SAT and ACT Test Anxiety: What the Research Says Parents Should Try

As of October 2026: Two studies point to simple moves. In Brooks (2014), people who said "I am excited" spoke 167 seconds versus 132 seconds for those who said "I am calm." In Ramirez and Beilock (2011), the writing group's accuracy rose 5% from pretest to posttest while the control group's accuracy fell 12%.

What the Brooks (2014) excitement studies and the Ramirez and Beilock (2011) writing studies say about helping a nervous test taker.

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

Nick Standlea on this question -- Nick Standlea, Test Prep Gurus (YouTube).

Saying "I am excited" beat saying "I am calm" in Brooks (2014) In the speech experiment reported by Brooks (2014), 140 participants prepared and gave a recorded persuasive speech. Those told to say "I am excited" spoke for a mean of 167 seconds. Those told to say "I am calm" spoke for a mean of 132 seconds. Independent judges rated the excited group as more persuasive, competent, persistent, and confident. The same paper reports better karaoke singing and higher math scores for the excitement groups, though the measures differed by task. The karaoke result comes to us through a summary page, and that summary gives no number for the size of the gain.

Writing about test worries moved math accuracy by a wide margin in Ramirez and Beilock (2011) The lab study in Ramirez and Beilock (2011) used 20 college students, 10 per condition, before a high-pressure math test. The writing group's accuracy rose 5% from pretest to posttest. The control group's accuracy fell 12%. The posttest gap favored expressive writing, with t(18) = 5.55, P < 0.01, and Cohen's d = 2.48. In their classroom study of 106 ninth-graders, test-anxious students who wrote about their feelings scored about 6 percentage points higher on a biology final.

The 2004 choking paper ties pressure to working memory, and practice to protection A 2004 paper on choking under pressure reports three experiments on math problem solving. Pressure hurt unpracticed problems with heavy working-memory demands. Problems practiced 50 times each were not hurt. We name one limit up front. The pack we drew from flags the 2004 citation details as inconsistent, since the abstract points to another article's journal details, so we have not treated the author and journal line as verified. We quote the direction of the finding and the 50-times figure, and we skip sample sizes, which the available abstract does not state.

What do the two anxiety studies actually report?

These are the findings we cite throughout this page, taken from the Brooks (2014) paper and the Beilock line of research. Where a study does not report a number, the row says so.

Study	Task or sample	Reported result
Brooks (2014), overall finding	Singing, public speaking, and math tasks	Reappraising anxiety as excitement improved performance across singing, public speaking, and math tasks.
Brooks (2014), karaoke	Participants randomly assigned to say "I am anxious," "I am excited," or nothing	The "I am excited" group performed better on the karaoke game's objective singing measure; no numeric effect estimate is given in the available excerpt.
Brooks (2014), speech sample	140 participants (63 men and 77 women) prepared a persuasive public speech	Speech was recorded and judged.
Brooks (2014), speech length	"I am excited" vs. "I am calm"	Mean duration 167 seconds vs. 132 seconds, respectively.
Brooks (2014), speech ratings	Independent judges rated the recorded speeches	The excited group was rated more persuasive, competent, persistent, and confident.
Brooks (2014), math	Prompts to "Try to remain calm" or "Try to get excited"	Significantly higher math scores for excitement reappraisal in two experiments.
2004 choking paper (citation details not verified)	3 experiments on math problem solving under pressure	Infrequently practiced high-demand problems were still performed poorly under pressure; problems practiced 50 times each were not. Sample sizes are not stated in the available abstract.
Beilock and DeCaro (2007)	2 experiments on working memory and strategy choice	Under pressure, higher working-memory individuals performed optimally by using the simpler strategies lower working-memory individuals employed. No numeric effect size in the available abstract.
Ramirez and Beilock (2011), lab study	N = 20 college students, 10 per condition, high-pressure math test	Writing group accuracy rose 5% from pretest to posttest; control group accuracy fell 12%.
Ramirez and Beilock (2011), lab posttest	Expressive writing vs. control at posttest	t(18) = 5.55, P < 0.01, Cohen's d = 2.48.
Ramirez and Beilock (2011), classroom study	N = 106 ninth-graders writing before a biology final	About a 6-percentage-point advantage for test-anxious students who wrote about exam-related feelings.

