Why Homework Is the Key to SAT and ACT Score Gains: 5.7 and 190 Points

As of October 2026: Homework matters because the test rewards what a student has learned deeply, not time logged. In our published results, ACT students starting at 27 or below gained an average of 5.7 points, and SAT students starting at 1350 or below gained an average of 190 points, with 63% of them gaining 200 points or more.

Nick Standlea on the three-step homework method, the research behind it, and what our 782 students actually gained.

By Nick Standlea, Test Prep Gurus. Figures may be quoted with attribution to Test Prep Gurus.

Homework is Key to Your Long Term Test Prep Success on SAT/ACT -- Nick Standlea, Test Prep Gurus (YouTube).

Our completers gained 5.7 ACT points and 190 SAT points on average Those are the two headline numbers from our published results. ACT students who started at 27 or below gained an average of 5.7 points across a cohort of 491 students from 2017 to 2023. SAT students who started at 1350 or below gained an average of 190 points across a cohort of 291 students in 2024. Both figures describe students who completed the program. Our results page counts a student as a completer with 12 or more instruction hours and at least 80 percent homework completion. Homework is named right in that count, next to the hours.

63% of our SAT completers gained 200 points or more Among SAT students who started at 1350 or below, 63% gained 200 points or more. That is close to two out of every three students in a 291-student cohort. A 200-point move is a large change on a test scored out of 1600. It is also the kind of change that takes months of steady self-testing rather than a single long weekend of cramming. We publish the figure as a result for completers, not a promise for every student.

The two study methods rated high utility are both built into homework Dunlosky and colleagues (2013) reviewed ten common study techniques and gave a high utility rating to only two: practice testing and distributed practice. Five techniques got a low rating, including rereading and highlighting, the two things most students do by default. Homework done as practice testing, spread over weeks, runs both high-rated methods at once. The total pool of students behind our own published results is 782.

What do our published results show?

These are the figures we publish for students who completed the program. We quote them exactly as printed on our results page.

Measure	Value
ACT students starting at 27 or below: average gain	5.7 points
SAT students starting at 1350 or below: average gain	190 points
SAT students starting at 1350 or below who gain 200 points or more	63%
ACT cohort	491 students, 2017-2023
SAT cohort	291 students, 2024
Total students in the published results	782
Founded	2009

Quote these exactly as printed; they are averages for students who completed the program. The results page counts students who completed the program: 12 or more instruction hours and at least 80 percent homework completion (prepgurus.com/results-at-test-prep-gurus).

How did researchers rate common study methods?

Dunlosky and colleagues (2013) sorted ten study techniques into three utility ratings. We list them here as the study reported them.

Utility rating	Techniques
High utility	Practice testing; distributed practice
Moderate utility	Elaborative interrogation; self-explanation; interleaved practice
Low utility	Summarization; highlighting/underlining; keyword mnemonic; imagery use for text learning; rereading

Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Psychological Science in the Public Interest, 14(1), 4-58.

What did the retrieval and spacing studies actually test?

Two studies sit behind the learning science on this page. Here is what each one reported, in its own terms.

Study	What it tested	What it found
Roediger & Karpicke (2006)	Two experiments; free recall without feedback	Repeated study beat repeated testing at 5 minutes; repeated testing beat repeated study at 2 days and 1 week
Roediger & Karpicke (2006)	Final retention test given at one of three delays	5 minutes, 2 days, or 1 week
Cepeda and colleagues (2006)	839 assessments; 317 experiments; 184 articles	Longer retention interval means a longer best gap between study sessions
Cepeda and colleagues (2006)	Spacing versus massed practice	No single overall effect-size or percentage is reported in the abstract

Exact percentages for the Roediger and Karpicke intervals are not established from the available primary-source text, so we do not print any.

Why does homework matter so much for SAT and ACT score gains?

Homework matters because our own results are built on it: ACT students starting at 27 or below gained an average of 5.7 points, and SAT students starting at 1350 or below gained an average of 190 points. Both numbers describe students who completed the program. Our results page counts a completer as a student with 12 or more instruction hours and at least 80 percent homework completion. Homework is named right there with the hours.

