How Interviewers Rationalize Decisions They've Already Made

A well-documented finding in hiring research is that interviewers often form an overall impression of a candidate very early in the interview, sometimes within the first four minutes, and then spend the rest of the conversation gathering evidence that confirms that impression.

This is not a character flaw. It is a feature of how human cognition works under uncertainty. The problem is that it undermines the stated purpose of an interview.

The thin slice phenomenon

Research on “thin slices,” brief exposures to a person’s behavior, shows that people form reasonably accurate impressions of some traits (like extraversion) from very limited information. Hiring managers often interpret this as evidence that their early impressions are reliable signals.

They are not reliable for predicting job performance. Traits that are readable from a thin slice (how confident someone presents, how articulate they are in opening small talk) correlate weakly with actual job performance. The competencies that matter, how someone handles a project that goes wrong or how they navigate a disagreement with a peer, require more than first impressions.

The thin slice literature shows that initial impressions are accurate at measuring some things. It does not show that those things are what matter for the job.

Confirmatory bias in interviews

Once an initial impression forms, interviewers shift into confirmation mode. They ask questions that are more likely to elicit confirming evidence. They weight positive answers more heavily when they like the candidate. They probe harder when they are skeptical, producing more evidence of weakness from candidates they already doubt.

Research by Dougherty et al. found that when interviewers had a positive initial impression, they evaluated candidates more favorably, sought out confirming information, and remembered evidence consistent with their impression better than disconfirming evidence. The pattern reversed for negative initial impressions.

The interviewer doing this is not aware of it. The bias operates below the level of deliberate decision-making.

What the note-taking problem reveals

One indicator of confirmatory bias is what interviewers choose to write down. In unstructured settings, interviewers take notes selectively, recording what seems important rather than what the candidate actually said. When they like a candidate, they tend to record their answers in positive language. When they do not, they record the same content in more negative framing.

This is why note-taking guidance matters as much as question design. The instruction should be explicit: record what the candidate said, not what you thought about it. Behavioral summary notes, not evaluative impressions.

Structural countermeasures

The solution is structural, not motivational. Telling interviewers to be more objective does not work. Research on bias-awareness training shows it has limited impact on actual behavior.

What works:

Standardized questions asked in order: Interviewers who follow a script cannot selectively ask easier questions to candidates they like.

Behavioral anchors for each rating level: When the scoring criterion is “describe behavior that matches anchor 3,” interviewers have less room to apply a general impression.

No discussion before independent scoring: Interviewers who score independently before the debrief are anchoring to their own evidence, not conforming to a shared impression that forms in conversation.

Interview-blind reviews where practical: Removing irrelevant information from the review (where it does not compromise fairness) reduces the surface area for initial impression formation.

None of these eliminate the bias entirely. They reduce its influence on the final decision, which is the achievable goal.