How the two adaptive modules work and what a score means
Two modules, 22 questions each, 35 minutes each. Every question inside a module is worth the same — which decides almost everything about how you should spend the clock.
Which plan earns the most points?
With 5 minutes left in Module 1, Maya has 4 questions unanswered: one two-step linear equation she can finish in about 40 seconds, two medium questions of about 90 seconds each, and one hard geometry question she has already worked on for 3 minutes without progress.
Every question in the module is worth the same, so the only sensible order is cheapest first: 40 + 90 + 90 = 220 seconds buys three near-certain points and still leaves about 80 seconds to click a choice on the fourth. B is the only plan that collects all three cheap points and still answers the hard one. A spends the whole 5 minutes on one point and abandons three easier ones — the single costliest mistake on this test. C gives every question 75 seconds, which is not enough to finish either medium question, so it risks converting three sure points into three half-finished ones. D is B minus the guess: a blank and a wrong answer score exactly the same, so the blank throws away a free one-in-four shot.
1
The maths section is two separate modules, and the clock does not carry over.
22 questions in 35 minutes, then 22 more in 35 minutes. That is 44 questions in 70 minutes, and about 95 seconds per question.
Questions
Time
Per question
Module 1
22
35 min
95 s
Module 2
22
35 min
95 s
Maths section
44
70 min
95 s
35 × 60 = 2,100 seconds
what one module actually gives you
2,10022 = 95.45 seconds
about 1 minute 35 per question, on average
Time left over in Module 1 is gone when Module 1 ends; you cannot bank it. Neither can you return to Module 1 once you have moved on. Inside a module, though, you may move freely: answer, flag, skip and come back are all one screen away, and there is a review page listing every question with its status.
calculator on screen throughoutreference sheet on screenflag and return inside a moduleno return to a finished module
2
Module 2 is chosen by how you did in Module 1, so Module 1 sets your ceiling.
This is the one way the digital test differs from the paper one that actually changes your strategy. Module 1 is a fixed mix of easy, medium and hard questions and is the same for everybody. Your performance on it routes you into one of two versions of Module 2.
A student routed into the easier Module 2 can still answer all 22 of its questions correctly and will not reach the top of the 200–800 scale. Module 1 is where the ceiling is set; Module 2 decides where under that ceiling you land.
Two consequences worth acting on. First, treat Module 1 as the important one — bring your full attention to it rather than warming up. Second, if Module 2 feels noticeably easier than Module 1 did, that is information, not relief: answer every question anyway, because within the easier module each point still moves your score.
3
Inside a module every question is worth exactly the same.
Question 3 and question 20 are worth one point each. The hard one is not worth more for being hard.
budget per question = 95 s
the average you have to hold
5 minutes on one question = 300 s
what a stubborn question actually costs
30095 = 3.2 questions
you paid three points for one
One hard point bought at the price of three easy ones is a losing trade — every time
The trap in the opening problem
Choice A — “go back and finish the geometry question” — is how it feels from inside the test: you have already invested 3 minutes, so abandoning it feels like wasting them. Those 3 minutes are gone either way. The only question is what the next 5 minutes buy, and 5 minutes spent on the geometry question buys at most one point while three cheaper points expire. Sunk time is not an argument for spending more.
read the questionno route in 20 seconds → flag itno answer in 2 minutes → guess and flag itnever a third visit before the sweep
4
A wrong answer costs nothing, so a blank is a point you threw away.
There is no deduction for a wrong answer on the SAT. A blank and a wrong answer are scored identically — zero — but a blank cannot come good and a guess can.
What you do
Chance it scores
Expected points per 4 such questions
Leave it blank
0
0
Guess from 4 choices
14
1
Eliminate one, then guess
13
1.33
Eliminate two, then guess
12
2
4 blanks × 14 = 1 extra question right
for about 20 seconds of clicking
1 question ≈ 10 to 15 scaled points
see the conversion in the next step
Pick a letter now — say C — and use it for every blind guess on test day. A fixed letter removes one decision from the most rushed part of the test, and it stops you from staring at four choices you have no information about.
