Structured vs unstructured interviews: what the research says

MatchCard · 19 July 2026 · 6 min read

Quick answerStructured interviews, where every candidate is asked the same job-specific questions and scored against a fixed rubric, are nearly twice as valid as unstructured 'gut feeling' interviews at predicting job performance, according to Schmidt and Hunter's landmark meta-analysis of hiring research.

"I'll know it when I see it" is one of the most confidently stated, least defensible claims in hiring. Industrial-organisational psychology has been measuring exactly how well unstructured, conversational interviews predict job performance for decades, and the answer is settled science, not opinion. It's just rarely applied.

The headline number: nearly double the predictive power

The foundational work here is Schmidt and Hunter's 1998 meta-analysis, still the most-cited study in personnel selection. It found that structured interviews have a predictive validity of .51 for job performance, against .38 for unstructured interviews, meaning structured interviews are roughly twice as effective at actually predicting who will perform well on the job.

The finding has held up under replication. A 2023 update by Sackett and colleagues placed structured interview validity at .42, the highest-performing predictor among all commonly used selection methods, while unstructured interviews scored a considerably lower .19 in the same analysis. Whichever version of the numbers you use, the ranking never changes: structure wins, and it isn't close.

Why unstructured interviews perform so badly

"Gut feeling" hiring isn't just imprecise. It's measurably vulnerable to specific, well-documented biases: interviewers over-weight the first few minutes (first-impression bias), unconsciously favour candidates who resemble themselves (affinity bias), and let one strong or weak answer colour every answer that follows (the halo effect). None of this requires bad intent. It's how unstructured conversation works, and it's exactly what standardisation is designed to interrupt.

Structured interviews close the gap through three specific mechanisms: the same core questions for every candidate, in the same order, scored against predetermined behavioural anchors rather than a vague overall impression. That structure is what makes candidates comparable to each other at all: an unstructured conversation with candidate A and a different unstructured conversation with candidate B were never actually measuring the same thing.

Where "culture fit" fits into this

"Culture fit, decided on a gut feeling" is one of the most common unstructured judgements in hiring, and one of the least examined. It's rarely a defined, scored criterion. It's a feeling formed in the room, which means it inherits every bias listed above, and it's nearly impossible to challenge after the fact because it was never written down as a specific claim in the first place.

What this means for how you screen, not just how you interview

Worth noting: not every "structured" tool is equally structured in the way that matters. A fixed personality-trait questionnaire is standardised, but it's standardised around an abstract trait model disconnected from the actual role. See how MatchCard compares to personality tests and other traditional assessments for where that distinction bites.

The research above is specifically about the interview stage, but the same principle governs everything before it. A screening step built the same way as a good structured interview (fixed questions engineered to reveal something specific, scored against evidence, not a general impression) should carry the same advantage over an unstructured one. MatchCard's Match Cards are role-specific scenario questions, engineered per assessment to probe something particular about the actual job, and every judgement in the resulting report is required to quote the evidence it's based on, the same discipline that makes structured interviews outperform "I'll know it when I see it" by a factor of two.

Frequently asked