What One ACT Point Actually Costs, in Questions

Going from a 24 to a 27 on the Enhanced ACT costs twelve questions. Four more correct in English, four more in Math, four more in Reading. That is the entire distance.Points and questions are different units, and almost nobody converts between them. Score reports are denominated in points. Studying happens in questions. Below is the conversion, computed from the published raw-to-scale table, with the arithmetic shown.

The denominator is 131

The Enhanced ACT composite is the average of three scale scores: English (50 questions), Math (45) and Reading (36). Science is optional and reported on its own, so it does not enter the composite. That leaves 131 scored questions deciding a number from 1 to 36.

Because the composite is a rounded average, a composite of C requires the three section scores to sum to at least 3C minus 1. A 27 needs the three sections to total 80. A 24 needs 71.

What one point costs

The table below uses a balanced reference profile: the same scale score in all three sections, at the fewest correct answers that reach it.

The first thing to notice is the size of the numbers in the last column. Across composites 18 through 35, a single point never costs more than 6 questions and never fewer than 2.

Where the curve is steep, and where it is flat

Single-point costs bounce between 2 and 6 because of rounding, so the three-point jump is the stabler unit. Grouped by starting composite:

• Starting at 12 through 17, a plus-three costs 14 to 17 questions, averaging 16.
• Starting at 18 through 28, a plus-three costs 10 to 13 questions, averaging 11.5.
• Starting at 29 through 32, a plus-three costs 6 to 8 questions.
• Starting at 33, the final three points cost 12.

The list starts at 12 on purpose. Below that, the low end of every section curve is compressed so hard that the balanced profile returns increments that mean nothing in practice: one composite point costs zero additional questions between 4 and 5, and one question each between 2 and 3, 6 and 7, and 8 and 9. Those are artifacts of the conversion, not study advice.

Inside the range that describes real test takers, the expensive stretch is the low teens, not the top. The costliest plus-three anywhere on the scale is 17 questions, and it sits at starting composites 12, 13 and 15. A student moving from 15 to 18 pays 17 questions. A student moving from 27 to 30 pays 11. The three cheapest single points anywhere from 18 up are 30 to 31, 31 to 32, and 32 to 33, at two questions each.

The one place the curve genuinely bites is the last point. Going from 35 to 36 costs 6 questions, more than any single point between 18 and 35.

For anyone in the high teens or low twenties, this means the answer to "where should I spend my time" is close to "anywhere," because every point in that band costs between 3 and 5 questions. For a student already at 29 or 30, the next three points are the cheapest stretch above a composite of 18: 7 questions from 29, and 6 from 30, which is the cheapest three-point jump on the charted range. Stopping there is the most expensive habit the curve exposes.

The per-section picture

Sections are not equal currency. Divide the questions each section needs to travel from the bottom of its scale to a 36 by the 35 points of travel:

• English: 50 questions over 35 points, or 1.43 questions per scale point
• Math: 44 questions over 35 points, or 1.26
• Reading: 36 questions over 35 points, or 1.03

Reading is dense. It covers 35 points of travel in 36 questions, so a Reading question is worth about a full scale point on average, where English takes 50 questions to cover the same ground.

Across 17 overlapping six-point windows, from 14-to-20 up through 30-to-36, Reading is the cheapest or tied-cheapest section in 15 of them. Math is the sole cheapest in exactly 2, the windows starting at 20 and 21.

Concretely: to move a section score from 24 to 30, English needs 8 more correct, Math needs 9, Reading needs 6. To move from 30 to 36, English needs 5, Math needs 8, Reading needs 5. Math is the expensive section at the top.

The table also contains questions worth double. Above a section score of 15, five single Reading questions and three single English questions each move the section score by two points at once. Math has none.

And it contains questions worth nothing. Above a section score of 15, 13 raw steps in Math buy zero scale points, along with 10 in English and 5 in Reading. One more Math question right, at the wrong place on the curve, changes the reported score by nothing at all.

That cuts both ways in Reading, which is why "on average" is the right qualifier: across the whole section, 10 of its 36 raw steps buy nothing and 9 of them buy two points.

A related quirk worth knowing: not every scale score is reachable in every section. Reading skips 9 of the 36 possible scores, including 27. A student who "gets a 27 in Reading" is actually getting a 28.

Why "get everything right" is the wrong target

At a composite of 24, a student can miss 42 of 131 questions. At 27, 30. At 30, 19. At 33, 13. Only at 36 does the margin nearly close: 130 of 131 correct, with the single allowed miss falling in Math, which reaches a 36 at 44 of 45.

"Answer everything correctly" is not a target. It is a description of a 36 and of nothing else on the scale. It also has a cost in time, because minutes spent on the questions least likely to come out right are minutes not spent on the ones that would have.

The target is a count. A student at 24 aiming at 27 needs 12 more correct answers on a schedule of 4, 4 and 4. That is a checkable weekly goal in a way that "get better at reading comprehension" never is. It is also small enough to be believed, which matters more than it sounds.

Method

The counts come from the published raw-to-scale conversion for the Enhanced ACT, structured as data. What is borrowed here is that conversion table, and nothing else. Three caveats belong with any citation of it.

First, ACT equates every administration separately, so the conversion describing one released form is not literally the conversion for the form a student sits.

Second, the released Enhanced ACT practice form scores only a subset of its items: 40 of 50 in English, 41 of 45 in Math, 27 of 36 in Reading and 34 of 40 in Science. The remainder are unscored field-test questions, so the published conversion covers only the scored range. The table used here stretches that curve across every question in each section and forces monotonicity. Treat every count as an estimate within roughly one scale point.

Third, these figures use the balanced path, meaning the same scale score in all three sections. A pure mathematical minimum exists and is lower, but it reaches its totals through extremely lopsided profiles, such as near-perfect Reading paired with almost nothing in English, which exploits the compressed low end of the curves and describes no real test taker. The balanced reference is the honest one.

Nothing here is a forecast. It is a conversion between two units, points and questions, and it says what a target costs, not what any particular student will score.

Prep36 questions are original practice questions written to the Enhanced ACT format. Not affiliated with ACT, Inc.

Frequently asked questions