A candidate gives a confident, well-structured answer about leading a migration. You finish the interview and can't say what they personally did. That happens a lot. Interviews collect claims, and hiring decisions need evidence.
This post covers behavioral evidence in interviews: the definitions, the red flags in polished answers, how to evaluate what you hear, and how to probe when it's thin.
Behavioral Claims vs. Behavioral Evidence: Key Definitions
Behavioral claim. A general statement about oneself. "I'm a strong leader." "I handle conflict well."
Behavioral evidence. A specific past action, the decision behind it, and what came of it. Enough detail that you could check it.
Behavioral interview. Questions about past behavior ("tell me about a time...").
STAR. Situation, Task, Action, Result. A way to structure an answer, nothing more.
For scoring these against defined criteria, see competency-based interviewing.
Why Polished Interview Answers Can Be Misleading
Polished answers sound complete, which is why they get waved through. Four patterns to watch for:
Rehearsed STAR. Every box is ticked, but the Action part describes what the team did, not the candidate.
Fluency. Good communicators can make this experience sound solid for ninety seconds.
"We." "We cut costs by 15%." Who ran the analysis? Who made the call?
"I would." A hypothetical is a claim about the future. It isn't evidence from the past.
Most of these candidates aren't lying. They're describing, and description sits at the self-reported end of verified capability vs. self-reported skills.
How to Identify Specific Actions, Decisions, and Outcomes
Real evidence has five parts:
Context. The actual situation, with real constraints.
Their own action. What they did, not the team.
The reasoning. Why that option over the others.
A result. Something that happened, ideally with a number.
What they'd change. Hindsight shows whether they learned anything.
Standard behavioural interview questions tend to produce claims. "Tell me about a time you handled conflict" gets you "I always try to listen first." A sharper version gets evidence: "Tell me about the last time you disagreed with a peer over a deadline. What did you say, and what happened next?"
Claim | Evidence |
|---|---|
"I'm great at stakeholder management." | "Our VP wanted the launch moved up. I showed her the three tasks that would slip, and she kept the date." |
"We improved onboarding." | "I cut the manager sign-off at day three, because it never caught anything. Time to first ticket went from nine days to five." |
"I'd always escalate early." | "In March, a vendor missed a handoff. I escalated that afternoon instead of waiting for Friday's sync." |
Evaluating Behavioural Evidence Against Role-Specific Competencies
Start with four to six competencies that matter for this role, not the generic list every company uses. Then rate what you heard against each one: strong, partial, or missing. "Good answer, bad answer" hides too much. A candidate can nail ownership and say nothing at all about prioritisation, and an average score covers up both. Weigh the ratings by role criticality, since missing evidence on a must-have counts for more than partial evidence on a nice-to-have. That's the core of competency-based interviewing, and it's what makes candidate competency assessment comparable across people. Behavioural assessment in hiring works the same way whether a human or an AI runs the conversation. For one way of putting it into practice, see deep candidate assessment.
How Follow-Up Questions Uncover Missing Evidence
Three probes turn a claim into evidence: "What did you personally do?" "What was the alternative, and why not that one?" "What was the result?" Ask them after any answer that stayed general. Adaptive follow-ups do the same thing at scale. They go deeper when an answer is vague, move on when it isn't, and hold every candidate to the same standard. The full breakdown is in adaptive follow-up questions.
How NaviHyr Maps Behavioural Evidence to Competencies
Behavioural evidence mapping means taking what a candidate describes (past actions, decisions, outcomes) and sorting it by competency. A story about a missed vendor handoff might count toward ownership and prioritisation, but not stakeholder communication. The output shows each competency as supported, partly supported, or lacking evidence, so a recruiter can see where the interview found proof and where it ran dry. The format is covered in hiring intelligence reports.
Conclusion
Claims tell you what candidates believe about themselves. Evidence shows what they've done. Evidence-based hiring starts with asking for the second. See how NaviHyr surfaces behavioural evidence, and what's still missing, before the offer.