Brooks, A. W. (2014), Journal of Experimental Psychology: General, 143(3), 1144-1158, DOI: 10.1037/a0035325. Ramirez, G., and Beilock, S. L. (2011), Science, 331(6014), 211-213, DOI: 10.1126/science.1199427. Beilock, S. L., and DeCaro, M. S. (2007), Journal of Experimental Psychology: Learning, Memory, and Cognition, 33(6), 983-998, DOI: 10.1037/0278-7393.33.6.983. For the 2004 choking paper, the citation details in the available search results are inconsistent, and we have not treated them as verified.

What do our own published results show?

These are our published results for students who completed the program. They are averages, not promises, and they are not a test of the anxiety methods on this 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.

How do I help my child with SAT or ACT test anxiety?

Start with two moves the research backs: have your child name the feeling as excitement, and have your child write about test worries before the exam. Brooks (2014) found that people who reappraised anxious arousal as excitement felt more excited and performed better. Ramirez and Beilock (2011) found that writing about test worries raised scores for anxious students.

Neither move takes long. The excitement move is a short self-statement. The writing move is a few minutes with paper before the test starts. Both are free, and both can be practiced at home.

The third move is slower and matters most. A 2004 paper on choking under pressure found that pressure hurt work on unpracticed problems with heavy working-memory demands. Problems practiced 50 times each were not hurt. We name that paper with care, because the citation details in the available search results are inconsistent and we have not treated them as verified. In plain terms, deep practice takes the load off a mind under stress.

We are not saying nerves cause every low score. Some scores are low because content is missing. But when a student knows the math and still blanks, these three moves are where we start.

None of these studies tested the SAT or the ACT. So read them as direction, not as a point forecast. The rest of this page reads each one in turn and says what a family can do with it.

What is the "I am excited" technique, and does it work?

It is a short self-statement said out loud before a stressful task, and Brooks (2014) reports that it beat trying to calm down. The paper's summary line is direct: compared with those who attempt to calm down, individuals who reappraise their anxious arousal as excitement feel more excited and perform better.

The logic is simple. A racing heart before a test feels the same as a racing heart before something fun. The body does not sort the two. The label the mind puts on the feeling is the part a student can change.

Trying to calm down asks for a big switch. The body is revved up, and the student is asking it to shut off. That rarely works in the minutes before a proctor reads directions. Moving from nervous to excited is a smaller step in the same direction.

Brooks (2014) tested this three ways. In the karaoke task, people were assigned at random to say "I am anxious," "I am excited," or nothing. A summary of the research says the excited group did better on the game's singing measure, and it gives no number for that gain. In the speech task, the excited group spoke longer and was rated better. In math, people read prompts to "Try to remain calm" or "Try to get excited," and the excitement group scored higher in two experiments.

For a test day, we teach students to say it out loud in the car, not in their head. Said out loud, it lands.

How big was the effect in the speech experiment?

In the speech experiment reported by Brooks (2014), participants in the "I am excited" condition spoke for a mean of 167 seconds, while those in the "I am calm" condition spoke for a mean of 132 seconds. The sample was 140 participants, 63 men and 77 women, who prepared and delivered a recorded persuasive speech.

Speaking time is not a score, so read it as a signal. A nervous speaker cuts the speech short and sits down. A speaker who stays in the task keeps going. The excited group stayed in the task longer.

Judges backed that up. Independent judges rated the excited group as more persuasive, competent, persistent, and confident. The press write-up of the research says the subjects who said they were excited gave longer speeches and were more persuasive, competent and relaxed.

Persistence is the part that maps onto the SAT and the ACT. A hard reading passage tempts a student to skim and move on. A hard geometry item tempts a guess. Staying in the problem for another twenty seconds is often the whole difference.

We do not claim the speech result predicts a point gain on a test. Brooks (2014) measured different outcomes in different tasks. What carries over is the direction: labeling nerves as excitement kept people working.

