Ever wondered how to spot test questions that have been compromised or students who got a sneak peek before their exam? In this blog, I walk through a new statistical model that combines both how fast and how accurately people answer test items to flag compromised questions and examinees with prior knowledge—all at once. You’ll find a full walkthrough: model explanation, code, diagnostics, and example results using simulated data. If you’re into psychometrics, cheating detection, or just love seeing R and Stan in action, you’ll get a kick out of this hands-on guide!
For attribution, please cite this work as
Zopluoglu (2025, June 11). Cengiz Zopluoglu: Deterministic Gated Hierarchical Item Response Model to Simultaneously Identify Compromised Items and Examinees with Item Preknowledge Using Both Response Accuracy and Response Time Data. Retrieved from https://github.com/czopluoglu/website/tree/master/docs/posts/duolingo_dghirt/
BibTeX citation
@misc{zopluoglu2025deterministic,
author = {Zopluoglu, Cengiz},
title = {Cengiz Zopluoglu: Deterministic Gated Hierarchical Item Response Model to Simultaneously Identify Compromised Items and Examinees with Item Preknowledge Using Both Response Accuracy and Response Time Data},
url = {https://github.com/czopluoglu/website/tree/master/docs/posts/duolingo_dghirt/},
year = {2025}
}