UX Glossary · R
Research synthesis
Turning raw research data into themes and decisions, which is where most of the value is created or lost.
Synthesis is the step between collecting data and knowing something: reviewing notes and recordings, identifying patterns, and forming findings that can be acted on.
It is the least visible part of research and the part where quality is decided. Ten excellent interviews synthesised badly produce a slide of quotes; ten adequate ones synthesised well produce a decision.
The method
Break the data into individual observations, group them by similarity, and name the groups only after they form. This is affinity diagramming, and the discipline is in resisting the urge to name categories first, because pre-set categories can only confirm what you already believed.
What makes a finding a finding
Support from several participants, ideally recruited independently. One person’s frustration is a quote. The same frustration from six of eight is a pattern.
Findings should also be stated as observations rather than solutions. “Users could not tell which numbers were editable” is a finding. “Add an edit icon” is a proposal, and collapsing the two hides the reasoning from everyone who reads the report later.
In practice
Synthesise as a group where possible. Everyone who sorts the notes has read the raw data, which makes findings considerably harder to wave away in the meeting where they are inconvenient.