All frameworks and cheat sheets

Reference table

AI feature research focus by trust and stakes

Table placing AI features by the trust they need and their stakes, with what to research in each case, from perceived usefulness to calibrated confidence.
Find your feature's cell before you plan the study. A 5-person test suits one cell and is risky in another.

A two-by-two table for planning research on AI features: how much trust the feature needs and how high the stakes are, with an example feature and the research focus for each cell. Use it to choose the method and sample size before a study, since a light usability test suits low-stakes suggestions but not diagnostic advice. Koji runs the interviews and concept tests behind each cell.

Where this appears

What's in this reference table

Low stakesHigh stakes
Low trust neededTone suggestions, emoji recs: test perceived usefulnessAutocomplete in medical notes: test error recovery
High trust neededAI-written tweet drafts: test voice fidelity and edit costDiagnosis, financial advice: test calibrated confidence and human-in-the-loop limits

Use this image

Free to use in your docs, decks and posts under CC BY 4.0, as long as you credit Koji with a link to this page. To embed it, paste this into your page:

html
<a href="https://www.koji.so/visuals/ai-feature-research-focus-by-trust-and-stakes"><img src="https://www.koji.so/visuals/ai-feature-research-focus-by-trust-and-stakes/ai-feature-research-focus-by-trust-and-stakes.png" alt="Table placing AI features by the trust they need and their stakes, with what to research in each case, from perceived usefulness to calibrated confidence." width="1200" height="630" /></a>
<p>Source: <a href="https://www.koji.so/visuals/ai-feature-research-focus-by-trust-and-stakes">AI feature research focus by trust and stakes</a>, Koji</p>