How it works, and why

Every step of the daily 30-minute session has to earn its place with research on how people learn. If a step asked for time and the evidence didn’t back it, we removed it. This page lists each step, the evidence behind it, and how strong that evidence is.

Grades: A = meta-analyses or randomized experiments that agree · B = large observational or quasi-experimental studies · C = indirect or our own judgment. The grades rate the learning method. Most of those studies used college students and school material, not SAT questions, so for SAT points specifically every step is untested.

The questions

Official College Board digital SAT questions: the full-length practice tests and the SAT Suite Question Bank. Two practice tests are kept out of the daily session, for full-length, timed practice.

The daily session

1Study cardonly when needed
Method: A
A question missed again on a re-test comes back as a worked solution, next to the answer picked last time. Nothing to answer that day; the retry comes three days later.
Why: For a skill still being learned, studying a worked solution teaches more than attempting another problem unaided. Using it after a repeat miss is our judgment call (C).[8][15]
2Warm-up~4 min
Method: A
Up to three earlier misses that are due, back for another try; a correct answer marked as a guess counts as a miss. They return after 1, 3, 7 and 10 days. When more are due, up to three more ride unlabeled in the mixed set.
Why: Retrieving an answer from memory strengthens it more than rereading it, and spreading the retrievals out makes it last. The 1/3/7/10-day schedule itself is our choice, not a tested one (C).[1][2][3]
3Drill~8 min
Method: A
Five questions on the weakest skill (early on, the least-practiced one), labeled, with feedback and an explanation after each. When a math and a reading-and-writing skill are about equally weak, math goes first.
Why: Targeted practice with immediate explanatory feedback. The math tie-break rests on controlled coaching studies finding math gains about twice verbal gains; applying that to one day’s choice is a convention (C).[5][6][9]
4Mixed set~13 min
Method: A math / C reading
Ten questions from both sections, unlabeled, with a countdown and a quiet per-question pace hint. Three more questions of the drill skill are hidden among them.
Why: Mixing problem types makes the student decide which method applies before solving. Interleaving beats blocked practice on later tests for math; for reading and word learning the evidence is weak or mixed (C there).[4][10]
5Review~5 min
Method: A
Every miss, with the official explanation of how it is solved.
Why: Explanations of the correct answer outperform right/wrong feedback, and they help most on new questions, not just the one that was missed.[5][6][7]

What we deliberately left out

Streaks and points
Nothing in the research ties them to learning, and a broken streak punishes a missed day.
Required written error notes
We used to ask for a sentence explaining each miss. The evidence that error logs raise scores is thin, so reading the solution replaced it.
A “why it tempts” label on every wrong choice
No study shows labeling each trap helps, and reading wrong answers without a correction can make them stick. A short hint appears only for the choice actually picked, only when it says something specific, and always with the explanation.
Score predictions from practice accuracy
Practice accuracy isn’t a test score. Judge progress on official full-length practice tests.

On trap labels: [7][11]

What to expect

Honest numbers: controlled studies of SAT coaching find gains of about 20–30 points for a typical program, with math gaining about twice as much as the old verbal section.[9] A 2025 meta-analysis of experimental test-prep studies finds a modest average effect, and little evidence that practice tests alone help.[14] The much larger numbers in advertising come from studies without a comparison group.

Our own rough estimate for a semester of daily practice, a few official full-length practice tests and one retake is about 50–100 points. That combines several studies; no study measured it. Judge progress on official full-length practice tests, not on accuracy inside the app.

No study has tested this session on the digital SAT. Each piece rests on learning research done elsewhere, and whether it carries over to SAT reasoning questions is unproven.

For parents