“I ran out of time” is usually “I left five blanks”
Running out of time is not what costs the points; leaving the unreached questions empty is. Those five questions were going to be wrong anyway — guessed, five multiple-choice blanks are worth about 1.25 points, roughly 15 scaled points, for no extra thinking at all.
One caveat, because it changes what a blank is worth: about a quarter of the maths section is typed rather than clicked, and on those there is nothing to guess between — a random entry is worth essentially zero. So the free points are in the multiple-choice blanks. Click a letter on every one of those before the clock stops, and spend any seconds left on a typed question you half-know rather than on one you have not read.
5
44 raw points become a section score between 200 and 800.
Maths is scored out of 800 on its own; Reading and Writing is scored out of 800 on its own; the two add to the 400–1600 total you hear quoted. Every test form has its own conversion, so the table below is approximate — but the shape of it is stable.
Questions right (of 44)
Approximate maths section score
44
800
40
about 760
35
about 700
30
about 650
25
about 590
20
about 520
14
about 450
8
about 360
0
200
30 right → 650, 35 right → 700
5 more questions, 50 more points
505 = about 10 points per question
nearer 15 lower down the scale
Two things follow. A 200 is what you score for answering nothing, so nobody scores below 200. And one careless question — a misread sign, an unfilled blank — is worth roughly ten points, which is the same as an entire topic you have not studied yet.
Set a target in questions, not in points. “I want a 650” is not something you can act on in a module. “I need 30 of 44, so I can miss 14” is: it tells you that you may walk away from three hard questions per module and still hit the number, which is exactly the permission most students need in order to stop grinding.
6
What you actually do in the last five minutes of a module.
Three checkpoints keep you on the 95-second pace, and the final stretch has a fixed job that is not “solve one more hard question”.
5:00 left — stop starting anything new
finish only a question already half solved
4:30 left — open the review page
it lists all 22 with their status
4:30 to 3:30 — fill every blank with your letter
about 5 seconds each, no thinking
3:30 to 0:20 — revisit flagged questions, cheapest first
each already has an answer, so any change is a gain
0:20 left — check the review page shows zero blanks
the last thing you do in every module
The order matters: blanks first, then flagged questions. Filling blanks is guaranteed expected value at five seconds a question; reworking a flagged question is a maybe at ninety. Never do the maybe before the guarantee.
✓
Check: the five numbers that run the section.
the module22 Q in 35 min95 seconds each ✓
the routingModule 1 → Module 2Module 1 sets the ceiling ✓
the weightingeasy = hard = 1 pointcheapest question first ✓
the guess4 blanks × 14 = 1a blank is worth strictly less ✓
the scale44 raw → 200–800about 10 points a question ✓
Remember the flow
Cheapest first · Guess and flag, never blank · Sweep the last five minutes
Leave every module with 22 answers on the screen.
Now you try
Practice problem
Which statement best describes the two section scores?
Two students each answer 26 of the 44 maths questions correctly. Student X answers 17 correctly in Module 1 and 9 in Module 2. Student Y answers 9 correctly in Module 1 and 17 in Module 2.
Within one module every question counts the same, but the two modules are not the same test. Student X's 17 in Module 1 routes her into the harder Module 2, whose questions map onto the upper part of the 200–800 scale; Student Y's 9 routes him into the easier version, whose ceiling is roughly 600 however well he then does. A treats 26 as the whole story — true on a paper test, false once Module 2 is chosen for you. C invents a rule: the test rewards the questions you get right, not the shape of your progress through them. D confuses blanks with information — a blank is simply a question answered wrongly, and both students' totals already account for every question.
Hint: ask which module each student did well in, not how many they got right overall. One of those two facts changes which Module 2 they sat.
