Spotlight: Recent Graduate

Karina G. Diaz

Pronouns: she/her/hers

Tell us a bit about yourself.

My name is Karina Diaz, I’m a research specialist at the Consortium for Policy Research in Education (CPRE) at the University of Pennsylvania. I’m primarily responsible for the design and execution of quantitative analyses for funded research studies, including efficacy evaluations and large-scale projects with experimental and quasi-experimental designs. Prior to this, I collaborated as a research assistant at the STEPP Center at the Institute for Policy Research (IPR) at Northwestern University.

How did you get into systematic review and/or meta-analysis research?

While working as a researcher in Chile, I wanted to understand what the most effective type of professional development educational programs were. To my surprise, I found a great number of studies, which provided a variety of different recommendations. I found this to be both exciting and confusing. I finally came across a meta-analysis that addressed this research question and that proved to be very useful in clarifying conclusions previously published on this topic.

Years later, while I was working on my master’s degree at Teachers College, I took a statistics class with Beth Tipton. She mentioned how her field of study was meta-analysis and explained many of the statistical challenges of systemic reviews. I started participating in her students’ research group, where we reviewed different theoretical and practical articles on meta-analysis.

What work do you do now that is related to systematic review and/or meta-analysis?

My research interests include methods for meta-analysis, design and analysis of experiments, and advances to missing data approaches. Much of my methodological work is motivated by statistical inference questions arising in social research, particularly in educational research.

I’m currently working on methods to handle missing data of covariates in meta-regression and on the development of effect size estimators used in systematic reviews. I’m also collaborating on a meta-analysis that seeks to evaluate the utility of Emotional approach coping (EAC) in response to a variety of stressors (i.e., general, health-related, trauma-related, and interpersonal). These findings may help determine the most beneficial applications of EAC and serve as a framework for evaluating the utility of other coping mechanisms.

What do you love about systematic review and/or meta-analysis?

One of the things that I value the most about systematic reviews is that they provide clarification on conclusions previously published on a particular topic of interest. It does this by increasing the statistical power through combined analysis. Meta-analysis results can then help policymakers and practitioners make evidence-based decisions about specific interventions, programs, and practices.

Systematic reviews are not only used to aggregate evidence relating to a specific topic, but also to make clear what is not known and, thus, to direct new primary research into areas where there might be a gap in the body of knowledge.

What advice do you have for graduate students and early career researchers about working in systematic review and meta-analysis research?

Conducting a meta-analysis is an extensive, meticulous process that may take months to complete. Before staring a meta-analysis, I would suggest identifying the question that you want to answer. Understanding if there is a gap in the literature and whether a systematic review is needed. Second, make sure your meta-analysis includes both published and unpublished sources for data collection. Third, explore missingness in your data and select the appropriate approach to handle missing data. I would recommend implementing model base methods instead of ad hoc approaches, such as complete-case analysis and shifting units of analysis.

Finally, I would encourage practitioners to provide raw data underlying their systematic review. Providing access to raw data could contribute to the advance of science and the improvement of research synthesis practices.