Critical Incident Method

What the critical incident method is

Critical incident method (طريقة الحوادث الحرجة, also called الوقائع الحرجة) is a way of appraising an employee’s performance in which the appraiser records, over the whole period, specific events in the employee’s work behaviour, whether effective or ineffective, and then returns to them at the appraisal to base a judgement on them. The record holds real examples with their dates and context. It holds no general impressions and no scores on a scale.

The name goes back to the critical incident technique that John Flanagan set out in an article titled “The Critical Incident Technique”, published in Psychological Bulletin in 1954 (volume 51, issue 4, pages 327 to 358). The summary of that article in the PSNet database of the US Agency for Healthcare Research and Quality describes the technique as a method for recording specific behaviours, not opinions or estimates. Its use in performance appraisal is an applied form of that idea, and its shape varies from one establishment to another.

What “critical” means in the name

The word does not mean dangerous or urgent. A critical incident is an event with a visible effect on the result of the work, positive or negative, so that the behaviour in it serves as an example of what the employee does when the situation matters. The event may look ordinary: a trainer’s reply to an angry customer, a report delivered on time despite a system failure, or a review step that was skipped and led to rework.

An event has qualities that make recording it useful:

• Specific. It describes what happened on a particular day and not what recurs from day to day.
• Behavioural. It describes what the employee did or said and not the employee’s temperament.
• Observed. The appraiser saw it, or a reliable party documented it. It is not an account without a source.
• Consequential. It has a result that can be described in the work of the team, the customer or the establishment.

What is recorded for each event

One or two lines for each event are enough, covering these elements:

• The date of the event and where it took place.
• The situation: what called for action.
• The employee’s behaviour: what was actually done, in neutral terms.
• The effect: what followed from the action.
• The general label, effective or ineffective, meaning whether the behaviour served the purpose or weakened it.

The difference between an event and an impression shows in the wording. “The employee is disorganised” is a judgement on the person. “Delivered the March report two days after the deadline, so the financial review was late” is an event. The second can serve as the basis for a conversation in which the employee knows what is meant, while the first opens an argument about temperament.

How an event is worded well

Take two hypothetical examples for a customer service employee. The first is an effective event: “In the third week of April, she received a complaint from a customer about a late order, contacted the warehouse and gave the customer the updated delivery date within an hour, and the customer kept the order.” The second is an ineffective event: “In the first week of June, she promised a customer a date she had not checked, the order arrived late and the customer lodged a second complaint.”

The examples have what makes the wording sound: a specific time, an action that can be seen and an effect that can be described. Neither passes judgement on the employee by calling her careless or excellent. The judgement comes after the events are taken together, not in the line that records an event. This is our own view and not a quotation: what is written in descriptive language is easier for the employee to verify or to contest, and what is written in the language of judgement may open an argument that never reaches a conclusion.

How the method differs from neighbouring appraisal tools

The method sits beside other tools in the area of performance, and it differs from them in what the appraisal is made of: a recorded event, as against a score or an impression.

• The wider frame. The performance management cycle is the yearly sequence that includes goals, monitoring, appraisal and decisions. The critical incident method is a way of gathering evidence that is used within the monitoring and appraisal stage of that cycle. It is not the cycle itself.
• The immediate remark. Continuous feedback means giving the employee remarks close to the event. A record of critical incidents preserves what happened. A remark to the employee may be built on it, or the record may stay with the manager until the appraisal. The first is speech that reaches the employee, and the second is evidence that is kept.
• The tool with a scale. A graphic rating scale asks the appraiser to choose a position on a scale for each item. The critical incident method has no scale. It consists of narrative, and the judgement is drawn from it after the events are examined. Incidents of this kind also serve as the material from which the anchors of a Behaviourally Anchored Rating Scale (BARS) are built.
• Several sources. A 360 degree appraisal gathers views from several parties. The critical incident method gathers events that the appraiser observes in the first place, although events narrated by others can be added.
• The role, not the event. A critical role is a description of a position whose absence stalls the work. It is a concept about positions and not about an employee’s behaviour. The resemblance is in the word only.