Here is how Nick puts it in his own words. School is all about putting in the minimum amount of effort to achieve the maximum grade possible. Test prep is reversed. There is no grade for effort at the end. There is an official test that measures the concepts a student has mastered against the ones the student has not.

That changes what homework is for. In a class, homework is a task you turn in. In test prep, homework is the place where the learning happens. Nick says we have to prioritize learning and long-term retention when we attack our homework. Much of that time happens at home, where a parent sees only the edges of it.

So when a score stalls, the first place we look is not the lesson plan. Nick says the real learning and the real long-term retention happen in the steps after the answers are scored. That is not a scolding. It is a map of where the fix lives.

What are the three steps of doing SAT or ACT homework the right way?

There are three steps, and steps two and three should take just as long as step one took. Step one is doing all the problems and checking right answers against wrong answers. Nick calls that obvious, and he is right. The point he makes is that a student who stops there is only halfway done.

Step two is going back to the ones you got wrong and trying them again, armed only with the knowledge that you did not get them right the first time. No solution. No notes from anyone else. Just a second honest attempt on a problem you now know is hard for you.

Step three comes after that. Some problems will still not crack. Now the student reads the explanation in the book, or the solution attached to the practice test. Then, as Nick says, read it, close the explanation, and attempt that problem again. Not following along with the solution. Closing it and working the problem cold.

Only after three tries does a problem go to the next one-on-one session. By then there are usually three or four left. Those are the ones Nick says a student should bring in. The read for a parent is simple: if a student finishes homework in the same time it took to do the questions once, steps two and three did not happen.

Why is struggling with a hard problem better than being shown the answer?

Struggle helps because, as Nick puts it, the research shows that struggling with the problems is key to internalizing the information. He calls this leaning into desirable difficulties. The more a student struggles with a concept, the more likely that student is to retain it once they figure it out. The catch is that they do have to figure it out eventually.

This is why Nick asks students not to hand over every wrong answer right away. He understands the pull. He says he knows students are busy, over-scheduled, and want to speed the process up. But handing a wrong answer to an adult ends the struggle early, and the struggle is the part that sticks.

Homework is where this works best because the student is alone. There is no adult to lean on. Working a problem two or three times without help is uncomfortable, and that discomfort is the point. Nick says a student will learn it more deeply by figuring it out on their own.

For a parent, this reframes a familiar scene. A child sitting at the table frustrated by one math problem is not stuck in a bad way. That is the method running. The move that helps is not rescue. It is time and quiet.

Does the research really say that testing yourself beats studying again?

Roediger and Karpicke (2006) found that repeated study beat repeated testing after 5 minutes, but repeated testing beat repeated study at 2 days and 1 week. That timing is the whole story. The students in the study read prose passages and either took immediate free-recall tests without feedback or studied the material again.

The final retention test came at one of three delays: 5 minutes, 2 days, or 1 week. In the short window, study looked better. Over days and a week, the students who had tested themselves held on to much more. The exact percentages are not established from the available primary-source text, so we do not print any.

Think about which window matters for a junior. No one takes the official SAT or ACT five minutes after a study session. The test is weeks or months out. On that clock, the method that looked weaker in the lab is the one that wins.

This is the gap between feeling ready and being ready. Rereading notes feels smooth and fast, and it leaves a student sure they know it. A practice section feels rough and slow, and it leaves a student with real evidence. Nick's phrase for homework is self-testing, and that is close to what the study measured.

Which study methods do researchers rate as high utility?

Dunlosky and colleagues (2013) gave a high utility rating to only two techniques: practice testing and distributed practice. Three more got a moderate rating: elaborative interrogation, self-explanation, and interleaved practice. The remaining five got a low rating.

The low-rated five are the ones most families picture when they hear the word studying. They are summarization, highlighting and underlining, the keyword mnemonic, imagery use for text learning, and rereading. The authors wrote that the benefits of rereading are limited and that highlighting does not reliably improve performance.

Practice testing earned its high rating because the benefits held up across different learners and conditions, and because it does not take much extra setup. Distributed practice earned its rating because the gains are not limited to one age group or one subject, and it is easy to put in place.

