Home Company Interview Guides

Situational Judgement Test: What It Is and How to Answer It

By Jacob Gates 9 min read Updated Nov 19, 2026

Job candidate in a navy sweater thinks through a question at a laptop with an open notebook on the desk

A situational judgement test asks what you would do in a work scenario, then scores your choice against answers built by the employer or the test publisher. It measures judgment, not knowledge, and the format tells you how your answer gets read. US employers usually spell it judgment.

You meet one as an online stage, usually after the application and before a person talks to you. It can arrive as a leadership test in retail, a warehouse screening, a public-safety application or a health professions admissions requirement. Scenarios are short. Scoring isn’t as obvious as it looks.

What is a situational judgement test?

A situational judgement test (SJT) is a scored set of workplace scenarios. You pick, rank or rate the responses, or write your own, and the result goes into the hiring decision. No item checks whether you know the job. Every item checks what you would do next.

Two other tests get confused with it. A personality test asks who you are and has no right answer. A skills test asks what you can do and has a correct one. An SJT sits between them: several answers work, and the scorer wants the one that holds up under pressure.

Viewpoint Psychology, which sells one test catalogue, describes the usual task as choosing which responses are most and least appropriate, with 24 role-specific versions covering jobs from barista and customer service to warehouse operative and firefighter. Each version is written around situations from that job. A generic “be professional” answer fits every role, which is why it proves nothing about yours.

Two colleagues discussing a workplace situation

Four formats, four different scoring rules

The format is the scoring rule. Two candidates can give the same advice and score differently, because one clicked a button and the other wrote a paragraph. Which one are you sitting?

  1. Most and least effective. You mark two options out of four or five. The scorer reads one pair. Picking the most effective is the easy half. The least effective box is where careless candidates lose points, because “harmless” and “least bad” aren’t the same answer.
  2. Rank every option. You order all of them. A near-miss in the middle still costs you, so two options that both sound reasonable still need an order. Ask which one you’d do first if you could do only one.
  3. Rate each option on its own scale. Every statement gets a score, from strongly disagree to strongly agree. Agreeing with all of them reads as no judgment at all. One weak option should get a weak rating, even when the scenario makes you want to look helpful.
  4. Written or recorded answers. Casper, a test used in health professions admissions, asks open questions about each scenario and accepts typed and video responses. A trained rater reads the reasoning, so the wording carries the score.

In the first three formats, a human sees nothing but your clicks. Nobody reads your tone or your intent.

That’s why the same person can pass one SJT and fail another: the second test asked for an ordering, and the candidate answered as if a single best choice was enough.

So which format is your test? The invitation email usually names the publisher, and the publisher’s practice page shows the task before you start. If it doesn’t, the first two screens tell you: clicking one option means most/least, dragging rows means ranking, and a sliding scale per statement means rating. Reading that shape early is worth more than another practice set, because it tells you what the scorer can and cannot see.

How the score is built, and what counts as a good one

Most SJTs are scored against a key built by the publisher or the employer, not against a fixed pass mark. Acuity Insights says Casper scores go to the programs you select, and each program decides how to use them. So “good” depends on the school or the employer in front of you.

Casper is also scored by people. The company says its certified human raters work through a structured process, and it states plainly that the test isn’t scored by AI. That single sentence changes how you should write a free-response answer: a rater is checking whether your reasoning names the people involved and the trade-off you accepted.

For employer tests, the useful question isn’t “what score passes” but “what is this stage for”. Some SJTs screen candidates out. Others rank them against each other and the hiring manager reads the band. The employer’s own instructions usually say which, and if they don’t, ask the recruiter before you sit the test.

One more wrinkle comes from the same catalogue. When a test is used for development rather than hiring, the candidate report shows no scores at all. It gives suggestions instead. A test with no score can’t be passed or failed, and candidates who treat feedback as a verdict waste the exercise.

Run three questions on every option

Run the same three questions on every option before you commit. They take about fifteen seconds once you’ve practiced, which is what a time limit punishes: not slow reading, but re-reading. In the coaching calls I run, this filter fixes most answers before the practice set does.

