How Focus on Prep Works

Our questions are original items written to match the format, style, and difficulty of the Digital SAT. They are produced by an in-house content pipeline rather than copied from any official test.

Before a new item can enter the live question bank, it has to pass a set of automated verification checks. An item that fails these checks does not get served to students. This keeps obvious errors, malformed answer choices, and unsupported claims out of the pool before anyone sees them.

Automated checks at creation time are only the first step. We keep watching questions after they go live, using three layers.

a) You can report any question

Every live question has a Report control. If something looks wrong, you can flag it in a few seconds, and your report goes straight into a triage queue that our team reviews.

b) Automatic statistical monitoring

Automated checks recompute each question’s statistics from real answer data as students answer it. They look for statistical warning signs, such as a question where stronger students do worse than weaker ones (a negative point-biserial correlation) or a wrong answer that pulls in an unusual share of responses (a dominant distractor). Questions that trip these thresholds are flagged for a person to look at.

c) Human review, and retirement when needed

Reported and flagged questions are reviewed by our team. A question can be corrected and returned to the pool, or retired so that it is no longer served. Retiring a question removes it from practice and tests going forward.

We show your estimated score as a range, not a single number, because a range is an honest reflection of the uncertainty involved.

The estimate comes from simulating many full-length adaptive tests for you. For each section we run a set of simulated tests that follow the same two-module routing the real exam uses, then report the middle band of the results as your range. The simulations are driven by your per-section skill ratings, which we update from your answers over time.

A section needs at least 30 answered questions before we show a score range for it. Before you reach that point, anything shown is an early projection based on limited data, and it is labeled that way so you know not to lean on it yet.

The suggestions you see are meant to point you at the practice where you can make the most visible progress soonest, not at an abstract ideal study plan. To rank areas, the recommender weighs several signals together:

• How much room you have to improve in an area, since there is more to gain where you are not yet solid
• Where you are weakest relative to your own average, so attention follows your real soft spots
• How long it has been since you last practiced an area, so nothing goes stale
• Questions you have gotten wrong and not yet cleared

Every recommendation shows the reason it was chosen, so you can see why an area came up rather than being asked to trust a black box.

Full practice tests follow the same two-stage adaptive structure as the Digital SAT. You take a first module of mixed difficulty, and the difficulty of the second module is then set by how you did on the first, the same routing the real test uses.

Tests run on a real clock, with the on-screen tools you get on test day, including a reference sheet and, for the Math section, a graphing calculator. When you practice, questions you have already seen are held back for a period before they can appear again, so you are not re-drilling the same items day after day.

We think it is fair to tell you exactly where AI is involved.

AI tools help our team draft practice questions and explanations inside the content pipeline. Anything drafted this way then goes through the same steps as everything else: automated verification before it is served, ongoing statistical monitoring once it is live, and human review when it is reported or flagged.

AI does not grade your answers, and there is no AI chatbot in the product. Your score estimates and your practice recommendations are produced by statistical models with rules we have described above, not by an AI making a guess about you.

Your study data belongs to your account and is private. Other learners cannot see your results, and there are no public profiles or shared leaderboards that expose your progress to anyone else.

• Payments are handled by established payment processors, and we never store your full card number.
• You can ask us to delete your account at any time, and we remove your synced study data when we do.
• Questions are delivered to your browser without their answer keys, and answers are graded on our server, so the correct answer is never sitting in the page waiting to be read.

For the full details on what we collect, how long we keep it, and how to request deletion, see our Privacy Policy.

SAT is a trademark of College Board. Focus on Prep is an independent study tool and is not affiliated with, authorized by, or endorsed by College Board or ETS.