United States
Hong KongWhy there is no single FTCE pass percentage
Because there are three separately scored multiple choice subtests with three different published cuts - at most 63 percent, at most 70 percent and at most 63 percent - and no combined score at all. A candidate at 68 percent across the board passes two subtests and fails one.
Type the question into a search box and you will be given a number. Usually one number, stated confidently, sometimes with a decimal point on it. It is invented, and the arithmetic below shows why it has to be.
Three cuts, not one
Each subtest passes at a scaled score of 200. The Department separately publishes what that scaled 200 is worth as a share of correct answers, per subtest, and the three figures are not the same.
| Subtest | Questions | Published maximum for a scaled 200 |
|---|---|---|
| English Language Skills | 30 | At most 63 percent |
| Reading | 30 | At most 70 percent |
| Mathematics | 35 | At most 63 percent |
Reading sits seven percentage points above the other two. That is not a rounding artifact. It is a different subtest with a different standard, set separately.
The worked example
Take a candidate who scores 68 percent on all three multiple choice subtests. A single blended pass mark would say they passed comfortably, or failed comfortably, depending on which blended figure you invented.
| Subtest | Their share correct | Published maximum | Outcome |
|---|---|---|---|
| English Language Skills | 68 percent | At most 63 percent | Pass |
| Reading | 68 percent | At most 70 percent | Fail |
| Mathematics | 68 percent | At most 63 percent | Pass |
Two passes and a fail, from one identical performance. There is no single percentage that produces that result, because the result is not made of one percentage. This is the entire argument and it takes about ten seconds to check.
There is also no combined score
Even if the three cuts were identical, there would be nothing to combine. The Department's rule is that where an examination has multiple subtests, an examinee must achieve passing scores on all of them to be recorded as having passed at assessment level.
Not an average of them. All of them. Add the essay, which is not scored in percentages at all - two raters, scores summed, passing total 6 - and any attempt to express this test as one number has to silently drop a quarter of it.
Why averaging the three is the worst version of the error
Average 63, 70 and 63 and you get a figure in the middle sixties that describes no subtest anybody sits. Use it to judge readiness and you have set a target too low for Reading, which is the subtest with the highest cut, and slightly too high for the other two.
So the blended number is not just imprecise. It is wrong in a specific and predictable direction: it points people away from the subtest that actually has the highest bar. That is our reading of it, and it is the reason this page exists.
The maximums are ceilings as well
One more layer, and it is the one most third-party pages miss. Those percentages are published as maximums. A harder version of a subtest may need slightly fewer correct answers for a scaled 200. Never more.
So even the three real numbers are upper bounds rather than thresholds. Treat them as "no worse than this" and you are reading them correctly. Maximum percent correct, explained works through what that means for a target.
What to aim at instead
Set a target per subtest, above that subtest's published maximum, and hold yourself to it separately. Reading needs the highest target of the three. Mathematics and English Language Skills need the same target as each other and it is lower than Reading's, which surprises almost everyone.
The concession: our own readiness target is a number we picked, measured against our own papers, and it is not a pass mark. Nobody outside the Department can tell you the raw count that clears a scaled 200 on the version you will sit, and we are not going to pretend otherwise.
Common questions
What percentage do you need to pass the FTCE General Knowledge Test?
There is no single percentage. Each subtest has its own published maximum for a scaled score of 200: at most 63 percent for English Language Skills, at most 70 percent for Reading and at most 63 percent for Mathematics. The essay is not scored in percentages.
Can you average the three percentages?
No, and doing so is actively misleading. An averaged figure sets a target below Reading's cut, which is the highest of the three. Each subtest is passed separately, so each needs its own target.
Why is Reading's cut higher?
Because standards are set per subtest. Reading's published maximum for a scaled 200 is at most 70 percent, against at most 63 percent for the other two. It is also the subtest with the highest first-attempt pass rate, at 78 percent.
Is 63 percent the pass mark for two subtests?
It is the published maximum for a scaled 200 on those subtests, which is a ceiling rather than a fixed threshold. A harder version may require slightly fewer correct answers. It will never require more.
What about the essay?
The essay is scored by at least two independent qualified raters whose scores are summed, with a passing total of 6. No percentage applies to it at all, which is another reason a single number cannot describe this test.