Firstly I’d like to not take anything away from the sad loss of life, which should be the main focus of the article and the issues its trying to highlight.
I’ve just used the calculations I normally use to look at the statistical significance of response rates to junk mail campaigns on these small sample sizes.
If you assume a 50:50 gender split in cyclists on the road. You can still, even with this small sample of 9 riders say with 99% statistical significance that the [unfortunate] result is true, that women are more likely to be killed by HGVs than men.
NB:It does assume that this sample of 9 is either random, or all HGV road deaths in a given period.
Also a higher proportion of male cyclist (Which I’m sure is the case) would only increase the significance of the results.
I don’t really want more deaths to improve the statistical significance, 9 is already 9 to many…
Firstly I’d like to not take anything away from the sad loss of life, which should be the main focus of the article and the issues its trying to highlight.
I’ve just used the calculations I normally use to look at the statistical significance of response rates to junk mail campaigns on these small sample sizes.
If you assume a 50:50 gender split in cyclists on the road. You can still, even with this small sample of 9 riders say with 99% statistical significance that the [unfortunate] result is true, that women are more likely to be killed by HGVs than men.
NB:It does assume that this sample of 9 is either random, or all HGV road deaths in a given period.
Also a higher proportion of male cyclist (Which I’m sure is the case) would only increase the significance of the results.
I don’t really want more deaths to improve the statistical significance, 9 is already 9 to many…