The percentages are somewhat misleading - actually insignificant since
the sample sizes are too small particularly at 14.1, 14, and 16.1. The
confidence intervals are just to large.
The normal way to deal with that is to combine "like categories." If 14
and 14.1 are combined with 14.3 that gies 23 horse in that category. If
16 and 16.1 are combined with 15.3 that gives 35 inthat category.
Given the numbers are I would suggest that the probability of finishing
the Tevis is pretty much independent of height.
Truman
DreamWeaver wrote:
I didn't
get the height on all of the horses this year, but looks like I had 216
of them, from the entry information provided on the Tevis site. Here
they are sorted, showing height and how many -- these were from the
start list.