems to be more pronounced among students with previous higher performance and more ambitious educational expectations.
I participated in the elaboration of the following policy reports. All were collective efforts and as such they have institutional authorship ascribed.