Wednesday, November 23, 2011

On drawing dubious conclusions from data (or, How to push your barrow along regardless)

This article in the Sydney Morning Herald on Monday set my teeth on edge, and it's taken me until today to pin down why.

The article is, basically, reporting the results of the NSW Schools Physical Activity and Nutrition Survey, but with the inevitable glosses and essentialist positions that seem to go hand in hand with any reporting on childrens' health or activity levels. It leads off with an absolute statement and doesn't improve from there:
MANY Australian children are too inactive, with more than half of the primary students and almost three-quarters of the high school students in a survey spending more than the recommended time each day in front of a TV, computer or other screen.
My editor's eye immediately caught the phrase "too inactive" and started unpicking it. What's too inactive? I thought. According to whom? Is the only measure how much screen time children are having, or are other factors considered too?

As the article continues, it becomes clearer what the study means by "too inactive" - it means a) more time in front of screens than 'experts' think is optimal b) an asserted decline in 'basic' physical skills such as jumping and catching a ball and c) a steady rate of consumption of fat-rich foods and sugary drinks (not an increase, it is noted in small words).

The article, to do it justice, does point out that the proportion of overweight and obese children is stable at 22.8%, having shown no increase since the 2004 study despite the widespread panic about a "rising" obesity crisis in young people. However, this is presented as the 'good' news in amongst a generally dismal set of results.

However, I'll be honest; it's not the actual data or even the conclusions that the article draws from it that I find problematic. I think aggregated data about diet and habits can be useful thing to gather in terms of planning public health interventions. I think messages about nutrition and giving growing bodies the best dietary support to do their growing, as well as initiatives to positively increase activity levels (such as Victoria's Go For Your Life program) are great, especially when they are done in a way that supports and encourages achieving the best health your body is capable of, rather than focusing on dubious measures of wellbeing such as weight.

No, the part of the article that really bugged me was the conclusions voiced by some of the people involved in the study. For instance:
Dr Hardy said it was ''appalling'' that fewer than 10 per cent of girls in year 4 and 6 could throw a ball overarm correctly.
REALLY? That's your measure of the decay of modern childhood - the ability to throw a ball overarm "correctly?"

The (weak) premise behind this is that if children lack physical skills, they are less likely to enjoy physical activity and / or get picked for team sports. If the lack could be demonstrated to be global, across ALL activity types, then perhaps there might be half a leg to stand on here. But it entirely fails to consider that some children - hell, some adults - may lack ball skills but still be perfectly physically competent, and active, in other ways. My own daughters, for instance, are somewhere below mediocre in ball sports, but are great swimmers, good gymnasts (and growing those skills all the time), and love to ride their bikes and scooters everywhere.

Is this study really asserting that THE MEASURE of physical wellbeing in children is whether or not they have the capacity to be good at a limited number of ball sports? I think, actually, that the assertion arises because of the data, in that the results showed a drop in ball skills, the commentators have an interest in showing an overall picture of failure and crisis, so they seize on that element and tut-tut about how shocking it is. I agree that helping kids to acquire good ball skills is a worthwhile goal, for parents or for PE teachers - it helps hand-eye co-ordination, and it's fun to play ball games! Do I think it a universal measure, though? NO.

I guess what I find the most disappointing in these stories is their assumption that a) there IS an obesity crisis, and it's getting WORSER and WORSERER b) kids these days. They need to get outside more! Screens are bad, eleventy!!! and c) parents are to blame. (Naturellement). I think what studies like the NSW one can do, at their best, is give a statistical picture that can help focus public health messages across the board, and also, in this case, refute some folk wisdom that's actually based on erronoeous assumptions (that childhood obesity is worsening. As this data shows, it isn't). It's a shame that instead of these actually useful outcomes, the focus always springs to demonising weight, parents, and children who use technology.

This post is part of NaBloPoMo. 23 down, 7 to go!

No comments:

Post a Comment