Bayesian Research Analyst – Research Lab, Northwestern University

Sep 11, 2024

About the Global Poverty Research Lab
The Global Poverty Research Lab (GPRL) is a research center based at Northwestern University that generates empirical evidence on the effectiveness of policies and programs in more than 10 countries in the Americas, sub-Saharan Africa, China, and Southeast Asia. Our projects examine the interaction between poverty and topics such as finance, entrepreneurship, education, gender, psychological well-being, agriculture, and the environment.
One main goal of GPRL is to produce evidence that can inform public policy. Recognizing that policy decisions cannot rely solely on individual studies, the lab emphasizes the importance of meta-analyses—studies that aggregate evidence from multiple sources to draw conclusions about the effectiveness of various interventions across different contexts and implementations. The core methodology for these meta-analyses is Bayesian Inference, with a particular emphasis on Bayesian Hierarchical Models.

Our Team
GPRL is co-directed by Dean Karlan, Christopher Udry, and Nancy Qian, professors at the Kellogg School of Management and the Weinberg School of Arts and Sciences’ economics department at Northwestern University. Other Kellogg professors who are affiliates and Lab investigators include Andrew Dillon and Lori Beaman. GPRL’s research managers, research analysts, and administrative staff support our investigators’ projects and work closely with their coauthors at universities, NGOs, and research institutions around the world.
GPRL also works in close collaboration with research teams at Innovations for Poverty Action (IPA) to coordinate field data collection, disseminate evidence, and create shared training and research resources.

Job Summary
GPRL is hiring a full-time Research Analyst to work on projects that utilize Bayesian methods, especially hierarchical models, to summarize results across different studies. For instance, in one of our key projects, we analyze the impact of cash transfer programs by combining data from 114 studies conducted in 34 countries. Similarly, other projects focus on evaluating the effects of microcredit, entrepreneurship programs, savings reminders, and more.

In this role, the Research Analyst will have the opportunity to use large datasets to apply and improve their expertise in Bayesian methods while contributing to research that is informative for public policy. Additionally, the RA will have access to Quest, Northwestern’s high-performance computing cluster, providing the opportunity to work with a powerful server equipped with a high number of cores and extensive RAM.

This position offers a unique chance to engage with significant policy issues, especially for low and middle-income countries, and to help shape evidence that can be informative to policymakers.

Core responsibilities:

– Writing Bayesian Hierarchical Models for the joint analysis of several datasets, for example in meta-analyses, including coding, reporting and interpreting the results
– Conducting analysis using advanced statistical and econometric tools
– Cleaning survey or administrative datasets and harmonizing variables across different sources, preparing them for analysis
– Presenting analysis to investigators and incorporating feedback into subsequent analysis
– Conducting literature reviews to identify state-of-the-art methods for conducting the analyses
– Preparing tables and figures with empirical results for publications and presentations and assisting with paper revisions for peer-reviewed journal submissions

Additional responsibilities as needed:
– Supervising undergraduate RAs as required
– Supporting research staff with lab-wide management tasks including mentoring new RAs and creating research and training resources

Minimum

Competencies:

(Skills, knowledge, and abilities.)
– Strong programming skills in R or another programming language to implement Bayesian methods
– Knowledge of Bayesian Inference, including writing and interpreting models and their output
– Understanding of simulation techniques for solving Bayesian Inference models,…