Line that list up against a stack of SAT homework spread over several weeks and the overlap is hard to miss. A timed section is practice testing. Weekly sections over months are distributed practice. A student who skips the homework and rereads the book instead has traded the two high-rated methods for one of the five low-rated ones.

Why does mixing problem types beat drilling one topic at a time?

Mixing helps because Dunlosky and colleagues (2013) rated interleaved practice as a moderate utility technique, while drilling one topic in a block is not on the high list at all. Mixing problems up is harder in the moment. It is also closer to what the real test does.

Nick explains the reason in plain terms. When a student only sees one kind of problem in a row, they stop reading for the cue that tells them which kind it is. The test never lines problems up by topic. It mixes them, and the student has to spot the type first.

He uses a sports picture to make it land. Throw a hitter fifty curveballs in a row and the hitter looks great by the tenth one. In a real game, the pitch is a surprise, and the practice does not carry over. Mix the pitches up and the early work feels worse, but the hitter learns to read what is coming.

So a study set that looks scattered to a parent is not sloppy. Short-term it feels slower and less tidy. Over months, the student learns to see which problem is in front of them and which tool fits it. That recognition is a skill of its own, and it only grows when the problems arrive out of order.

How much time should spread-out practice cover before test day?

Cepeda and colleagues (2006) reviewed 839 assessments of distributed practice in 317 experiments located in 184 articles, and found that the gap producing the best retention grew as the retention interval grew. In plain terms: the further away the real test, the wider the spacing should be.

That review did not report one overall percentage or effect size for the spacing advantage in its abstract, and we will not invent one. What it did report is the direction and the scale of the evidence base. A pattern that holds across 317 experiments is not a fad.

For families, the practical read is about the calendar, not the stopwatch. Short sessions repeated over months beat one heroic push in the final two weeks. A student who starts in the fall and works steadily has the spacing the research describes. A student who starts in March does not.

This also lines up with the 12 or more instruction hours in our definition of completion. Hours alone are not the goal. Hours spread out, each one followed by homework that is itself spread out, is the shape the research supports. The abstract does not give us a best number of weeks, and we do not pretend otherwise.

What does 80 percent homework completion mean for my child?

Our results page counts a student as having completed the program with 12 or more instruction hours and at least 80 percent homework completion. That is the bar behind the 5.7-point ACT average and the 190-point SAT average. It is the line printed with the figures on our results page.

Eighty percent leaves room for a real life. A sick week, a tournament, a show, a college visit -- those happen, and the bar survives them. What the bar does not survive is a student who turns in the easy sections and skips the hard ones every time.

Parents sometimes ask whether the number is about compliance. It is not. It is about dosage. A student who does one section in four is not getting distributed practice and is not getting enough self-testing to find the gaps. The method only pays when it runs often enough.

The honest caveat is that our published figures describe completers. Our results page does not print a gain figure for students below that bar, so this page cannot tell you what happens at a much lower rate. What we can tell you is that the published averages were earned by students who met both parts of that definition.

What does a wrong answer on homework tell us?

A wrong answer tells us where the next score points live, which is why steps two and three of the three-step method exist. Nick is direct about it. The problems a student answered correctly are probably ones they have mastery over. The missed ones are where we dig in.

He describes homework as a small-scale self-test. Each section gives feedback on what a student knows, what they do not, and whether they are improving. The key, in his words, is that once you determine what you do not know, you correct that stuff.

A missed problem that never gets a second attempt does not go away. It shows up again in a new form on the official test. A missed problem that gets two more honest attempts turns into a concept the student holds for the long term.

There is one more habit Nick asks for. During step one, a student should circle any problem they had to guess on. A lucky guess still hides a gap. Circling it turns a coin flip into something to review later. Nick says the goal is to know the material cold, so guessing is a last resort rather than a plan.

What should the homework routine look like week to week?

Homework may be a full practice test or sections from a practice test, and the student should watch the clock from the beginning. Nick says that as a student moves further through the work, they become more aware of time and their pacing improves. Pacing is not a separate skill to add later.

A normal week has one shape. The student does a timed section. The student checks right against wrong. The student returns to every miss for a second try. The student reads solutions for what is left, closes them, and tries a third time. Three or four stubborn problems go to the next session.

