Selection method chooser: the right mix for the role.
Most roles need more than an interview. Which other methods, and in what order, depends on the role, the number of applicants and the cost of getting it wrong.
Answer seven questions about the role; get a mix of selection methods in three stages, with the reason for each. Nothing is stored.
The evidence behind the choices
In 2022 Paul Sackett and colleagues re-analysed decades of research on how well selection methods predict job performance. Among the common methods, structured interviews had the highest average operational validity (.42), ahead of job knowledge tests (.40), empirically keyed biographical data (.38), work samples (.33) and general cognitive ability tests (.31). These are averages across many studies and many jobs. A well-built method matched to the role can do better than its average, and a poorly built one much worse. No single method is enough, which is why the chooser always recommends a mix.
Some methods in the chooser have no figure beside them on purpose. Personality questionnaires are recommended only as a way to plan interview questions. Situational judgement tests are recommended for high-volume, entry-level shortlisting because they suit people without a track record. References and checks are there to verify, not to predict.
How the chooser decides
| Stage | Method | Recommended when |
|---|---|---|
| Shortlisting | Read applications against agreed criteria | Fewer than 10 applicants per vacancy |
| Shortlisting | Scored application sift | 10 to 50 applicants |
| Shortlisting | Situational judgement test | More than 50 applicants and an entry-level role |
| Shortlisting | Short reasoning test, alongside the SJT | More than 200 applicants and an entry-level role |
| Shortlisting | Online job knowledge test | More than 50 applicants, not entry level, and specific job knowledge needed |
| Shortlisting | Online reasoning test | More than 50 applicants, not entry level, and no specific job knowledge needed |
| Assessment | Structured interview | Always |
| Assessment | Work sample | The core of the job can be shown in a short practical task |
| Assessment | Job knowledge test | Specific job knowledge needed and not already tested at shortlisting |
| Assessment | Reasoning test | Professional or leadership role, wrong hire high or very high cost, and not already tested at shortlisting |
| Assessment | Personality questionnaire, to plan interview questions | Wrong hire high or very high cost, or a leadership role |
| Assessment | Multiple exercises observed by trained assessors | Leadership role, wrong hire very high cost, and more than two weeks to decide |
| Before you confirm | References and pre-employment checks | Always |
| Before you confirm | Sector rules first | Regulated role, or work with children or adults at risk |
Two notes are added where they apply: a same-day assessment plan when you need to decide within two weeks, and a reminder to verify unsupervised online results in a supervised step whenever the mix includes an online test. For the reasoning behind that, see hiring in the AI era.
Going further
For each method: running a structured interview; situational judgement tests; before you use a personality questionnaire; taking up references.
Method mix and rules by great{with}talent from our assessment practice. Validity figures: Sackett, P. R., Zhang, C., Berry, C. M. & Lievens, F. (2022), “Revisiting meta-analytic estimates of validity in personnel selection”, Journal of Applied Psychology, 107(11), 2040–2068. Ability, personality, situational judgement and 360 assessment are part of FindingPotential.
The clearest thirty minutes in your week.
Book a walk-through with a real person, or start with a pilot or a 30-minute conversation.