Questions

Research themes

Questions first. Categories later.

Four themes organize the work. A project can belong to more than one theme.

What changes when evidence meets criteria, priors, persuasion, or a rating scale?

Bayesian modeling · signal detection · misinformation

What do models represent, and what gets hidden when a picture looks convincing?

NLP · embeddings · interpretability

Can an unusual question survive contact with messy data?

Visualization · exploratory analysis · caveats

What small instrument would make the next question easier to inspect?

Open source · reproducibility · local tools