Mental Models

Diagrams mapping how users conceptually organize a domain, revealing their mental categories, causal relationships, and rules of inference compared to the system model.

Process

Key Fields

Question it answers	How do users think about this domain? What categories and relationships matter to them? Where does their understanding diverge from the system model?
Participants & timing	8-12 in-depth interviews · 3-4 weeks analysis and diagramming
AI compatibility	AI extracts domain concepts from transcripts and suggests groupings; interpretation of shared vs. individual mental models requires human judgment.
Output	Mental model diagrams (1-3 core models), conceptual vocabulary guide, expert vs. user model gaps analysis, interface/labeling recommendations

Use when

Designing for a complex or unfamiliar domain (finance, healthcare, scientific tools)
Information architecture or labeling is causing user confusion
Choosing between multiple possible information structures
Onboarding needs clarity on where users start their understanding

Skip when

Domain is simple and mental models are well established
Available data is surveys only (need detailed interviews)

Common Mistakes

Creating the expert model instead of the user model

Documenting how the system works rather than how users think. Show users' misunderstandings, intuitions, and mental shortcuts - not the "correct" model.

Not validating the model

Creating diagrams in isolation without testing them with users. Ask users to explain the model back to you; iterate based on what is missing or wrong.

Mixing multiple expertise levels into one model

Novices and experts have very different mental models. Create separate archetypes and show how understanding evolves across experience levels.