Exam Stress: What Works and What Does Not

She revised for three weeks. She knew it at the kitchen table the night before. She walked into the hall, read the first question, and everything went.

For the wider school-age picture — homework, ADHD in teenage boys, classroom support — start at the school-age hub.

Or he cannot sleep for the week before. Or the stomach aches start each morning of exam week and stop the day it ends. Or she is fine until she turns the paper over and her hands will not work.

This is not laziness, and it is not a lack of preparation. It is a specific, heavily researched phenomenon with a name, a measurable size, and treatments that have been tested — including a very popular one that turns out not to work.

What test anxiety actually is

It is not simply "feeling nervous". The scales used to measure it in children separate three distinct components, and they need different responses:

• Thoughts — "I will fail", "everyone else finds this easy", "my parents will be ashamed", and the loop that runs during the paper instead of the answers.
• Autonomic reactions — racing heart, sweating, nausea, needing the toilet, shaking hands, the sensation of the mind going blank.
• Off-task behaviours — fidgeting, looking around, staring at one question, checking the clock, everything the body does instead of writing.

A child can be heavy on one component and light on the others. The child who vomits before an exam and the child who sits quietly and thinks catastrophic thoughts for two hours both have test anxiety, and they will need different work.

How big the problem is — measured

The most complete picture comes from a 20-year systematic review and meta-analysis in the Journal of School Psychology covering test anxiety in primary school children aged 5 to 12. It identified 76 studies, 85 independent samples, and 53,617 children.

What it found test anxiety is related to:

• Academic achievement — negatively related in both mathematics and literacy, with correlations of around 0.20 in magnitude
• Academic self-concept: r = −0.41 — how capable a child believes they are
• Self-efficacy: r = −0.39
• General anxiety: r = 0.62
• Social anxiety: r = 0.57
• Depression: r = 0.45

Three of those numbers deserve to be pulled out.

First: the correlations with general anxiety, social anxiety and depression are far larger than the correlation with achievement. Test anxiety is much more strongly an anxiety problem than an academic one. A child who is anxious in the exam hall is usually an anxious child, and treating the exam alone will not reach it.

Second: academic self-concept at −0.41. The belief "I am not good at this" travels with test anxiety more closely than actual marks do. Children who are performing adequately can still be carrying the self-concept of a failing child — and that belief is what determines whether they keep trying.

Third: it was measured in five-year-olds. This does not begin at matric or O-levels. By the time the board exams arrive, most children with test anxiety have had it for years.

Two findings specific to families like yours

The same meta-analysis reported two moderators that matter here more than they would in London or Toronto.

Test anxiety was higher in Asian samples than in European and North American samples. The review does not settle why, and we are not going to invent a mechanism — but any parent who has watched a Pakistani exam season, with the result lists, the comparisons between cousins, and the weight placed on a single set of marks, will recognise the environment as a plausible candidate.

Gender differences were smaller in Asian samples. Overall, girls scored higher than boys (d = 0.21) — but that gap was more pronounced in North American samples than Asian ones. Practically: do not assume this is a girls' problem. In our part of the world the boys are carrying a similar load, often with less permission to say so.

One more, on age: the association between test anxiety and mathematics achievement was stronger in older children. It compounds. A child who begins avoiding maths at eight has a much larger gap at fourteen, and the anxiety and the gap then feed each other.

What does not work — the popular one

You will have seen the advice: get the child to write down their worries in the ten minutes before the exam. It has circulated widely and it sounds plausible.

A meta-analysis examined it properly: 15 documents comprising 21 studies, 1,457 participants, 30 effect sizes across anxiety and test performance.

The effect on anxiety was negligible and non-significant (r = −0.05, p = 0.57), and remained non-significant after removing an extreme outlier. The effect on performance was initially significant (r = 0.09) but lost significance once the same outlier was removed. There was very little evidence of publication bias — meaning this is not a case of negative results being hidden. The technique simply does not do what it is said to do.

The authors' conclusion is worth having in your pocket the next time someone recommends it: the literature fails to support single sessions of expressive writing as a treatment for test anxiety — though other treatments with better evidence are available.

What does work

Exercise — with specific parameters

A systematic review and meta-analysis of 15 controlled trials on exercise for test anxiety produced unusually practical guidance:

• Twenty minutes of aerobic exercise, at any intensity, two to three times a week, for at least four weeks reduced test anxiety.
• A single session of 10 to 15 minutes did not. Sending a child for a quick run on exam morning is not the intervention.
• Exercise combined with psychological therapy worked better than exercise alone.

The authors note this is the minimum effective dose, not a ceiling. For a household preparing for exams, the implication is uncomfortable and clear: the physical activity that gets cancelled first, to make room for revision, is one of the few things with trial evidence behind it.

Mindfulness-based approaches

A meta-analysis of 18 studies with 1,275 participants found mindfulness-based interventions produced a moderate-to-large reduction in test anxiety (effect size −0.716, 95% CI −1.383 to −0.049). Age, number of sessions and delivery mode did not significantly change the result.

Two honest caveats. That confidence interval is wide and its upper bound sits very close to zero. And the authors found evidence of publication bias on one test (Egger's regression, p = 0.025), which means the true effect may be smaller than the headline. Worth trying; not worth guaranteeing.

Anxiety treatment itself

The large primary-school meta-analysis concluded that intervention studies targeting anxiety reduction have been successful in reducing test anxiety and improving related outcomes. Given that test anxiety correlates 0.62 with general anxiety, this is the least surprising and most important finding on this page: treat the anxiety, and the exam problem improves. Our page on helping an anxious child covers what that looks like, and childhood anxiety covers assessment.