Does writing about test worries really raise scores?

In the lab study by Ramirez and Beilock (2011), the writing group's accuracy rose 5% from pretest to posttest while the control group's accuracy fell 12%. The study used 20 college students, 10 per condition, before a high-pressure math test.

The gap at posttest was large and clear. The paper reports t(18) = 5.55, P < 0.01, and Cohen's d = 2.48. In plain terms, the two groups barely overlapped. Control participants choked under pressure, showing a 12% accuracy drop from pretest to posttest.

Twenty students is a small study, and we say that plainly. One lab result with 10 students per group is a starting point, not a law. That is why the classroom study in the same paper matters more for families.

The classroom study followed 106 ninth-graders writing before a biology final. Among test-anxious students, those who wrote about their feelings scored about 6 percentage points higher than students given an unrelated writing exercise. The title of the paper states the claim: writing about testing worries boosts exam performance in the classroom.

Note the group that gained. The benefit showed up among test-anxious students. A calm student may get nothing from the exercise. That is fine. The move costs a few minutes and a sheet of paper.

What should my child actually write before the test?

Write about the feelings and worries tied to the exam, which is what students did in Ramirez and Beilock (2011) before the biology final. The comparison group in that study wrote about something unrelated, and the test-anxious students in that group did worse.

So the content matters. A journal entry about the weekend is not the same exercise. The writing needs to name the fear in the student's own words.

We keep it loose. Ten minutes, full sentences, no one reads it. Students write what they are afraid will happen, why the score feels heavy, and who they think is watching. Spelling does not count.

A pattern shows up fast. Most students write some version of the same three worries: running out of time, blanking on a topic they studied, and letting a parent down. Seeing those three on paper shrinks them. They stop circling in the background during section two.

Try it on a practice test first, not on test day only. A new ritual on the morning of the real exam adds one more unknown. Used during two or three timed practice tests, it becomes part of the routine.

If your child refuses to write, do not force it. Talking it through in the car does some of the same work. The study tested writing, so we say writing has the evidence behind it.

Why does my child blank on math even though she knows it?

The 2004 choking paper offers the most useful answer: pressure harmed performance on unpracticed problems with heavy working-memory demands. Working memory is the mental desk space where a student holds steps while solving. Stress fills that desk with worry, and the steps fall off.

That is why a student can solve a problem at the kitchen table and freeze on the same problem in a test room. The knowledge did not leave. The space to use it shrank.

That paper reports three experiments. Its key line is clear: infrequently practiced high-demand problems were still performed poorly under pressure, whereas problems practiced 50 times each were not. Sample sizes are not stated in the abstract text available to us, so we quote the direction and not a size. We also flag that the citation details for this 2004 paper are inconsistent in the search results we have, so we do not pin it to one author line.

Beilock and DeCaro (2007) adds a wrinkle across two experiments. Under pressure, higher working-memory individuals performed optimally by using the simpler strategies lower working-memory individuals employed. Strong students often carry heavy methods in their heads. Under stress, the lighter method wins.

That is a coaching point we use often. On test day, pick the plainer route. Plug in numbers. Write every step on the page instead of in your head. A student who offloads work onto paper is protecting the desk space that pressure is trying to take.

Can practice alone fix test anxiety?

Practice does not erase nerves, but the 2004 choking paper found that problems practiced 50 times each were not hurt by pressure. That is the strongest case for volume we know of in this research.

The mechanism is retrieval. When a problem type has been practiced to the point where the answer comes back without a search, pressure has less to disrupt. The student is pulling, not building.

This is why we talk about reps and not tips. A student who has done one practice section is working from scratch under a clock. A student who has done many is recognizing shapes.

Here is the exception. Practice without feedback can harden a bad method. Doing the same wrong step 50 times is still 50 reps of the wrong step. Reps need review attached.

We should also be honest about the limits of that 2004 finding. The study used math problems in a lab, not a full college entrance exam. Its citation details are not verified in our sources. We treat it as a good guide, not a guarantee.

So the honest answer is both. Practice lowers the load. The excitement line and the writing exercise handle what is left on the morning of the test.

