Pre-Employment Assessment or Behavioral Interview: What Predicts?

Behavioral interviews and behavioral assessments sound like the same thing. They are not. One is a conversation about what a candidate has done; the other is a standardized instrument that produces comparable scores. The research verdict is less about which one wins and more about one property they can both have, or both lack: structure.

That is the honest answer, and it is a useful one. If you hire experts for missions, it changes what you should ask, how you should score it, and where a test actually helps.

What is a behavioral interview?

A behavioral interview asks candidates about real past situations: "Tell me about a time when a client changed the scope halfway through a mission. What did you do?" The idea is simple: the best clue to how someone will act is how they have already acted in similar situations.

In practice, two very different things go by the same name.

• The unstructured version. The interviewer improvises the questions, follows the conversation where it goes, and forms an overall impression at the end.
• The structured version. The questions come from an analysis of the mission, every candidate gets the same ones, and answers are scored against a grid defined in advance, ideally by more than one interviewer.

Both feel like "a good conversation". Only one of them produces something you can compare across candidates.

What is a behavioral assessment?

A behavioral assessment is a standardized instrument: the same questions or scenarios for everyone, answered in the same conditions, scored in the same way. What it produces that an interview cannot is a comparable profile. Candidate A and candidate B are described on the same dimensions, without depending on who interviewed them or how the conversation went.

The family is wide. It includes general personality questionnaires, situational judgment tests ("what would you do in this situation?"), and instruments that describe behavior at work specifically. That last distinction matters more than it seems, as the figures below show.

At Migbirds, our own assessment is the IDC. It describes how someone tends to act in a professional context, and we use it as one input to a fit prediction, never as a verdict on a person.

Which method best predicts job performance?

The most complete answer today comes from a 2022 re-analysis of decades of hiring research by Sackett, Zhang, Berry and Lievens, summarized for practitioners in an open-access 2023 article. They corrected a statistical habit that had inflated earlier estimates, and the ranking changed.

Here is what they found, as average correlations with overall job performance (predictive validity, on a scale from 0, no link, to 1, perfect link):

• Structured interviews: 0.42, the highest of all the methods reviewed.
• Personality measures framed around work, for example conscientiousness measured "at work" rather than in general: 0.25.
• Situational judgment tests: 0.26.
• General cognitive ability tests: 0.31 in older data, and 0.23 in a separate analysis of 21st-century studies cited by the same authors.
• Conscientiousness measured in general: 0.21.
• Unstructured interviews: 0.19.
• The other general personality traits (extraversion, agreeableness, emotional stability, openness): 0.11 or lower.

Two readings stand out.

First, the gap between the two interviews (0.42 against 0.19) is larger than the gap between an interview and a test. The format, conversation or questionnaire, matters less than whether it is structured and tied to the work.

Second, the same is true inside the assessment family. A personality trait measured in general predicts little. The same trait measured in a work context predicts more. The authors' own summary is that the methods at the top are those specific to the work, while general measures of psychological constructs do less well.

The limits, stated plainly. These are averages across many studies, and they vary a lot from one setting to another: for structured interviews, the authors give a range of roughly 0.18 to 0.66 around the 0.42 average. They concern overall job performance, not every outcome you may care about. And they describe each method used on its own. Read them as a map, not a promise.

Why are unstructured interviews unreliable?

Because without a grid, an interview measures the conversation as much as the candidate.

Think of two interviewers meeting the same data analyst for a six-month mission. One spends most of the hour on a shared passion for cycling and leaves convinced. The other asks about a failed project, gets a hesitant answer, and leaves doubtful. Same person, two opposite conclusions. Neither interviewer is wrong to trust their impression; the problem is that the impression depends on them.

An unstructured interview rewards ease in conversation, shared references and a good first five minutes. Those can be pleasant qualities. They are rarely what makes a mission succeed, which depends more on how someone organizes their work, handles friction with a client and adapts when the scope moves.

Structure fixes this in a boring, effective way: same questions, same scale, notes taken during the interview, scores given before any discussion between interviewers. It takes preparation, and the 2023 article notes that structured interviews are harder to run at high volume. For a handful of expert missions, that cost is small.

Can you combine both approaches?

Yes, and it is the practical recommendation.

The same authors modeled what happens when methods are combined. Putting together six common ones (cognitive ability, conscientiousness, biodata, structured interviews, integrity tests and situational judgment tests), they estimate a combined validity of about 0.61, against 0.42 for the best single method. That is a statistical estimate built from the meta-analyses, not a measured hiring process, but the direction is clear. The point is not to stack tests. It is that each method captures something the others miss.

A simple sequence for an expert mission:

1. Assessment first, to generate hypotheses. A standardized profile tells you where to look: this person tends to work alone, prefers a stable scope, reacts quickly under pressure.
2. Structured interview next, to test them. You turn each hypothesis into a behavioral question for this specific mission: "Tell me about a mission where the client kept changing priorities. How did you organize yourself?"
3. Scoring before discussion. Each interviewer scores on the same grid, then you compare.

The assessment saves the interview from being generic. The interview saves the assessment from being read as a label. And the decision stays with people who have looked at both.

What are the limits of personality tests in hiring?

Naming them is what makes the rest credible.

• General traits predict little on their own. In the 2022 figures, most general personality traits sit at 0.11 or below. A test that only describes "who you are" in general will not tell you much about a specific mission.
• Candidates know they are being evaluated. In a hiring context, people answer with a goal. Sackett and colleagues list differences in "incentives to fake" between applicants and current employees as one reason results vary between studies.
• A score is not a diagnosis. A behavioral profile describes tendencies in a context. It does not say what someone is worth, and it should never decide alone.
• Quality varies. Not every instrument is built or validated to the same standard. Ask any provider what their tool measures, how it was validated and on whom.

None of this disqualifies assessments. It tells you how to use them: as structured information, combined with a structured interview, inside a decision that people own.

The bottom line

Behavioral interview or behavioral assessment is the wrong question. The useful one is: is what we are doing structured, and is it tied to the work? An improvised interview and a generic personality test share the same weakness. A structured interview and a work-focused assessment share the same strength. Used together, they predict better than either alone.

If you want to go further on what science can and cannot predict, read our article on whether job performance can really be predicted, and our practical guide on how to assess soft skills before you hire.

References

• Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting meta-analytic estimates of validity in personnel selection: Addressing systematic overcorrection for restriction of range. Journal of Applied Psychology.
• Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2023). Revisiting the design of selection systems in light of new findings regarding the validity of widely used predictors. Industrial and Organizational Psychology, 16(3), 283–300. doi.org/10.1017/iop.2023.24
• French Labour Code, Article L1221-6 and Article L1221-8.