How Casper Is Scored — and What Your Quartile Actually Means

Every Casper response is scored by a trained human rater — Acuity Insights states plainly that "Casper does not use AI for scoring, rating, or evaluating Casper responses" — and what you eventually see is a single quartile, about 4-5 weeks after test day. That's the whole disclosure. No number, no per-scenario breakdown, no feedback. This page explains what happens between your last typed answer and that quartile, what programs see that you don't, and what a "good" result honestly means.

Who scores your answers

Your responses are split up, anonymized, and sent to trained, certified, monitored human raters — a different rater for each scenario, so no rater sees more than one set of your responses. Raters are instructed to focus on the substance and quality of what you said, not the polish: typos are explicitly not penalized, and grammar isn't the axis. Acuity's official rating page is the load-bearing source here, and it's worth reading in full.

There is no answer key. Casper is an open-response judgment test, which means your answer isn't checked against a correct one — it's judged against how everyone else handled the same situation. That single fact explains almost everything else on this page, including why nobody can honestly promise to raise your quartile.

From 22 answers to one z-score

A full test is 11 scenarios with 2 questions each — 22 answers. Each question gets rated, question-level ratings are averaged, and the result is standardized against the applicant pool that took the same test instance. The statistical output is a z-score: a measure of where you sit relative to everyone else who tested when you did. Scoring is relative by construction. There's no official difficulty scale and no fixed passing bar; the cohort is the bar.

That's also why identical-feeling performances can land differently. Test takers commonly report score swings between cycles with similar answers — that's reported experience, not a measured fact, but the mechanism makes it plausible: a different cohort is a different denominator.

What you see vs. what programs see

The disclosure is asymmetric, and it's worth being precise because most prep pages aren't:

Programs receive your z-score and percentile, roughly 2-3 weeks after you test.

You receive only a quartile, roughly 4-5 weeks after you test. The quartiles are Acuity's own published bands: first quartile is the 0-24th percentile, fourth is the 75-100th, with the second and third covering the 25% bands between. Those are band definitions, not our estimates. And there is no per-topic breakdown for anyone — what programs get is a single combined number.

One genuine ambiguity, documented rather than smoothed over: Acuity's Terms of Use state the quartile reflects "how you have performed on the typed section" relative to other applicants, while its Help Centre article on quartiles doesn't specify a scope at all. One official page says typed-section; the other is silent. We won't assert either reading, and you should be suspicious of any site that does. Acuity reports a quartile; what exactly it covers is a question only Acuity can answer.

What counts as a "good" Casper score?

Fourth quartile means you landed in the top 25% of your test instance — by definition, per Acuity's bands. That's the honest ceiling of what anyone can tell you, because how much your quartile matters is a program-level decision, not a test-level one. Some schools put explicit, large pre-interview weight on Casper. Some use it as one signal among many. At least one — Dalhousie Medicine — applies a minimum-score cutoff below which, per the AFMC admission guide, applicants "will not be considered"; the school does not publish the cutoff number.

So "is my second quartile bad?" has no universal answer. It depends on where you applied and how that program weighs the test — which is exactly why our full format guide keeps school-specific stakes separate from test mechanics. If you're deciding where your quartile leaves you, the program's own admissions page beats any prep site's guess, ours included.

What you can and can't do with your result

Three rules bound your options, all from Acuity's official policies. Your score is valid for one admissions cycle, one test type, and one language — it dies with the cycle. You get one Casper per test type per cycle. And retakes are considered only for a technical problem verified by Acuity's proctoring and security team, reported within one week — a disappointing quartile is not grounds for a second sitting.

Practically, that makes a weak quartile fixed information for this cycle; the decision it actually informs is next cycle's. Whether to test again is a judgment call. Walking in with the format's time pressure already familiar is not — test takers repeatedly report that their first sitting was also the first time they felt the 3.5-minute typed clock, and our timed practice sim exists so the clock isn't news on the sitting that counts.

A disclosure about our own numbers, because this page is about scoring: when we built our practice scoring we deliberately made it a 0-100 coaching signal computed from your own answers — an independent AI practice score, not an official Casper result. It is not a quartile, not a percentile, and not calibrated to Acuity's raters. Any tool that claims its practice score predicts your quartile is claiming something no one outside Acuity can know.

Scoring facts on this page were verified against Acuity's official pages on September 10, 2026.