PSYED 3472 - CAUSAL INFERENCE IN EDUCATIONAL RESEARCH
Many key questions in the field of education are framed causally. Do investments in full-day kindergarten pay off in terms of improved school readiness? Does project-based learning in mathematics and science increase the pipeline of students into STEM-related fields? Does the introduction of a generous merit-based scholarship program improve students' motivation to prepare rigorously for postsecondary education? Despite this causal framing, analytic tools commonly applied to questions such as these allow for statements about relationships but not about causation. For example, we may observe correlational evidence that communities with full-day kindergarten also have higher levels of school readiness. These same communities, however, may also serve children from higher-income families. Given students' backgrounds, their levels of readiness may have been unchanged by participation in full-day kindergarten.
In this course, we will focus on framing research questions with a causal lens and on research designs and analytic techniques that provide the tools for answering these key questions in a causal framework. Specifically, we will learn about research designs for drawing causal inferences, including randomized trials, regression discontinuity, differences-in-differences, instrumental variables, and propensity score and other matching techniques. Our learning will be grounded through the semester in reading scholarly articles in which these techniques are applied to questions in education. Assignments throughout the semester will include preparation for class participation, a referee report to critique the work of another scholar, and a final course project. At the start of the semester, students will be asked to identify an area of focus and potential sources of data for the final course project.