Is it better to tell my child to calm down or to get excited?

Brooks (2014) tested exactly that choice, and excitement won on the measures reported. In the speech task, the "I am excited" group spoke 167 seconds versus 132 seconds for the "I am calm" group. In math, participants who reappraised their anxiety as excitement scored significantly higher.

This changes what a parent says in the car. "Calm down" asks a child to fight the body. "This is going to be a good one, you are ready, let's go" moves the same energy somewhere useful.

Tone matters more than the script. A parent who is tense while saying calming words sends the tense signal. Children read the face, not the sentence.

We also keep the morning talk off the score. Talking about a target number right before the test puts the stakes back on the desk. Talk about the plan instead: pacing, which section comes first, when to skip.

One caution. Brooks (2014) studied karaoke, speeches, and math tasks in a lab, not the SAT or the ACT in a real test center. The direction is useful. The size of the gain in a test room is not something the paper reports, and we will not invent one.

If your child hates the phrase, pick another one with the same meaning. The label is the active part, not the exact words.

How early should we start working on test nerves?

Start at the first full practice test, because the 2004 choking paper suggests the protection comes from practice done long before the test. Nerves built over months do not drop in a week.

A good order helps. First, run a full-length timed practice test to see the real starting point. Second, build content and reps. Third, add the test-day rituals, which are the writing exercise and the excitement line.

Adding the rituals last keeps them from becoming a crutch. A student who uses them to cover a content hole will be let down. A student who uses them on top of solid practice gets the benefit the studies describe.

Timing in the school year matters too. Reps take weeks, not days. A plan that leaves room for review is calmer than a plan built on the last free weekend.

For most families, that means starting in the fall or winter before the spring test dates. It also means planning for two sittings. A second test with a known routine is almost always calmer than a first.

One more thing. If anxiety is showing up outside of testing, in sleep or school or friendships, that is bigger than a test plan, and a licensed professional is the right support.

What can I do on the morning of the test?

Keep it short: ten minutes of writing about worries, one spoken line of "I am excited," and no score talk. Those are the two moves from Ramirez and Beilock (2011) and Brooks (2014) that fit into a test morning.

The writing comes first, at home or in the car. Ramirez and Beilock (2011) had students write just before the exam, and among test-anxious ninth-graders the advantage was about 6 percentage points on a biology final.

The spoken line comes last, right before your child walks in. In Brooks (2014), participants made the statement just before performing. Keep it out loud and keep it brief.

Everything else should be boring. Same breakfast as a practice test morning. Same bag, packed the night before. Same arrival window. Each new variable is one more thing for a nervous mind to track.

What we skip: last-minute drilling, a review of missed problems, and any sentence about colleges. Beilock and DeCaro (2007) found that under pressure the simpler strategy served students well, and a cram session right before the test adds pressure without adding skill.

After the test, ask about the experience, not the score. "How did it feel" is a question a tired student can answer. "How do you think you did" is a guess that will be wrong either way.

What do your own published results say, and what do they not say?

Our published results cover 782 students, and they are score gains, not a test of the methods on this page. Across 491 ACT students from 2017 to 2023, students starting at 27 or below gained 5.7 points on average. Across 291 SAT students in 2024, students starting at 1350 or below gained 190 points on average, and 63% gained 200 points or more.

Read those rows for what they are. They are averages for students who completed our program. They do not split out the effect of nerves, the effect of content work, or the effect of any one ritual.

We cannot tell you how much of a gain came from calmer test mornings. Our data does not track that. No study on this page tracks it either. So we report the gains and the research side by side, and we do not tie one to the other.

What the rows do show is range. Students who start lower have more room, and the rows are drawn from students at or below a set starting score. A student who starts above those lines is not described by these numbers.

We have run this program since 2009. The honest frame for a parent is this: a plan with reps, review, and a steady test-day routine is a reasonable plan. A single number from our table is not a promise for one child.

What do these studies not tell us?

They do not tell us how many SAT or ACT points these moves are good for, because no study in this pack measured the SAT or the ACT. Brooks (2014) measured karaoke, a recorded speech, and timed math tasks. Ramirez and Beilock (2011) measured a lab math test and a ninth-grade biology final.

