Behavioural interview questions: what to ask, and how to score the answers
Behavioural questions rest on a simple premise: past behaviour in a comparable situation predicts future behaviour better than a candidate’s theory about themselves. The premise holds. The execution is where most interviews lose it.
- Ask for a specific past instance, not a general policy — "tell me about a time" rather than "how do you handle".
- STAR is a probing structure for the interviewer, not a script to impose on the candidate.
- The Action and Result are where evidence lives; most candidates over-narrate the Situation.
- Situational ("what would you do") questions measure judgement; behavioural questions measure track record. Use both deliberately.
- A rehearsed answer is detected by probing for specifics the story would not have prepared — the second decision, the thing that went wrong, the person who disagreed.
Behavioural, situational and the difference that matters
| Type | Form | Measures | Best for |
|---|---|---|---|
| Behavioural | Tell me about a time you… | Demonstrated track record in a comparable situation | Experienced candidates with relevant history to draw on |
| Situational | What would you do if… | Judgement and reasoning about a job-relevant scenario | Graduates, career changers, and roles where the situation is new to everyone |
| Knowledge | How does X work? | Domain knowledge | Roles with a genuine technical floor — used sparingly, and never as a proxy for capability |
| Self-assessment | Are you a good communicator? | Confidence, essentially | Nothing. Avoid. |
The last row is not a joke. Self-assessment questions survive in interview scripts because they feel like they are asking about a competency. They correlate with self-confidence and with cultural norms about self-promotion, both of which vary systematically by gender and background, and they produce no ratable evidence at all.
STAR is a probing tool, not a format to demand
STAR — Situation, Task, Action, Result — is widely taught to candidates as an answer format and to interviewers as a scoring frame. It is more useful as neither: treat it as a checklist of what you still need before you can rate the answer.
- Situation. Enough context to judge difficulty. One or two sentences is sufficient; more is usually avoidance.
- Task. What specifically was theirs to do. This is where "we" answers get separated from "I" answers.
- Action. What they personally did, in sequence, including the decisions they made and rejected. This is the evidentiary core.
- Result. What happened, ideally with a number, and what they would do differently. A candidate who cannot say how it turned out was probably not close enough to the outcome to own it.
When an answer is missing an element, probe for that element specifically rather than restating the question. "What did you personally do first?" recovers an Action; "How did it turn out?" recovers a Result. Both are neutral probes that any candidate can answer, which keeps probing equal across your pipeline.
A question bank by competency
Adapt the wording to the role and the level. The competency names below follow the MERIT-8 framework, but the questions work against any comparable rubric.
Interpersonal and service orientation
- Tell me about a time you dealt with someone who was upset or frustrated with your organisation. What did you do first?
- Describe a time you had to say no to someone who really wanted a yes. How did you leave the relationship?
- Tell me about a colleague you found difficult to work with. What did you change about how you worked with them?
Work discipline and reliability
- Tell me about a commitment you made that became much harder than expected. What did you do?
- Describe a time you noticed something was going to be late before anyone else did. What happened next?
- Walk me through how you kept track of your work in your last role, on a normal week.
Judgement and integrity
- Tell me about a time you were asked to do something you thought was wrong, or unwise. What did you do?
- Describe a decision you made with incomplete information. What did you do to reduce the risk?
- Tell me about a mistake that had a real consequence. Who did you tell, and when?
Adaptability
- Tell me about a time a plan you were relying on fell apart at short notice.
- Describe something significant about how you work that you have changed in the last two years, and why.
- Tell me about the last time you had to learn something quickly to do your job.
Problem solving
- Walk me through the hardest problem you solved in the last year. Start with how you worked out what the problem actually was.
- Tell me about a time your first solution did not work.
- Describe a time you decided a problem was not worth solving.
Telling a real example from a rehearsed one
Preparation is legitimate and you should not penalise it. What you are separating is a prepared account of something that happened from a constructed account of something that did not — and the distinguishing feature is depth beyond the narrative.
- 01Probe the second decision. Prepared stories have one decision. Ask what they decided next, after the thing they just described.
- 02Ask who disagreed. Real situations contain other people with other views. A story with no friction usually has no source.
- 03Ask for the timeline. Real events have durations, sequences and dead time. Constructed ones compress.
- 04Ask what they would do differently. A rehearsed answer often has a polished version of this; ask them to be specific about the cost of the alternative.
- 05Ask about the boring part. What did the handover look like, who signed it off, what did you write down.