Statistical Results Somehow Interesting, Guardian Finds

Guardian reporter Warwick Mansell lets us know about the results of a large-scale study on child development. Well, sort of. The title of the piece:
Children can fall behind as early as nine months
"Fall behind" suggests a movement from an earlier state of parity. But no earlier measurements are reported. The title makes sense if you assume that children must have started out all the same at birth. Which, of course, makes sense given that we know with certainty that nature plays no role whatsoever in human development. None whatsoever!

The article starts:
Children who do not reach key developmental milestones at just nine months old are far more likely to struggle at school, according to an important study published today.
Not a bad first sentence. But then:
The Millennium Cohort Study of nearly 15,000 children says that babies who were slow to develop their motor skills at nine months were significantly more likely to be identified as behind in their cognitive development, and also likely to be less well behaved at age five.
Ah, the S-word makes an early appearance. In a study of 15,000, pretty much any difference is going to be statistically significant. I'd like to know how big it is.
The correlation between performance at nine months and five years was said to be significant even after the researchers considered the impact of poverty on children's development.
"Impact" suggests causality.
The children's cognitive development was assessed at the age of five through a series of vocabulary, spatial reasoning and picture tests, and their results compared with those from separate assessments years earlier.
That sounds very much like an IQ test. If it is, you should call it that. In fact, I'm pretty sure it is, as the next sentence talks about "cognitive ability", which is one definition of intelligence, which is what an IQ test is supposed to measure. The sentence actually starts well...
Children who failed at nine months to reach four key milestones in gross motor development, relating to sitting unaided, crawling, standing and taking their first walking steps, were found to be five points behind on average in cognitive ability tests taken at age five, compared to those who passed the milestones.
Once you've stopped wondering about the wording, you'll wonder: How much is five points? What does it mean? The next sentence has the lowdown:
This equates to the difference between being in the middle of the ability range in the cognitive tests, and being below average.
I can't be sure, but being at the 49th percentile of the ability range would probably qualify as "being in the midde" in most people's interpretations of the concept "the middle". It is also below average. Should we conclude that 5 points equates to no difference whatsoever?

To mention something positive, though, the artcle has an interesting quote, from "[t]he Department of Health":
Children's health and wellbeing is a key priority for [the] government.
In this commenter's humble opinion, though, the reporter should have taken the time and called someone from the Tories. After all, the above quote would have more punch when contrasted with something along the lines of: "The conservative party, and I can't stress this enough, doesn't give a fuck about the children."

Ah, what can you expect? They're giving this stuff away for free on the web.

Related post: What statistics should be taught in school?

No comments: