3 ms·
Unfortunately n, in this paper, doesn't seem to mean sample size like you would expect in statistics. It is an extrapolated number of concussions nationwide bas
by camiller 9y ago
Unfortunately n, in this paper, doesn't seem to mean sample size like you would expect in statistics. It is an extrapolated number of concussions nationwide based on a study of 100 high schools.
- kelnos 9y agoThat still doesn't mean you can just add the percentages. You cannot determine the percent of boys+girls soccer head injuries without knowing the sample sizes. Or, if you can figure out the relative difference between the number of boys and girls who play soccer, you can do a weighted average of the percentages. Either way, adding the percentages is completely incorrect. EDIT: http://www.usyouthsoccer.org/media_kit/keystatistics/ http://www.usyouthsoccer.org/media_kit/keystatistics/ suggests that the split is roughly equal -- in 2008 it was 52% boys and 48% girls. With that said, the correct weighted value is more like 21.5x.48 + 15.4x.52 = 18.3%, which is significantly less than the 40.5% you quoted for football.
- camiller 9y agoper the source material there were a total (extrapolated) of 135901 concussions across the 9 sports studied. 55007 for football - 40.5% of 135901 20929 for boys soccer - 15.4% of 135901 29167 for girls soccer - 21.5% of 135901 50096 combined soccer 50096/135901 = 36.9% of all concussions were from soccer.
- jotux 9y agoYou are correct. That is such an odd way to represent data like this.