Independent Consultant

John Deke, PhD

I am an economist and nationally recognized expert in experimental and non-experimental impact evaluation methodology, Bayesian inference, and the interpretation of impact and descriptive findings. I bring more than two decades of experience designing and analyzing experimental, quasi-experimental, and descriptive evaluations for OPRE, IES, CMS, and OPA.

As a Senior Fellow at Mathematica, I co-led the development of BASIE (Bayesian Interpretation of Estimates), an approach now used across federal agencies to provide transparent, probability-based interpretation of both impact estimates and descriptive differences between groups or across time. BASIE empowers more useful, comprehensive, and integrated interpretation of heterogeneous research findings.

Areas of Expertise
Bayesian inference Program evaluation Causal inference Impact evaluation Applied statistics Education research Health policy
John Deke

BASIE Workbench

BASIE Workbench is a free, browser-based tool that applies the BASIE framework to a portfolio of impact estimates. Upload your own estimates — or explore the pre-loaded example — to see how Bayesian reinterpretation changes the picture. The tool produces posterior probabilities, credible intervals, and a predictive distribution for the true effect in a comparable future study.

The pre-loaded example illustrates the PIP Foundation: a fictional grant-making organization that discovers how the standard significance-based approach to picking winners is systematically misleading — and how BASIE provides a better path forward.

Open BASIE Workbench →
Built with Chart.js · Runs entirely in your browser · No data is uploaded or stored

About BASIE

BASIE (Bayesian Interpretation of Estimates) replaces the question "Is this statistically significant?" with decision-relevant questions: What is the probability this program truly works? What is the probability the effect exceeds a meaningful threshold? BASIE combines prior knowledge about the plausible range of effects with the current study's estimates to produce probability statements that map directly onto decision criteria.

BASIE is described in detail in the resources below and has been applied across evaluations sponsored by IES, OPRE, CMS, and other federal agencies.

OPRE Report → IES Toolkit → Gelman Blog →
Journal Article 2023

A Strong Case for Rethinking Causal Inference

Deke, John
Journal of Research on Educational Effectiveness
doi.org/10.1080/19345747.2023.2203683 →
Journal Article 2022

Bayesian Interpretation of Cluster-Robust Subgroup Impact Estimates: The Best of Both Worlds

Lipman, Erin R., Deke, John, and Finucane, Mariel M.
Journal of Policy Analysis and Management, 41(4), 1204–1224
Journal Article 2021

Asymdystopia: The Threat of Small Biases in Evaluations of Education Interventions That Need to Be Powered to Detect Small Impacts

Deke, John, Thomas Wei, and Tim Kautz
Journal of Research on Educational Effectiveness, 14(1), 207–240
Report 2022

The BASIE (BAyeSian Interpretation of Estimates) Framework for Interpreting Findings from Impact Evaluations: A Practical Guide for Education Researchers

Deke, J., Finucane, M., & Thal, D.
National Center for Education Evaluation, Institute of Education Sciences, U.S. Department of Education
Download PDF →
Report 2019

Moving Beyond Statistical Significance: The BASIE (BAyeSian Interpretation of Estimates) Framework for Interpreting Findings from Impact Evaluations

Deke, J., and Finucane, M.
OPRE Report #2019-35. Office of Planning, Research, and Evaluation, ACF, U.S. DHHS
View Report →
Report 2023

Using Bayesian Meta-Analysis to Explore the Components of Early Literacy Interventions

Walsh, E., Deke, J., Robles, S., Streke, A., and Thal, D.
National Center for Education Evaluation, Institute of Education Sciences (WWC 2023008)
whatworks.ed.gov →
Journal Article 2017

The WWC Attrition Standard: Sensitivity to Assumptions and Opportunities for Refining and Adapting to New Contexts

Deke, John, and Hanley Chiang
Evaluation Review, 41(2), 130–154
Journal Article 2016

Design and Analysis Considerations for Cluster Randomized Controlled Trials That Have a Small Number of Clusters

Deke, John
Evaluation Review, 40(5), 444–486
Report 2015

Preview of Regression Discontinuity Designs Standards

Deke, J., Cook, T., Dragoset, L., Reardon, S., Titiunik, R., Todd, P., Van der Klaauw, W., and Waddell, G.
Institute of Education Sciences, U.S. Department of Education
Journal Article 2003

Study of the Impact of Public School Spending on Postsecondary Educational Attainment Using Statewide School District Refinancing in Kansas

Deke, John
Economics of Education Review, 22(3), 275–284
jdeke73@gmail.com
📍
Overland Park, Kansas
📄
BASIE Workbench