Maths anxiety is real and it costs marks: what the research says

Maths anxiety consumes the same working memory the problem needs, so an anxious child underperforms what they actually know — the effect is largest on problems that require carrying or holding steps in mind. It also transmits: in a year-long study, children of maths-anxious parents learned less maths, but only when those parents helped with homework frequently.

Anxiety runs in the same mental space the maths needs

Working memory is what holds a part-finished calculation while the next step is worked out. Anxiety competes for it. In the classic experiment, highly maths-anxious students made far more errors under a memory load, and worst of all on problems requiring a carry — precisely the problems that demand you hold something in mind.

This is why a child who demonstrably knows the method at the kitchen table blanks in the exam hall. It is not a character failure and it is not laziness; it is capacity being spent on worry instead of arithmetic. (Ashcraft & Kirk, 2001.)

Anxious parents transmit it — but only while helping

In a year-long study of first and second graders, children whose parents were maths-anxious learned less maths across the year and ended it more anxious themselves. The striking part is the condition: this held only where those parents helped with homework frequently. Where help was infrequent, there was no link at all.

The mechanism is modelling, not method. How you sound at the table outranks the technique you use. If maths makes you tense, the useful move is not to withdraw from your child's education — it is to keep your own commentary neutral, and to be honest that finding something hard is not the same as being bad at it. (Maloney et al., 2015 — an observational field study, so it shows association rather than proving cause.)

Mixing question types beats drilling one

Children practised four kinds of problem, either blocked by type or shuffled together. Shuffling made the practice session harder and lower-scoring — and then doubled scores on a test the following day.

Blocked practice teaches the steps. Mixed practice teaches which method the question is asking for, which is the skill a contest paper actually tests. The implication for a worksheet of twenty near-identical questions is not flattering. (Taylor & Rohrer, 2010.)

Let the errors happen

Learners recall a correct answer considerably better when they first attempted the problem and got it wrong. The struggle is not wasted time before the learning — it is the learning. The instinct to supply the method the moment a child stalls is a well-meant way of removing the most valuable part of the exercise. (Kornell, Hays & Bjork, 2009; reviewed in Metcalfe, 2017.)

Brain-training apps do not transfer

A major review found no broad improvement in cognition or academic achievement from practising cognitive skills stripped of subject content. Gains stay stuck to the trained task. What moves maths is time spent on maths. (Simons et al., 2016.)

What we can't know

Most of this research is not Singaporean and none of it was conducted on olympiad candidates. The parent-anxiety study is observational: it shows a relationship across a group, not a mechanism proven in any individual family. Early maths skill is the strongest school-entry predictor of later achievement (Duncan et al., 2007) — but prediction across large groups is not a verdict on one child, and it is not cause.

Treat all of this as a reason to change your tone and your practice structure, not as a diagnosis of your child.

Sources

• Maloney, Ramirez, Gunderson, Levine & Beilock — Psychological Science (2015)
• Ashcraft & Kirk — Journal of Experimental Psychology: General (2001)
• Taylor & Rohrer — Applied Cognitive Psychology (2010)
• Kornell, Hays & Bjork (2009); reviewed in Metcalfe — Annual Review of Psychology (2017)
• Simons et al. — Psychological Science in the Public Interest (2016)
• Duncan et al. — Developmental Psychology (2007)

Last verified: 2026-08-04