Sample sizes are another limit. The lab study in Ramirez and Beilock (2011) used 20 college students, 10 per condition. The classroom study used 106 ninth-graders. These are small groups by the standards of education research.

Some numbers simply are not reported. For the karaoke task, the available excerpt of Brooks (2014) gives no numeric effect estimate, and the result reaches us through a summary page rather than a full data table. For the 2004 choking paper and for Beilock and DeCaro (2007), sample sizes and effect sizes are not stated in the abstract text available to us. We have left those gaps open rather than filling them.

There is also a citation problem we will name again. The 2004 citation metadata in the available search results is inconsistent, with the abstract pointing to a different article's journal details. We have not treated those details as verified, which is why we call it the 2004 choking paper in the prose.

Our own results carry their own limit. They are averages for students who completed the program, across 782 students total. They are not a forecast for any one child, and they do not measure the anxiety methods described here.

What is the move this month?

Run one full-length timed practice test, then add the two rituals on the next one. That order gives you a real baseline before you change anything about test day.

We offer a free mock SAT and a free mock ACT. A full-length, timed, proctored test tells you two things at once: where the content gaps are, and how your child handles the clock. Most families find the second answer more useful than the first.

On the second practice test, add the writing exercise before the start and the spoken line right before the first section. Then compare. One student is not a study, but you will see the pattern across a few sittings.

If the scores hold steady and the stress stays high, the problem is probably content, not nerves. If the scores jump between a calm practice test and a tense real one, the problem is probably the test room.

Nick Standlea founded Test Prep Gurus in 2009 after working as a Research Associate at the Quality of Life Research Center at Claremont Graduate University under Mihaly Csikszentmihalyi. The two short talks at the top of this page cover the same ground in his own words.

To set up the free mock test, call 949.229.5359 or go to https://www.prepgurus.com/take-the-next-step.

Questions parents ask

Does saying "I am excited" really help more than calming down?
Brooks (2014) found that people who reappraised anxious arousal as excitement felt more excited and performed better.

How long did the excited group speak compared with the calm group?
In the speech experiment reported by Brooks (2014), participants in the "I am excited" condition spoke for a mean of 167 seconds, while those in the "I am calm" condition spoke for a mean of 132 seconds.

How much did writing about worries change math accuracy?
In the lab study by Ramirez and Beilock (2011), the writing group's accuracy rose 5% from pretest to posttest while the control group's accuracy fell 12%.

Did the writing exercise work in a real classroom?
The classroom study followed 106 ninth-graders writing before a biology final.

Why does my child blank on math she already knows?
The 2004 choking paper offers the most useful answer: pressure harmed performance on unpracticed problems with heavy working-memory demands.

Can practice by itself protect against pressure?
Practice does not erase nerves, but the 2004 choking paper found that problems practiced 50 times each were not hurt by pressure.

What score gains do your students see?
Our published results cover 782 students, and they are score gains, not a test of the methods on this page.

What are the limits of this research for the SAT and ACT?
They do not tell us how many SAT or ACT points these moves are good for, because no study in this pack measured the SAT or the ACT.

Method and sources: The research on this page comes from two lines of published work: Brooks (2014) in Journal of Experimental Psychology: General, and the Beilock line of research from 2004, 2007, and 2011, including Ramirez and Beilock (2011) in Science. The populations were college students and, in one classroom study, 106 ninth-graders taking a biology final; none of the studies tested the SAT or the ACT. We print only figures the primary sources state, so several rows note that sample sizes and effect sizes are not reported in the abstract text available to us, and the karaoke result carries no numeric effect estimate and reaches us through a summary page. We also note a gap we did not fill: the 2004 citation metadata in the available search results is inconsistent, and we have not treated those details as verified, so we call it the 2004 choking paper rather than tying it to one author line. Our own results rows are averages for students who completed our program, 782 students in total, published at prepgurus.com; they are not adjusted for starting score or hours of practice, and they do not measure the anxiety methods described here.