Sunday, 20 September 2015

Physical activity and achievement in Maths – it’s scientific nonsense

An American educational website recently tweeted this:


An alarm bell that rang for me was, why only mention Maths? If this thin layer helps Maths then it should be useful for other things too. So I was suspicious enough to check the original research, a paper by Laura Chaddock-Heyman, published last month.

This was her methodology. Find 48 children, half with much better than average fitness levels, half much worse than average. Check that there are no other major differences between the children (e.g. age, social deprivation, cognitive ability) and then give them tests in reading, spelling and mathematics to see if their fitness level correlates with their ability in these areas.

She stated that on testing, in reading and spelling there was no difference. We were left with only mathematics.

The scattergraph produced by Chaddock-Heyman does indeed point to a thinner cortical layer in the fitter children and also a higher performance in Maths.  

Remaining dubious,  I then looked more closely at the methodology. As I said, her respectable scientific approach was to ensure that the children were the same in all possible ways except for their fitness. So, what if I dug deeper into the selection of the subjects themselves – were there any errors there? Sure enough, among the higher-fit participants there were 14 boys and 10 girls. Among the lower-fit participants, there were 8 boys and 16 girls.

In other words, when she measured their mathematical ability, she observed lower mathematical ability among a group that was two-thirds girls.

Why does this matter? Well, like it or not, there is something wrong in the US educational system with mathematics. Boys consistently out-perform girls in this subject, and have done for the last forty years

Many other countries have little or no gender differential. But Chaddock-Heyman’s research was done with American subjects, so it needed to take the known US mathematical inequality into account. When does the mathematical difference between boys and girls start to be apparent? At some point between elementary school and middle school.  And what was the age of these children that Laura Chaddock-Heyman examined? The final year of elementary school!

So this research had an imbalance of subjects who were already predisposed to be better at mathematics owing to the gender inequality in this subject in the USA. It was no surprise that the lower-fit group, being mainly composed of girls, performed less strongly in the maths test, and no surprise that the reading and the spelling was the same in each group, because there is not the same gender differential in these subjects.

As this research is flawed, it should be ignored. If we really want to know if there is a link between physical fitness and mathematical ability, then lower-fit boys must be tested against higher-fit boys, and the same for girls – or better still, the US first needs to get its educational act together and address the mathematical under-performance by girls.


But what’s the betting that this erroneous conclusion will be promoted in teacher-training colleges and repeated in staff rooms for years to come?

No comments:

Post a Comment