 From: ohmiee 20060422 07:15 pm (UTC)
Hmm...  (Link)

Forgive me, as a mere humble undergrad suffering through the first year of a math probability/stats course.
Could you do it as a population proportion test? In that case, let's say in a year, there are S amount of hate crimes (poisson), XS (binomial) of which are hate crimes against queers. Set p = X/S, a beta random variable. Would it be wrong to assume the null hypothesis is that p should be .5 and construct a confidence interval, etc?
We haven't technically covered testing, yet, so I might be way in left field.