Research & publications
AI evaluation and applied statistics, with a focus on the quality of the evidence behind a decision.
Explore the work ↓Research record
Publication status is listed for each work.
2024 · arXiv preprint
How should the quality of an AI-generated summary be evaluated?
Evaluates GPT-generated summaries for quality, accuracy, and consistency. This work contributes to assessing generated text before it is relied on in an analytical workflow.
Manuscript · Prepublication
What changes when sentiment and bias become part of the evaluation?
Develops an iterative approach to evaluating bias in summarization, extending the focus beyond fluent output to the way a summary represents its source.
Applied statistics · Research project
Which predictors help explain ten-year coronary heart disease risk?
Models risk using Framingham data, connecting statistical modeling with interpretable predictors. An applied analysis of how quantitative evidence can inform understanding of risk.
Putting the method to work
My independent research system preserves investment theses and assumptions, monitors new evidence, and compares outcomes with prior expectations. It applies the same discipline to research over time: keep a record that can be tested.
Academic foundation
In progress
University of Colorado Boulder
2025
University of Kansas
2024
University of Kansas
2007
Northern Illinois University
Let’s connect
Let’s talk about decision science, analytical leadership,
and applied AI that earns its place in the workflow.