Critical Incident Technique: Definition, Steps and Research Uses

The critical incident technique (CIT) is a qualitative research method in which participants or observers report specific events, called incidents, so that the researcher can work out what helped or hindered an activity. It was set out by John C. Flanagan in 1954 and is still used across psychology, nursing, education and user research.

Short answer: what is the critical incident technique?

The critical incident technique is “a set of procedures for collecting direct observations of human behaviour in such a way as to facilitate their potential usefulness in solving practical problems and developing broad psychological principles” (Flanagan, 1954, p. 327). In practice a researcher collects concrete incidents, groups similar ones into categories, and reports those categories. It records specific behaviour and events rather than general opinions.

Definition and scope

Flanagan developed the technique for job analysis, to identify the critical requirements for job success, and it grew out of the US Army Air Forces Aviation Psychology Program of the Second World War (Butterfield et al., 2009). Flanagan defined an incident as “any observable human activity that is sufficiently complete in itself to permit inferences and predictions to be made about the person performing the act” (1954, p. 327).

Butterfield and colleagues describe five features that distinguish CIT from other qualitative methods: a focus on critical events or factors that promote or detract from effective performance; origins in industrial and organizational psychology; data gathered mainly through interviews; analysis that sets a frame of reference, forms categories and decides how general or specific to be; and a report in the form of categories with operational definitions and self-descriptive titles (Butterfield et al., 2009). The same authors say CIT is exploratory and suits little-understood events, factors or psychological constructs, because it asks what helps or hinders in a particular experience.

What would be a critical incident?

In Flanagan’s sense, a critical incident is a reported event that made a difference to the outcome of an activity and is complete enough to support inference. An example from Flanagan’s job-analysis tradition is a first-hand report of a satisfactory or unsatisfactory execution of an assigned task, where success or failure was determined by specific reported causes (Flanagan, 1954, p. 329, as quoted by the University of Warwick). Critical incidents are not always dramatic. Tripp (1993) argues that most are “straightforward accounts of very commonplace events” that become critical through analysis.

What is a critical methodology?

“Critical” in critical incident technique refers to the incident making a difference to the outcome being studied, not to critical theory or critique of a discipline. A “critical methodology” in the sense people usually search for here means the method that works from critical incidents: CIT, and its extension, the Enhanced Critical Incident Technique (ECIT).

How the method works: Flanagan’s five steps

Butterfield et al. (2009) summarise Flanagan’s five steps and show how each is carried out in an interview study.

1. Ascertain the general aims of the activity. The researcher answers two questions: what is the objective of the activity, and what is the person who engages in it expected to accomplish.
2. Make plans and set specifications. This means defining the types of situations to be observed, deciding their relevance to the general aim, understanding the extent of the effect an incident has on the aim, and deciding who will make the observations. It includes building an interview guide and training interviewers so that procedures are consistent.
3. Collect the data. Interviews, either in person or by telephone, are the main vehicle for psychological topics. Flanagan advocated expert observation, which is often not possible when the subject is a person’s own experience. Interviewing continues until exhaustiveness (redundancy): no new incidents and no new categories appear.
4. Analyse the data. The researcher determines the frame of reference, forms categories by grouping similar incidents, and decides the level of generality or specificity for reporting. Butterfield et al. note that analysis is considered the most difficult step.
5. Interpret and report. Results are presented as categories with operational definitions, supported by credibility checks.

Outside academic psychology the same logic is applied more lightly. The Usability Body of Knowledge describes CIT as a method of gathering facts (incidents) from domain experts or less experienced users to find the causes of human-system problems, in five stages: preparation, obtaining existing materials such as accident or investigation reports, gathering facts by asking people to describe an event, what led up to it and what resulted, analysis by grouping incidents by problem type, and interpretation fed back into design.

A research-design checklist drawn from the ECIT credibility checks

Butterfield et al. (2009) describe nine credibility checks that extend Flanagan’s original approach, together with two other enhancements: contextual questions at the start of the interview, and “wish list” items (people, supports or information that were absent but would have helped) in addition to helping and hindering incidents. The checks they describe in detail can be used as a design checklist:

• Audio-record interviews and work from recordings or transcripts so the participants’ own words are captured. See transcription conventions in qualitative research.
• Interview fidelity: have a CIT-experienced person listen to every third or fourth recorded interview to check that the method and guide are followed and that no leading questions are asked.
• Independent extraction: give about 25% of transcripts to a second person who identifies incidents without seeing the first coder’s work, then compute percentage agreement.
• Exhaustiveness: keep a log of new categories created per interview. In the worked example, no new categories appeared after the fifth of twenty interviews.
• Participation rate: count how many participants contributed to each category. The cited minimum for a viable category is 25% of participants (Borgen and Amundson, 1984, as reported by Butterfield et al.).
• Independent judge: an independent person places a sample (25%) of incidents into the categories; a match rate of 80% or better is the cited guideline (Andersson and Nilsson, 1964).
• Participant cross-checking: a second contact in which participants confirm their incidents and the categories they were placed in.