The revision itself is usually the wrong shape

Most children revise by re-reading and highlighting. There is good evidence that this is close to the worst use of the hours.

An analysis of learning techniques among 155 students, assessing retention up to three months, found that re-reading, highlighting and summarising were the most known and most used techniques — and had detrimental effects on academic outcomes. The most effective technique was metacognition — planning, monitoring and checking one's own understanding — and up to 92% of students did not know what it was.

What has better support is retrieval practice: closing the book and trying to produce the answer from memory. On how to schedule it, five experiments plus a meta-analysis of 19 studies found that testing spread through the learning — a small test after each block — beat one large test at the end, with a small-to-medium effect (g = 0.25).

Why this belongs on a page about anxiety: a child who has already answered questions from memory, twenty times, in a low-stakes setting, walks into the hall having done the thing before. Re-reading produces a feeling of familiarity that collapses the moment a blank page demands recall — which is precisely the "I knew it last night" experience.

What to do at home

1. Keep the exercise. Twenty minutes, two or three times a week, four weeks minimum. Do not cancel it for revision.
2. Change re-reading to self-testing. Cover the page. Say it out loud. Check. This is more uncomfortable and more effective, and children resist it precisely because it exposes what they do not know — which is the point.
3. Test little and often through the term, not once at the end.
4. Protect sleep in the last week. An all-night revision session trades a small amount of content for a large amount of recall — see helping your child sleep better.
5. Separate your disappointment from your love, out loud. Children carrying "my parents will be ashamed" into the hall are carrying it because someone said something once, probably years ago, probably not meaning it that way.
6. Stop the comparisons. Cousins, siblings, neighbours' children, the result list on the wall. Academic self-concept has the second-largest correlation on this page.
7. Do not offer money or large rewards for marks. It raises the stakes of the thing that is already too high-stakes.
8. Practise the conditions, not just the content. A past paper, timed, at a desk, in silence. Familiarity with the situation reduces the autonomic component more than more content does.
9. Have the "what if it goes badly" conversation before the exam, not after. A child who knows there is a life on the other side of a bad result is measurably freer in the hall.

What makes it worse

• "Just relax, you know it all." True, useless, and it tells the child their reaction is unreasonable.
• Increasing revision hours as the anxiety rises. Past a point, more hours produce more anxiety and no more learning.
• Discussing results in front of relatives.
• Framing one exam as deciding a life. It is said to motivate. It reliably does the opposite.
• Withdrawing sport, play and friends as punishment for poor marks — removing the things with actual evidence behind them.
• Treating physical symptoms as fabrication. The nausea and the racing heart are the autonomic component. They are real, they are measured on the scales, and calling them an excuse teaches a child to hide them.

When it is something else

Some children labelled as exam-anxious are struggling with something the exam merely exposes. Worth checking before you treat the anxiety alone:

• A specific learning difficulty. A child who reads slowly, or cannot organise written answers under time pressure, is not primarily anxious — see dyslexia, dysgraphia and dyscalculia.
• Attention difficulties. A child who loses the thread of a long paper, or misreads instructions repeatedly, may have an attention problem — see ADHD or autism.
• Vision or hearing. Simple, common, and missed for years.
• Anxiety that is not about exams at all — social anxiety at r = 0.57 means the room, the invigilator and the other children may matter as much as the paper.
• Depression. At r = 0.45, low mood is a serious candidate in a teenager whose motivation has gone, not just their nerve.

When to get help

• Physical symptoms outside exam periods — daily stomach aches, headaches, nausea
• School refusal, or refusing to attend on exam days — see school refusal
• Sleep that has not returned after the exams finished
• A child who has stopped trying — this often looks like laziness and is frequently the endpoint of a collapsed academic self-concept
• Panic attacks — sudden overwhelming fear with breathlessness, chest tightness or the feeling of dying
• Marks that have fallen sharply without an obvious reason
• Any talk of self-harm or of not wanting to be here. Exam seasons are a known period of risk. This is a same-day conversation with a doctor, not something to manage until results day.

Where we fit

We are not a tuition centre and we will not promise better marks.

What we do is the anxiety underneath — which the evidence says is the larger part of this — and the assessment that distinguishes an anxious child from one with an unrecognised learning difficulty, because those two look identical from the outside and need completely different help. Where the difficulty is genuinely in reading, writing or number, learning difficulties covers what assessment involves.

A consultation is Rs 1,500 and lasts up to 50 minutes, we work in Urdu or English, and you do not need a referral.

Sources

• Robson DA, Johnstone SJ, Putwain DW, Howard S. Test anxiety in primary school children: a 20-year systematic review and meta-analysis. Journal of School Psychology. PubMed 37253582
• O'Meara M, et al. Single-session expressive writing interventions for test anxiety: a meta-analysis. Anxiety, Stress, and Coping. PubMed 40968670
• Zhang Y, et al. Effects of exercise intervention on students' test anxiety: a systematic review with a meta-analysis. International Journal of Environmental Research and Public Health. PubMed 35682293
• Yılmazer E, et al. Effects of mindfulness on test anxiety: a meta-analysis. Frontiers in Psychology. PubMed 38993347
• Tümtürk MA, et al. A reliability generalization meta-analysis of Children's Test Anxiety Scale. Psychological Reports. PubMed 39962704
• Urrizola A, et al. Learning techniques that medical students use for long-term retention: a cross-sectional analysis. Medical Teacher. PubMed 36306371
• Don HJ, Boustani S, Yang C, Shanks DR. A grain of truth in the grain size effect: retrieval practice is more effective when interspersed during learning. Journal of Experimental Psychology: Learning, Memory, and Cognition. PubMed 39556402