After a round of interviews, the team has dozens of observations and several competing explanations. One researcher sees a problem with trust, another with timing, and a third keeps returning to a detail nobody else recorded. Affinity mapping gives those interpretations something concrete to work against by arranging individual notes into groups based on their relationship.
Also called affinity diagramming, the method helps people organise research material or ideas and examine possible patterns. In research synthesis, the notes should remain connected to their sources. The arrangement is an analytical aid, not evidence that a finding becomes true once enough people agree on a heading.
Make the notes intelligible before grouping them
A useful note contains a specific observation and enough context to understand it. “Privacy concern” is difficult to evaluate. “Participant 4 declined contact access because the phone is shared with a partner” gives the team a behaviour, an explanation and a source to revisit.
Keep different kinds of material distinguishable. A quotation, an observed action, a researcher’s interpretation and a proposed solution can all be relevant, but mixing them without labels makes speculation look like a participant’s account. Use participant identifiers rather than unnecessary personal information, and retain links to the underlying session where access is appropriate.
Nielsen Norman Group’s guide to affinity diagramming describes grouping related material and naming the resulting clusters. The technique can support collaborative synthesis, but the quality of the notes determines how much meaning the grouping can preserve.
Let the grouping expose disagreement
Begin by reading and arranging notes, then discuss the relationships people see. A period of individual consideration can give quieter team members room to form an interpretation before the discussion gathers around one explanation.
Some notes will belong in more than one place. Others will resist the emerging categories. Keep them visible rather than forcing a tidy board. In a fictional study of a shared calendar, a note about hiding an appointment might concern both privacy and control over notifications; treating it as only one of those could narrow the design response prematurely.
Label groups at the level the evidence supports. A topic such as “notifications” may be a useful intermediate category. A finding such as “people need to separate private reminders from shared events” requires a further interpretive step, supported by the notes and their original context. Naming every cluster as a confident conclusion can disguise work that has not yet been done.
Check the pattern against the underlying accounts
Count sources when counts are relevant, not just sticky notes. Ten notes may come from one talkative participant or from a recording that one researcher documented more densely than the others. Neither the size of a cluster nor the number of votes it receives establishes prevalence in the wider audience.
Return to contradictory cases and examine whether the proposed interpretation still holds. Compare what the notes say with the full interview or observation before turning a group into a recommendation. For more systematic interpretation across a dataset, consider the requirements of thematic analysis, rather than treating a workshop as a substitute for the entire process.
The map should leave the team with supported interpretations, unanswered questions and a route back to the evidence. An empathy map or journey map may help communicate part of that understanding later, but the next format should follow what needs explaining, not a desire to produce another board.