The article describes nine checks in total; the seven above are those it details most fully.

Where the term is used: a source-by-source comparison

Field	How “critical incident” is used	Source
Job analysis (original)	Reports of satisfactory and unsatisfactory task execution used to determine critical requirements for job success	Flanagan (1954), quoted by the University of Warwick
Counselling and organizational psychology research	Interview-based, exploratory study of what helps or hinders an experience; categories reported with participation rates and credibility checks	Butterfield et al. (2009)
User research	Gathering facts about events from domain experts or less experienced users to find causes of problems between people and systems	Usability Body of Knowledge (UXPA)
Health care and patient safety	A methodology “aimed to record specific behaviors, rather than opinions or estimates, in evaluating a given activity”; applied in health care, including studies of anaesthesia mishaps	AHRQ PSNet
Teaching and reflective practice	An incident becomes critical through the significance a practitioner attaches to it; written up in a critical incident file for reflection	Tripp (1993), quoted by the University of Warwick

Where it matters: critical incidents in teaching and performance management

What is an example of a critical incident in teaching?

In Tripp’s (1993) reflective-practice usage, a critical incident is not a dramatic event but an ordinary classroom or professional event that, once analysed, shows something about underlying values, motives or structures. Tripp writes that “critical incidents are produced by the way we look at a situation: a critical incident is an interpretation of the significance of an event.” A teacher’s critical incident file is written for the writer first, then a critical friend, then a wider readership of colleagues and researchers. This is a reflective-practice use and differs from Flanagan’s data-collection method, where incidents from many participants are categorised.

What is a critical incident method in a performance management system?

Performance appraisal is one of the fields in which CIT has been applied, according to the 2005 Butterfield et al. review of the method’s fifty-year history (quoted by the University of Warwick). The idea carries over from Flanagan’s job analysis: specific, first-hand reports of effective and ineffective execution of a task, rather than general impressions, form the evidence.

Common misunderstandings

• It is not root cause analysis. Safety reporting and investigation often start from an incident, but CIT is a research method that collects many incidents and categorises them. For one regulatory use of the word “incident”, see critical safety incident reporting under California SB 53, which is a legal reporting duty rather than a research method.
• It is not an opinion survey. AHRQ PSNet describes the aim as recording specific behaviors rather than opinions or estimates.
• “Incident” does not mean a sudden event. Cope and Watts (2000), quoted by Warwick, found the term can understate prolonged critical episodes and that boundaries of incidents are hard to define.
• It does not fit every research question. It explores what helps or hinders. For processes use grounded theory; for individual experience, phenomenology (Butterfield et al., 2009, citing Creswell).
• It is not a single fixed procedure. Flanagan’s five steps are stable, but the credibility checks and wish-list questions were added later as the ECIT.

Related methods in the same cluster include participant observation, maximum variation sampling and the Colaizzi method of phenomenological analysis.

FAQ

What is the critical incident technique in simple terms?

A method for gathering specific events people experienced or observed, sorting them into categories, and using the categories to understand what helps or hinders an activity.

Who created the critical incident technique?

John C. Flanagan, in “The critical incident technique”, Psychological Bulletin 51(4), 327-358 (1954), DOI 10.1037/h0061470, from work in the US Army Air Forces Aviation Psychology Program.

How many participants does a CIT study need?

Flanagan’s criterion is exhaustiveness: continue until no new incidents or categories appear. Butterfield et al. (2009) note that in their example this occurred after the fifth interview, and that interviewing may continue for other reasons.

Is CIT qualitative or quantitative?

It meets standard descriptions of a qualitative method (words as data, inductive analysis, focus on participants’ perspectives), though categories can be reported with participation rates (Butterfield et al., 2005, 2009).

Sources

• Flanagan, J. C. (1954). The critical incident technique. Psychological Bulletin, 51(4), 327-358. PubMed record. Read 2026-11-07.
• Butterfield, L. D., Borgen, W. A., Maglio, A.-S. T., and Amundson, N. E. (2009). Using the Enhanced Critical Incident Technique in counselling psychology research. Canadian Journal of Counselling, 43(4), 265-282. Full text. Read 2026-11-07.
• University of Warwick, GlobalPAD Core Concepts: Critical Incidents (2013), quoting Flanagan (1954), Tripp (1993), Cope and Watts (2000) and Butterfield et al. (2005). PDF. Read 2026-11-07.
• Usability Body of Knowledge (UXPA), Critical Incident Technique. Read 2026-11-07.
• AHRQ Patient Safety Network, Critical incident technique. Read 2026-11-07.
• Tripp, D. (1993). Critical Incidents in Teaching: Developing Professional Judgement. London: Routledge.