Budget the time honestly. However long the timed section runs, the review should run about that long again. Nick is direct that steps two and three should take just as long as the section itself. A homework block that ends the moment the answers are scored was half a homework block.

Over months, this routine is doing all three things the research likes. It is practice testing. It is spread out. It is mixed. A parent does not need to grade anything to see whether it is running. The giveaway is whether the second half of the time ever happens.

Why does the first practice test score matter less than the homework habit?

A starting score sets the room for growth, which is why our published results are split at 27 or below on the ACT and 1350 or below on the SAT. What it does not predict is the gain. The gain comes from what happens in the weeks after that first score.

Every practice test is feedback. Nick describes it as a self-test that tells a student what they need to work on, what they do not, and whether they are improving. If the work being put in is not helping, something has to change. The test is a readout, not a verdict.

Homework works the same way on a smaller scale. Each section is a self-test. The key, Nick says, is that once you determine what you do not know, you correct that stuff. A wrong answer that never gets corrected is just a wrong answer repeated later on test day.

So a low first score is information, not a ceiling. Our ACT cohort of 491 students from 2017 to 2023 and our SAT cohort of 291 students in 2024 all started somewhere. What they shared was not a starting point. It was 12 or more instruction hours and at least 80 percent homework completion.

What should a parent do this month?

Start with a free mock SAT and a free mock ACT so your child has a real starting score and a real section to review. You can set that up at prepgurus.com/take-the-next-step or by phone at 949.229.5359. We have been doing this since 2009, and 782 students sit behind our published results.

At home, the one change that helps most is protecting the second half of homework time. If your child finishes a section and the review never happens, the method is not running. You do not need to know the content. You only need to notice whether the review time exists on the calendar.

Second, let the hard problems stay hard for a while. Nick's point about desirable difficulties applies at the kitchen table. A child who works a problem three times and still misses it has done exactly what we asked. That problem now has a job to do in the next session.

Last, think in months. Cepeda and colleagues (2006) found that the best gap between study sessions grows as the gap to the real test grows. Starting early gives the spacing room to work. Starting late takes that option off the table and leaves cramming, which the evidence treats as the weaker choice.

Questions parents ask

How much did your students gain on average?
ACT students starting at 27 or below gained an average of 5.7 points, and SAT students starting at 1350 or below gained an average of 190 points.

How many of your SAT students gained 200 points or more?
Among SAT students who started at 1350 or below, 63% gained 200 points or more.

What counts as completing the program?
Our results page counts a student as having completed the program with 12 or more instruction hours and at least 80 percent homework completion.

What are the three steps of homework?
Step one is doing all the problems and checking right answers against wrong answers. Step two is going back to the ones you got wrong and trying them again, armed only with the knowledge that you did not get them right the first time. Step three comes after that.

Does self-testing really beat rereading?
Roediger and Karpicke (2006) found that repeated study beat repeated testing after 5 minutes, but repeated testing beat repeated study at 2 days and 1 week.

Which study methods does the research rate highest?
Dunlosky and colleagues (2013) gave a high utility rating to only two techniques: practice testing and distributed practice.

Why is mixing problem types better than drilling one topic?
Mixing helps because Dunlosky and colleagues (2013) rated interleaved practice as a moderate utility technique, while drilling one topic in a block is not on the high list at all.

When should we start if we want spacing to help?
Short sessions repeated over months beat one heroic push in the final two weeks.

Method and sources: The score figures come from Test Prep Gurus' published results page as of October 2026, covering 491 ACT students from 2017 to 2023 and 291 SAT students in 2024, for 782 students in total. Those figures describe only students who completed the program, which that page counts as 12 or more instruction hours and at least 80 percent homework completion. The research figures come from three published papers: Roediger and Karpicke (2006), Dunlosky and colleagues (2013), and Cepeda and colleagues (2006). A known gap: the exact recall percentages in the Roediger and Karpicke experiments are not established from the available primary-source text, and the Cepeda review reports no single overall effect size in its abstract, so neither is printed here. Rows are ordered as the source pages print them, with our results first and the research tables after.