AI evaluation
For teams shipping systems that make claims about people.
- Dataset and benchmark design and collection
- Model evaluation
- Generative psychometrics
Psychology · Measurement · Machine learning
I work where psychological measurement meets machine learning. I build datasets, benchmarks and evaluations for systems that make claims about human behavior.
Associate Professor of Psychology at the University of Kansas, where I direct the Affective Communication & Computing Lab
I trained as a clinical psychologist (PhD, University of Pittsburgh, 2018), and my research is about measuring things that resist measurement: emotion, clinical state, the texture of a conversation. That work has appeared in Annual Review of Clinical Psychology, Journal of Consulting and Clinical Psychology and Affective Science, and it has produced open-source tools that other researchers use.
Through Girard Consulting I bring the same standards to applied work for teams building systems that reason about people. The research itself, and the students and staff who do it with me, live at the AffCom Lab.
For teams shipping systems that make claims about people.
For work that has to survive review.
For teams who want the capability in-house.
Tell me what you are trying to measure, and what would count as getting it right.
All 72, with preprints, data and code where they exist.
Corpora and benchmarks from the lab, and how to get them.
R packages, research tools and open course materials.