Selection interviews: what structure actually changes
Almost every organisation hires through interviews, and almost every one considers its own judgement reliable. Personnel psychology takes a more cautious position, and sharply distinguishes two practices that everyday language conflates: the unstructured and the structured interview.
What the reviews establish
The review by Julia Levashina and colleagues (2014, Personnel Psychology) synthesises a stable finding: structured interviews — identical job-anchored questions, rated against a scale defined in advance — show higher predictive validity than unstructured ones. Structure does not add intuition; it limits the variance introduced by the assessor.
An important revision of the magnitudes
Practitioners long cited the coefficients of Frank Schmidt and John Hunter (1998, Psychological Bulletin). Paul Sackett, Charlene Zhang, Christopher Berry and Filip Lievens (2022, Journal of Applied Psychology) re-examined those estimates and showed that several statistical corrections had inflated them. The revised validities remain useful, but more modest. This matters: it invalidates promises of near-mechanical prediction sometimes made to executive teams.
Why structure remains a minority practice
Scott Highhouse (2008, Industrial and Organizational Psychology) described practitioners' resistance to structured procedures: the belief that personal experience allows one to 'read' a candidate, and the sense that a rating grid dehumanises the exchange. The argument is understandable but unsupported: confidence in one's own judgement is precisely what correlates least with decision quality. (our executive and employee training programmes)
What this does not license us to conclude
Structuring an interview guarantees no successful hire. Reported validities are average associations across large samples, not individual predictions; they vary by role, context and the quality of the job analysis. A poorly designed structured process is no better than an unstructured interview run by an experienced assessor.
One practical implication, testable internally
A simple test before any redesign: take the last three executive hires and check whether two independent assessors had the same questions and the same rating grid. If not, the disagreement observed says nothing about the candidates. That is the starting point of our assessments in Geneva, ahead of any recommendation. To go further: explore the Executive Leadership training in Geneva, or browse our executive and employee training programmes in Switzerland.
In pictures: Executive Leadership in Geneva
- executive leadership
- Geneva
- research
- training geneva