How to test candidates before hiring: a practical guide
To test candidates before hiring, decide what you need to see, give every applicant the same job-relevant task under the same conditions, agree how you will review the work before you invite anyone, and tell applicants what is recorded. No single method is best, so most teams combine a practical task with a structured interview.
What are you trying to find out?
Start with the hiring question, not the format. “Can this person handle a refund dispute?” points to a work sample or simulation. “Do they know our terms?” points to a knowledge test. “How do they explain their decisions?” points to a structured interview. Each answers a different question, and none replaces the others. See the Evidence Ladder for a vocabulary.
Which method fits which question?
| If you need to see | Common method | What it cannot show |
|---|---|---|
| Whether someone knows a rule or term | Knowledge or skills test | What they do when a rule is buried in a record |
| Which response someone says they would choose | Situational judgement test | What they actually do |
| How someone describes past or future behaviour | Structured interview | The behaviour itself |
| What someone does with a piece of work | Work sample or job simulation | Their reasons, unless the exercise asks for them |
| Independent planning over days | Take-home assignment | How the work was done, or who did it |
How do you run a practical test well?
- Choose a bounded piece of real work. Pick a responsibility that matters in the job and that can be approached with information you are prepared to provide. See how to choose real work for a hiring exercise.
- Give every applicant the same brief and conditions. Same materials, same time limit, same instructions.
- Decide how you will review it before you invite anyone. Agree what you will look at and how you will record notes, and keep what you saw apart from what you conclude. See how to discuss applicant work without a score.
- Try it on someone who does the job. A colleague can tell you whether the brief is clear and whether the task is realistic.
- Give the right amount of context. Supply what a new hire would be given, and do not supply the judgement you want to see. See how much context entry-level applicants should receive.
- Tell applicants what is recorded and who sees it. See what we tell applicants before they start.
- Keep it proportionate. Calavion has no validated figure for how long applicants should spend, and the sources we reviewed give no reliable one. Match the length to the work, and consider paying for anything long or reusable.
- Use the result in the conversation. A specific moment in the applicant’s work gives both sides something to discuss. See how to debrief a practical hiring task.
Should you combine methods?
Most teams do. In the 2022 re-estimate by Sackett and colleagues, structured interviews ranked highest for predicting job performance (.42), work-sample tests were at .33 and job knowledge tests at .40. The same paper reports a wide spread between studies for structured interviews, so a structured interview is not guaranteed to outperform other methods in a given setting. No single method dominates. Work-sample estimates rest mostly on studies of current employees, so they may not match how applicants perform. The details are in the evidence review.
What about fairness and legal risk?
Any selection procedure, including a work sample, can be examined for adverse impact. Keep criteria job-related, give every applicant the same case, keep records and take legal advice for your location. Calavion does not claim its exercises are fair, valid or free of adverse impact, and this is not legal advice.
Where does Calavion fit?
Calavion prepares an exercise with you from work your team has handled. Applicants work inside it and you read a Replay of what they did. Calavion does not currently produce automatic scores, applicant rankings or hiring decisions. Employers inspect the work and make the decision. It has run one exercise end to end, a customer service cancellation case, and would build any other with your company.
Sources
- 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, 107(11), 2040–2068. doi:10.1037/apl0000994
- Roth, P. L., Bobko, P., & McFarland, L. A. (2005). A meta-analysis of work sample test validity: Updating and integrating some classic literature. Personnel Psychology, 58(4), 1009–1037. doi:10.1111/j.1744-6570.2005.00714.x
- Huffcutt, A. I., & Arthur, W. (1994). Hunter and Hunter (1984) revisited: Interview validity for entry-level jobs. Journal of Applied Psychology, 79(2), 184–190. doi:10.1037/0021-9010.79.2.184