The problem the method addresses

The method addresses a possible flaw in the yearly appraisal: that the judgement is built on what the appraiser remembers on the day of the appraisal. This is what the recency effect in appraisal describes, when the score is dominated by what happened in the final weeks. A written record of events kept over the period may limit that effect, because the early evidence stays written in front of the appraiser at the moment of deciding.

This is a reasonable hypothesis and not a guarantee. The same appraiser may record the events of the final weeks more than the rest, and the effect returns in another form. For that reason the spread of events across the period is reviewed before they are relied on.

A hypothetical example with figures

Suppose a hypothetical manager recorded 18 events about one employee during a year, spread across the quarters as follows: 4 in the first quarter, 5 in the second, 4 in the third and 5 in the fourth. The total is 4 + 5 + 4 + 5 = 18. The share of events in the last quarter is 5 ÷ 18 × 100 ≈ 27.8%, which is close to 25%, the share a quarter would hold if the events were spread equally. No clear leaning toward the end of the year appears.

The manager classified the events as 11 effective and 7 ineffective, so the share of effective events is 11 ÷ 18 × 100 ≈ 61.1%. This share is not a score, because one event can outweigh several others in its effect. If one of the ineffective events caused the loss of a customer, the employee cannot be judged by a percentage alone. The figures are hypothetical and are set out to illustrate; they do not represent an establishment or a study.

What matters is the content: which events recurred, what changed between the first quarter and the last, and which events had the larger effect. If it turned out that all seven ineffective events came in one task, what may be called for is training on that task and not a judgement on the employee in general.

How the method is run in an establishment

The method needs discipline in recording and not a complex form. The steps in running it include:

• Deciding who records. The direct manager, and possibly whoever supervises a particular project.
• Agreeing the format. A short form with the elements of an event, and a rule that an event is written within a few days of occurring and not months later.
• Recording both kinds. Effective and ineffective events are recorded together. Otherwise the record may turn into a file of faults.
• Informing the employee. We suggest that the employee know that the record is kept and what it contains, so that it does not become a secret file that appears at the end of the year.
• Reviewing periodically. The record is reviewed in the middle of the period, so that the format can be corrected and gaps in recording can be found.
• Using it at the appraisal. The events are taken together and a description of performance is drawn from them. It is then compared with the self assessment that the employee wrote, and the differences between the two accounts are discussed.

Events may also be brought to performance calibration as shared examples, so that appraisers agree on what effective performance means in their establishment through real events and not through general descriptions.

Uses in job analysis and interview design

Events may be recorded for a purpose other than appraisal: to describe what outstanding performers do in a job and what struggling performers do, so as to feed job analysis with behavioural examples. Behavioral interview questions are related to it as well, because they ask the candidate for an event from earlier work. They are close in idea to a record of events, although what is recorded in the first case is the candidate’s account and not the manager’s observation.

The limits of the method

• The burden of recording. It may cost the manager time throughout the year, and the manager may stop if time is short, leaving the record thin in some months and heavy in others.
• Selection of what is recorded. The appraiser chooses what to record and may lean toward what caught the eye or matched a prior impression. The record is therefore a selection and not a complete picture.
• Weak comparison. Because events are narrative and not scores, it may be hard to compare one employee with another or one team with another, so the method may not suit ranking employees on its own.
• Confusing an event with an opinion. If an event is written in a judgemental form, the impression returns to the record in new wording.
• No conversion into a decision. The method gathers evidence and does not say how it is turned into a score or a decision. That is a matter for the policy of the establishment.

A record of critical incidents remains a tool for gathering evidence and not a means of judging by itself. It shows the appraiser what actually happened in the employee’s work, and what is decided afterwards rests on taking the events together and on what the employee says about them.

This is an explanation of the concept, not legal advice.