Zarni Htet
Research Analyst, MDRC Center for Data Insights

Htet is a data analyst at MDRC’s Center for Data Insights. Since joining MDRC, in the fall of 2018, Zarni has co-created a reusable predictive analytics assessment tool (PAAT); currently, he is working on its methodological and codebase expansion on machine learning model fairness and interpretability. He is also co-authoring an internal guide on predictive modeling in the context of social policy programs. One of Htet’s favorite projects involved collaborating with researchers from a state agency to apply predictive modeling to improve the agency’s program outcomes while being rigorous about data and model ethics. He also worked on developing MDRC’s first public R Shiny app for statistical power estimation.

Before joining MDRC, Htet was a Data Science for Public Good Fellow at the Social Decision Analytics Lab in Ballston, MD, where he worked on data deduplication, probabilistic record linkage, and synthetic data creation, among other topics. Htet has a graduate degree in applied statistics from New York University, where his course work and independent projects included causal inference and multilevel models.

  • MDRC Publications


      A Toolkit for State and Local Agencies on How to Access, Link, and Analyze Unemployment Insurance Wage Data

      November, 2022
      Edith Yang, Sharon Zanti, T.C. Burnett, Richard Hendra, Dennis Culhane, Zarni Htet, Della Jenkins, Camille Préel-Dumas, Electra Small

      Temporary Assistance for Needy Families (TANF) agencies are increasingly focused on using administrative data to assess how well programs are working and to inform policies and best practices. This toolkit was created to help TANF professionals develop more robust practices using employment data for program monitoring, reporting, and evaluation.

      June, 2022
      Kristin Porter, Zarni Htet, Kristen Hunter, Luke Miratrix

      Multiple testing procedures reduce the likelihood of false positive findings, but can also reduce the probability of detecting true effects. This post introduces two open-source software tools from the Power Under Multiplicity Project that can help researchers plan analyses for randomized controlled trials using multiple testing procedures.


      Evidence from Child First

      May, 2022

      This brief presents results from a proof-of-concept exercise that examined the potential benefits of using predictive analytics to improve service delivery by Child First, a program that provides therapeutic support to families with young children. The information may be useful for other organizations interested in implementing these cutting-edge tools.

  • Other Publications

  • Projects