Every score carries measurement error
No test measures ability exactly. On a given day you might know a word you usually would not, or misread a question you would normally get right, and a different form of the same test would ask a slightly different sample of questions. The spread this creates is captured by a statistic called the standard error of measurement.
College Board makes the same point about the real SAT: every score report shows a score range, derived from the standard error of measurement, for how much the scores would likely vary on a different administration of the test under identical conditions. Read a practice result the same way: 1310 means "around 1310", not 1310 exactly.
Why a trend beats a single result
Measurement error partly averages out across several results, which is the idea the score band on the Studeous home page illustrates. Practice tests are only roughly comparable, though: forms differ, and you learn between attempts. So more tests are a reason to trust a trend, not a formula for how narrow your own range becomes.
Two things follow. A change between two single tests can be noise rather than progress. And a trend across several tests is far more trustworthy than any one of them.
What a practice score is good for
- Placing you on the scale, as a band, before you plan how much preparation you need.
- Showing which sections and question types cost you points, which is where study time pays off.
- Telling you whether time is part of the problem, if the test was taken under real timing.
- Tracking a trend across several tests taken a week or two apart.
What it cannot tell you
- Your exact score on test day. Nerves, sleep and the particular form you get all move it.
- Anything reliable if you took it untimed, with pauses, or with notes open.
- Progress, if you retake a form you have already seen. A second score on the same questions measures memory of that form as much as ability — which is why Studeous marks a form you have already sat.
Using practice scores well
Take full tests under the real conditions, space them out, and spend more time reviewing each one than taking it. Judge progress by the trend and by the shrinking list of error types in your review, not by a single number. For a plan built on that approach, see the digital SAT study plan.
Common questions
- Are practice test scores accurate?
- A well-built practice test is a good estimate of your current level, within the margin of error every test has. Read it as a range, and trust a trend across several tests more than any single result.
- Why did my practice score drop?
- A drop of a few tens of points between two single tests is often measurement error rather than a real decline, especially if the forms differ. Look at the trend over several tests and at which question types changed.
- Is it worth retaking the same practice test?
- Reviewing it is; rescoring it is not. A retake of a form you have seen measures familiarity with those questions, so the new score overstates your ability.
Sources
- College Board — What Do My SAT Suite Scores Mean?
- College Board — The SAT
- ETS — Understanding Your GRE Scores
Test owners change formats and policies. Where this guide and the owner’s own page disagree, the owner’s page is right — and we would like to know, at contact.