Luke Miratrix

The purpose of this grant is to provide guidance to applied researchers on the best estimands and estimators of those estimands for analyses of cluster randomized control trials (CRTs).

The purpose of this grant is to provide machine learning tools to improve the statistical power of an education impact analysis that relies on human-scoring of text passages written by students. Machine learning can be particularly useful in studies that have more text data than time, money, and/or coders to score all of the text data by human raters. By using the text and scores provided by the available human raters to train a machine learning algorithm, computer-scoring would then be used...

When a randomized control trial is infeasible in an education setting, researchers can use quasi-experimental research designs. A commonly used approach is to use repeated observations of aggregate data, known as group panel data, before and after a new policy or intervention is put in place. To estimate the effects, researchers typically rely on either a comparative interrupted time series (CITS) or, increasingly, the synthetic control method (SCM), but there is not a clear set of best prac...

The purpose of this project is to explore and analyze data collected from an evaluation of three professional development interventions in science. There is wide variation in the ways professional development programs improve teachers' experience and skills, making it difficult to determine which features of professional development interventions are associated with better teacher or student outcomes. To obtain a better understanding of the contributions of teacher professional development p...

The goal of this project is to create a framework that provides applied researchers with a set of practical tools and clearly lays out all the relevant assumptions for assessing treatment effect variation. When testing the impact of an intervention or treatment, education researchers are increasingly asking questions such as: (1) To what extent does the effect vary across units? (2) What accounts for this variation?