  1. Who is affected? Name the customer, the colleague, the patient or the driver. An option that touches nobody is usually a dodge, and dodges score low on customer-facing tests.
  2. What happens next for that person? The strong option tells them what you will do and who will decide. “I would inform my manager” beats “I would handle it”, because the second one hides the handoff.
  3. Which written rule binds you? Follow the policy in the order it is written. Breaking it to look helpful is the trap answer, and the trap is deliberate: it reads generous until you notice it puts the company at risk.

Here’s the filter on a common item. A customer wants a refund you aren’t authorized to give. The options are to refuse and move on, to approve it to keep them calm, or to explain the limit, offer the exchange you can do, and call the manager. The third names the person, states the next step and stays inside policy. Say that, not “I would go above and beyond for the customer”.

The least effective option is usually the one that tells nobody and changes nothing, or the one that solves today and creates a complaint tomorrow. If two options look equally good, prefer the one that keeps a record or brings in someone with authority. Scorers built those keys from what experienced staff actually do, and experienced staff escalate in writing.

For a written test, run a second pass against the list the publisher itself prints. Casper names nine qualities its raters weigh: collaboration, communication, empathy, ethics, fairness, motivation, problem solving, resilience and self-awareness. A draft that shows three of them in concrete terms beats one that gestures at all nine.

Where US job seekers actually meet one

For hourly and entry-level candidates, SJTs arrive through three doors: customer-facing and warehouse versions sold to employers, health professions admissions, and public-safety hiring. The role list in that catalogue includes barista, customer service, contact centre, warehouse operative, nurse, firefighter and train driver tests.

Retail workers describe the same shape. One Publix employee explained the department test this way:

“For the situational awareness answer with choice that best helps the customer and trains your associates at the same time. For the LSA test choose every option that makes it seem like you are the best thing since sliced bread and work well with everyone.”

That advice is half right. Helping the customer and the team at once is the pattern the key rewards. Ticking every box that flatters you isn’t, because rating scales punish a candidate who never disagrees with anything.

The other honest report comes from candidates who tried to outguess the key:

“I decided that I would respond authentically, rather than how I thought they’d want me to respond. I responded to each statement individually, not expecting that there would necessarily be a mix of good and bad responses.”

Where the test sits in the process sets its weight. When an SJT is one stage of five, a strong score won’t rescue a weak interview, and a middling score won’t sink an application that’s otherwise solid. Treat it as one input and stop rehearsing once your answers are consistent. Is a retake allowed? That’s the recruiter’s call, not the test’s.

Employers who use a broader pre-employment assessment often fold SJT items into it. Our breakdowns of the Walmart assessment test, the Home Depot assessment and the CVS assessment test show what each of those stages asks before the interview.

Is ChatGPT good at a situational judgement test?

Not on its own. On a written SJT a person reads the answer against a rubric, and Acuity says Casper is scored by certified human raters, not AI. A tidy answer that names nobody and commits to nothing gives that rater nothing to credit.

A language model helps with the part before the answer: sorting the options, spotting the one that breaks policy, checking whether your reasoning names a person. It’s poor at the part that earns marks, because its default is the balanced middle. Ask for a draft, then replace every general phrase with the actual people in the scenario and the actual next step.

Raters also read a lot of answers in one sitting. The one that sounds like a policy document reads as borrowed. The one that sounds like a shift supervisor explaining a decision reads as judgment, and judgment is the thing being scored.

I’ve read the same preference in interview notes for years. The interviewers I worked with at Wells Fargo and Bank of America wanted the same three things this test scores: the person, the next step, and the rule you followed.

Frequently asked questions

What is considered a good score on a situational judgement test?

There’s no universal cut-off. On Casper, programs decide how to use the score, so a number that’s strong for one school can be ordinary for another. On an employer test, ask whether the stage screens candidates out or ranks them, then aim for consistent answers rather than a guessed target.

How long does a situational judgement test take?

It varies by publisher. One graduate version is untimed and takes up to 15 minutes for 16 situations. Casper runs time-limited scenarios and scores each question separately, so the clock rewards a quick structure rather than a long paragraph. Check the invitation email for the format before you plan your time.

Can you fail a situational judgement test?

You can fail a stage that has a cut-off, and some public-safety and corrections candidates report retake limits after repeated attempts. You can’t fail a development version, because that report carries no score. Read the instructions to see which one you’ve been sent.

The short version: answer as the employee you’d want on that shift, name the person, state the next step, stay inside the policy. That’s the judgment the test was built to find, and it’s the same judgment the interview will ask about next.