How Parkrun times reflect the state of the economy

2 watchers

Muttley Sep 2023 12:41pm, 3 Sep 2023 2,633 posts	0 Muttley "The over-50s run faster in recessions" theguardian.com What the Parkrun data tells us about fitness and health relative to the economy
richmac Sep 2023 1:27pm, 3 Sep 2023 23,180 posts	0 richmac Fascinating, Overlay it with other data & a clearer picture could emerge
rf_fozzy Sep 2023 1:34pm, 3 Sep 2023 20,808 posts	0 rf_fozzy Hmm. I suspect there isn't enough longitudinal data for that to be conclusive. How many recessions have we had? Particularly since the wider spread of parkruns. There could be confounding factors. Hasn't the "average" parkrun time got slower over time (of which parkrun HQ are bizzarely proud)? It's possible it's true, but as with all seemingly unconnected statistical correlations, I remain somewhat sceptical.
DrPhleecingD Sep 2023 1:37pm, 3 Sep 2023 54,778 posts	0 DrPhleecingD Spurious at best, you'd think if he was an economist he shouldn't go round making dubious statistical claims
Derby Tup Sep 2023 1:39pm, 3 Sep 2023 61,093 posts	0 Derby Tup You'd think as an economist he'd be bound to make spurious statistical claims
DrPhleecingD Sep 2023 1:45pm, 3 Sep 2023 54,779 posts	0 DrPhleecingD Lol
SlownOld Sep 2023 1:46pm, 3 Sep 2023 5 posts	0 SlownOld As Tim Harford regularly points out on ‘More or Less’; correlation does not mean causation.
rf_fozzy Sep 2023 1:47pm, 3 Sep 2023 20,809 posts	0 rf_fozzy It's not Torsten Bell's paper. Nor is it from the resolution foundation, who are usually pretty robust to be fair. If you follow the link in the article, you can read the original paper. Which is a "discussion" paper (non-peer reviewed) from the university of Bradford. The discussion is interesting - they've apparently weighted for increased participation over time, but not the change in "running demographic" over time (when parkruns started out they were dominated by 'regular' club runners - which is not the case now). This may or may not be important. I don't know if the linear regression models they use are standard (they say they are).
DrPhleecingD Sep 2023 1:47pm, 3 Sep 2023 54,780 posts	0 DrPhleecingD This is where I miss Twitter, there would be someone on there eviscerating this analysis
rf_fozzy Sep 2023 1:49pm, 3 Sep 2023 20,810 posts	0 rf_fozzy "As Tim Harford regularly points out on ‘More or Less’; correlation does not mean causation" Yes, but sometimes it does The possible correlation and the possible explanation are plausible. My question more whether the analysis is robust from plausible confounding factors.