Wendy Chan

I am thrilled to share that, as of July 1st, I officially started as an associate professor (with tenure) at the Graduate School of Education at the University of Pennsylvania. Thank you to my family, friends and colleagues who supported me from the beginning. I am excited for this new chapter!

Thank you to the editor of JEE and to two anonymous reviewers for the helpful feedback.

We show in a series of simulation studies that the inclusion of weak proxies leads to increases in bias of the estimates.

We find that adding weak proxies improves assessments of generalization, leading to conclusions that a sample and population are similar.

In this paper, we examine the effects of augmenting data with weak proxy moderators on assessments of generalization and the estimation of population average treatment effects.

The Implications of Data Augmentation with Proxy Moderators for Generalization With the growth in experimental research in education, researchers and policymakers have become interested in understanding the mechanisms by which interventions work and for whom they work. Statis...

I am excited to share a new paper coming out in the Journal of Experimental Education. This was joint work with my students Jimin Oh, Chen Li, Jiexuan Huang and Yeran Tong.

www.tandfonline.com/doi/full/10....

A key tension, as highlighted by a reviewer, is when the average treatment effect (ATE) in the original population is the one of interest but we cannot estimate it well, but we can get a better estimate of the ATE in a smaller population (that may be of less interest).

A population can be redefined in different ways and in this study, we discuss the implications for bias reduction and precision when various methods are used.

My first post!

Excited to share our new paper on redefining populations for generalization. My students Jimin Oh and Kat Wilson and I learned a lot from this study, especially with the feedback we received from reviewers.

www.tandfonline.com/doi/full/10....