What is a work sample test? Definition, types and when to use one
A work sample test is a hiring exercise in which an applicant does a small, realistic piece of the job and the employer reviews what they produced. It differs from a quiz or an interview because it asks for the work itself, not an answer about the work. Employers use it for roles where the output can be read.
What this page is, and is not
- What it covers: a definition, the common kinds of work sample test, when one fits, what research says and where it falls short.
- What it is not: legal advice or a claim that any one format is best for every role.
- What Calavion does: prepares an exercise with a company from work it has already handled, and gives the employer a Replay of what each applicant did and wrote.
- What Calavion does not do: produce scores, rankings or hiring decisions. One customer service exercise runs today.
What is a work sample test?
A work sample test asks an applicant to perform a bounded task taken from the job. The US Office of Personnel Management (OPM) describes work samples as tasks or work activities that mirror the tasks employees perform on the job (OPM, work samples and simulations). The applicant produces something: a written reply, a reconciled sheet, a plan, a handover note. A reviewer reads it against what the job needs.
The idea is simple. A résumé lists what someone says they did, and an interview describes it. A work sample lets you see how they handle a piece of the work now. For the difference between the two kinds of evidence, see the Evidence Ladder for hiring.
What types of work sample test are there?
There is no single standard list, so these are the groups employers commonly use, named by what the applicant produces.
- Written task. The applicant writes a reply, a summary or a note from a set of records, such as a customer email or a handover.
- Analysis task. The applicant works with figures or documents, such as reconciling an invoice or checking two documents that disagree.
- In-tray or inbox exercise. The applicant handles a set of messages and records and decides what to do first, what to hold and what to pass on.
- Hands-on task. The applicant performs a procedure with the tools of the job, which usually needs the real tool or a close copy.
- Spoken or live task. The applicant handles a call or a role-play. This needs a person to play the other side and a way to record and review it.
| Kind | What the applicant produces | What the reviewer reads | Main limit |
|---|---|---|---|
| Written task | A reply, summary or note | The text and the choices behind it | Does not show voice or live manner |
| Analysis task | A worked answer from records | The answer and the steps taken | Needs clean, realistic records |
| In-tray exercise | Decisions across several items | What was done, held or passed on | Needs a clear rule for what a good order looks like |
| Hands-on task | A finished procedure | The result and how it was reached | Needs the tool or a close copy |
| Spoken or live task | A recorded conversation | The conversation | Needs a role-player and takes reviewer time |
When should you use a work sample test?
Use one when the job has output you can read and a short, bounded piece of it can stand in for the whole. These conditions help:
- The task is a real, common part of the job, not a puzzle.
- Every applicant gets the same brief, records and time.
- Someone who knows the job can review the work, and you have time to do it.
- You can say before you start what a sound answer must include, so reviewers read the same way.
- Applicants are told what is recorded, who reads it and whether AI tools are allowed.
It fits poorly when the core of the job cannot be shown in a short exercise, when no one has time to review the work, or when the task would need access to live systems or confidential records.
How is a work sample test different from a skills test, an interview or a job simulation?
A skills test usually scores answers against a key. An interview collects what applicants say. A work sample collects what they produce. A job simulation is the wider family: a work sample is one kind of it. See work sample vs skills test, work sample vs structured interview and what a job simulation is.
What does research say about work sample tests?
Sackett, Zhang, Berry and Lievens (2022) re-estimated validity across selection methods and put work-sample tests at .33, with assessment centres at .29 and situational judgement tests at .26. Most of the work-sample evidence came from studies of current employees rather than applicants. Moderate validity is not high validity, and it says nothing about how a specific exercise will perform for a specific employer. The full figures and their limits are in what research says about seeing applicants perform real work.
What are the risks of a work sample test?
- Inconsistent review. Two reviewers can read the same work differently unless they agree what to look for first.
- Applicant time. A long exercise costs applicants unpaid time and can drive away good ones. Keep it to one bounded task.
- Access and fairness. Offer a way to ask for adjustments and check the task does not depend on something unrelated to the job.
- Unclear AI rules. Say whether applicants may use AI tools, and review the work with that rule in mind.
- Using it to get free work. The task should be an exercise, not a piece of your live backlog.
How does Calavion fit?
Calavion prepares a work sample test with a company from a case the company has already handled. Applicants work inside it, and the employer reads a Replay of what each applicant did and wrote. Calavion does not produce scores, rankings or hiring decisions. The employer, not Calavion, makes the shortlist. Calavion has not demonstrated predictive validity, fairness or better hiring outcomes, and the only exercise it has run end to end is a customer service cancellation case. See the customer service work sample test, work sample tests by role and how to bring a case your team has handled.
Sources
- US Office of Personnel Management. Work samples and simulations. opm.gov
- 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
Common questions about work sample tests
What is a work sample test?
A work sample test is a hiring exercise in which an applicant does a small, realistic piece of the job and the employer reviews what they produced. It asks for the work itself instead of an answer about the work.
Is a work sample test the same as a take-home assignment?
Not quite. A take-home assignment is one way to deliver a work sample: the applicant does it on their own time over a stated period. A work sample can also be done on screen in a fixed session. What makes it a work sample is that the task mirrors the job, not where or when it is done.
Does a work sample test predict job performance?
Research reports moderate validity, not high. A 2022 re-estimate by Sackett and colleagues put work-sample tests at .33, and most of the evidence came from studies of current employees rather than applicants. That does not tell you how a particular exercise will perform for your roles.
Are work sample tests fair and legal?
It depends on the exercise, how it is reviewed and where you hire. Employers usually need to check that the task relates to the job, that applicants get the same brief and conditions, and that accommodations are available. This page is information, not legal advice; ask your own counsel before relying on any test.