The question comes up constantly, and it is poorly posed. “Training” can mean two very different things: improving your score on this kind of task, or improving the ability the task claims to measure. The first is easy. The second is not.

Improving your score: yes, clearly

The practice effect is among the best established findings in psychometrics. Retaking a test of the same type typically gains several points on the second attempt, fewer on the third, with a plateau arriving quickly.

The mechanism is unmysterious. On first encounter, a good part of your attention goes to side tasks: parsing the instructions, working out what is expected, managing the clock, getting used to the format. By the second attempt all of that is settled, and attention goes entirely to the problem.

It is why psychologists leave an interval — often six months to a year — before re-administering a test to the same person, and why serious protocols keep several equivalent forms.

Improving the ability: much more doubtful

Here the literature is unforgiving. Cognitive training programmes, brain-training apps in particular, produce well-documented gains on the trained task, modest gains on very similar tasks, and close to nothing on distant ones.

Put plainly: drill matrices and you will get better at matrices. There is little evidence that your capacity to reason in a new, untrained context benefits. Far transfer is exactly what these programmes promise and exactly what studies struggle to demonstrate.

That conclusion has been contested and the debate continues. But anyone promising a lasting, general increase in your capacities in return for a few weeks of exercises is claiming far more than the data support.

When training makes sense

In one specific case: when the score itself is what matters, rather than what it represents.

That is the situation with selection and aptitude tests — graduate schemes, civil service exams, transport, security, healthcare entry. The test is an obstacle to clear, formats are stable, and familiarity genuinely changes outcomes. Training there is not only legitimate but rational: it is preparation, like any exam.

The same applies if you are taking a supervised Mensa session: you are playing a threshold, and arriving familiar with the format is an advantage.

When it makes no sense

If you are about to have an assessment with a psychologist, training is counter-productive. The point of the assessment is a faithful picture of how you function, because decisions follow from that picture — an accommodation, a placement, an explanation for a difficulty. A score inflated by practice produces a false picture, and nobody gains. If you have recently taken many tests, simply say so: the practitioner will factor it in.

What actually helps on the day

Nothing dramatic, but the effect is real and immediate:

  • having slept;
  • taking the test in one sitting, without interruption;
  • not grinding on one question — on a timed paper, three minutes on a hard item usually costs more than it returns;
  • answering even when unsure, where nothing is deducted for a wrong answer, as is the case here;
  • choosing a moment when you are available rather than the end of a heavy day.

Those points are easily worth the points you would chase with weeks of drilling.