In general, there are three types of persona: proto-personas, qualitative personas, and statistical (or mixed-method) personas.
Proto-personas can spell out the assumptions a team has about users in the absence of research. Because they're not research-based, their prone to inaccuracy, but can be better than nothing in organizations with a low level of UX maturity or in an Lean UX environment where frequent testing and iteration will take place.
Qualitative personas are based on qualitative discovery research and document patterns in user needs, pain points, workflows, and tasks. Qualitative personas strike a nice balance between effort—they may require as few as 5 interviews per user group—and value. However, due to a small sample size, they might miss out some perspectives. Moreover, they may not indicate how representative an identified segment might be.
Statistical or mixed-method personas build on qualitative personas. Statistical personas use quantitative methods, like surveys, to validate qualitative findings and identify statistical patterns. This adds a veneer of scientific rigour to personas and provides assurance that the personas are statistically representative. However, the creation of statistical personas is far more time consuming than the other methods and may yield only incremental benefits.
- Methods to overcome the limitations of personas
- Landscape of UX research methods
- People fill information gaps with assumptions
- Personas alone can't help teams achieve empathy
- Personas lack rich context
- Personas may create space for assumptions
Laubheimer, Page. “3 Persona Types: Lightweight, Qualitative, and Statistical.” Nielsen Norman Group (blog). Accessed October 12, 2020. https://www.nngroup.com/articles/persona-types/.