Lies, Damned Lies and Sports Participation Statistics?

Sport England has recently published summary findings for sports participation in 2009/10 from the Active People Survey under the headline “Cycling and running boom shows appetite for sports participation”, with a growth in netball also highlighted. But what do these statistics really show, both for sports participation and for the London 2012 sport participation legacy effort?

——–—-———————————–

UPDATE (1st January 2011): When Mark Twain popularised the phrase “Lies, Damned Lies and Statistics”, his meaning was that statistics are often selectively used to bolster arguments, and that statistics that do not support one’s position are often discounted or disparaged.  This Blogpost was an invitation to consider whether this is the case with sports participation statistics (hence the question mark in the title).  My conclusion in the final paragraph below is that the selection of the statistic should depend on what policy outcome is valued and thus what question is important, not on what arguments one wishes to make.  In the cycling example, the long term participation trend is up, although in the last year participation has dropped.  If one wishes to examine whether any recent policy changes have had an immediate effect, the latter statistic is important; but if one wishes to examine a longer term trend, the former statistic is important (and it would take another year of falling participation to suggest a longer term downward trend).

The intention of this Blogpost is neither to make a particular argument through selecting a particular statistic, nor is it to critique the particular statistic Sport England prioritises.  It is intended to illustrate that the statistic that should be used depends on what policy outcome is valued, and therefore what question is important.  In this Sport England have been entirely consistent for at least three years – their headline statistic is the progress towards a million more people participating three times a week or more for 30 minutes or more at a time (3×30/week) by 2012/13 from a 2007/8 baseline.  In this respect there is no obfuscation and no selective use of statistics to bolster a particular argument, and for this Sport England are to be commended.  However, others may value a different policy outcome, and may therefore quite legitimately make a different statistical choice.

———————————————–

1) Pick of the Stats: Three different levels of participation are measured, once a month, once a week, and those participating three times a week or more for 30 minutes or more at a time (3×30/week).  It is this latter statistic that is used to measure the “one million target” – the Olympic legacy promise to get one million more people playing regular sport by 2012/13.  We can also choose from three “baselines” against which to compare this year’s participation: that from the first year of the survey, 2005/6, which gives a longer term picture; that from 2008/9 which tells us about the most recent changes; or that from two years ago, 2007/8, which provides the baseline for the “one million target” (there was no survey for 2006/7).  This means that we can choose from nine different statistics for sport participation change, and for individual sports, for which 3×30/week data has not yet been provided, there are six different participation change statistics to choose from.

2) Overall Participation: Progress towards the “one million target” for overall sport participation at 3×30/week has advanced 123,000 since 2007/8, a two year trend which, if it continued, would see the one million target being reached more than ten years late in 2023/24.  Of course, one argument is that the strategic investments towards the one million target need time to bed in.  In which case, we might expect to see a greater increase in participation in the last year as at least some of those investments take effect.  After all, we have known that London is to host the 2012 Games for five and a half years now.  However, the increase in overall participation at 3×30/week in the last year is only 8,000, a figure so small (0.1%) that it is within the margins of error of the survey, and therefore should be taken to mean no change at all.

 3) Netball: Overall, the number of women participating in sport has fallen, so the increase in those playing netball at least once a week by one fifth during the last two years is presented as a success that can help arrest this trend.  However, during the same period, those participating once a month or more has not changed (the small change reported is within the margins of error of the survey).  This means that the increase in women playing netball once a week is almost entirely drawn from those who were already playing once a month or more.  In other words, these are not new participants, but current participants who have increased their participation frequency.

4) Cycling and Athletics:  Unlike netball, cycling and athletics are both Olympic and Paralympic sports, and so a greater 2012-effect on participation might be expected.  In fact, looking at the longest term change available, those participating in athletics at least once a month has increased by over 700,000 since 2005/6, and in cycling by over half a million in the same period.  However, if we are looking for a link to strategic Olympic and Paralympic legacy investments, then the latest changes are most relevant.  In the last year, the increase in those participating in athletics at least once a month has been only 89,000, a figure that is almost entirely offset by a decrease in the number cycling at least once a month by 80,000.  In short, the net contribution of athletics and cycling to increases in those participating in sport at least once a month during the last year has been virtually zero.

5) The Olympic and Paralympic Games: So what does this say about the London 2012 Mass Participation Legacy effort.  In short, in the three sports that are highlighted as successes by Sport England, we have seen an increase in the frequency of participation among those already playing netball in the last two years, and in the last year we have seen a shift of participation from cycling to athletics that has an overall effect on monthly participation figures of all but zero.  Research on sport participation legacies by the Centre for Sport, Physical Education & Activity Research (SPEAR) shows that a ‘demonstration effect’, where people are inspired by elite sport, sports people and sports events to participate themselves, can result in three outcomes: those who are already playing may play more, those who have played before may play again, and people may give up one sport to try another.  The former and the latter are reflected in the recent participation statistics, and unless the recent London 2012 Mass Participation Legacy Plan is amended or extended to include measures that seek to raise demand for sport (rather than simply increasing supply), we are some way from delivering a participation legacy from the 2012 Games. (click HERE for an evidence-based analysis of the Mass Participation Legacy Plan)

In summary, and using cycling as an example, depending on the statistic chosen, participation has either increased by over half a million or fallen by 80,000!  The key question is what’s important?  Over the long-term (since 2005/6) cycling participation has risen significantly, but in the last two years, the rate of that rise has slowed, and in the last year participation has actually fallen.  You pays your money, and you takes your statistical choice!

About these ads