Lies, Damned Lies and Statistics

A friend of mine occasionally wears a t-shirt from the Royal Statistical Society and whilst this level of geekiness is beyond me, I do love the articles on statistics in School Leadership Today (mock me, go on!)

In the May issue, Stephen Gorard dissected Contextual Value Added data to show that there was no evidence that different types of school (Faith/Academy etc) achieved better or worse than others. He showed just how flimsy the “evidence” schools constantly measure themselves against really is.

In the most recent issue, Keith Savage delves into the recent reports on children and young people’s welfare. Whilst the headlines from the Children’s Society Report suggested that 7-10% of young people were unhappy or had low well being, Savage reminds us that the research is only about 10-15 year olds and that given the nature of those years, 90-93% being happy is really not that bad.

He then moves on to the 2010 DCSF report. This one tells us that 33% of 10-15 year olds were unhappy. Only, the thing is, this is apparently based on rather unusual measuring techniques: In order to be “happy” respondents to the survey had to agree firstly that “I have one or more good friends” and then that when they are worried they could to at least two out of “parent/friend/other adult”. As Savage argues “These might be important questions but to claim the responses are real and adequate measures of emotional health seems to be pushing it.”

So there you have it - things might not be so bad after all. (and then he goes on to “Idealised Youth “and it gets even better… but I’ll save that for another blog. In the mean time, see the discussion board and tell me what you think).

DCSF (2010), Local Authority Measures of National Indicators Supported by the Tellus4Survey 

Savage. K., (2010) How unhappy are our children, School Leadership Today, Volume 2.3

The Children’s Society (2010), Understanding children’s well-being: A National Survey of Young People’s Well Being