Critical Incident Technique

Description

Collects detailed accounts of specific events that were exceptionally good or bad, then classifies them to reveal what determines success or failure in an experience.

The Critical Incident Technique, developed by John Flanagan from aviation psychology work during the Second World War, collects detailed accounts of specific incidents in which behaviour or experience was notably effective or ineffective. Asking about concrete events rather than general opinions reduces vague answers and gives behaviourally specific data. Each incident is captured with its situation, actions, outcome and feelings, and incidents are then content-analysed and classified into categories that reveal the critical factors driving good and bad experiences. In UX, CIT works in interviews, surveys and diaries, and in reviews of support data. Rigour comes from a clear definition of what counts as critical, structured probes for each incident, collecting enough incidents for categories to stabilise (often 50 to 100 or more), and checking classification reliability.

Attributes

Loss type
Renamed
Stage
Discovery
Effort
Weeks

Steps

  1. Define "critical": Incidents that were either significantly more positive or more negative than the participant's baseline expectation. Not typical experiences — threshold-crossing events.
  2. Collection: "Think about the last [time period]. Tell me about a time when using [product] was particularly [difficult / helpful / frustrating / satisfying]." The temporal framing matters — recent incidents are more accessible and more accurately recalled.
  3. Structured probe: For each incident: (a) What was the situation? (b) What did you do? (c) What was the outcome? (d) How did you feel? (e) What would you have done differently?
  4. Diary integration: Deploy CIT as a diary trigger: "Tonight, if you had an experience with [product] that you would call particularly [good/bad], write it down using this template." This captures incidents in near-real-time rather than in retrospective interviews.
  5. Analysis: Code incidents for: type of event, trigger, participant action, outcome, and emotional response. Frequency analysis: which types of critical incidents are most common? Impact analysis: which types have the largest effect on overall satisfaction?

Sources

  • Research Methods in Psychology
  • Observing the User Experience (Goodman)
  • Universal Methods of Design (Martin)
  • Diary Methods
  • Lost Methods in UX — A Complete Treatise (Gopinath)
  • Flanagan JC (1954) The critical incident technique. Psychological Bulletin 51(4):327-358
  • Butterfield LD et al. (2005) Fifty years of the critical incident technique. Qualitative Research

Also known as

  • CIT
  • Critical incident interviews
  • Critical incident analysis

Categories

  • User Research
  • Interviewing & Elicitation