Question

I am trying to prove that the second half of two part names determines the gender of the whole name in a Romance
language.

I have 30 second parts used only by males in my population and 5 only by
females, with one used by both. It seems that I could reject the nul hypothesis based on 1 vs. 35, but what formula do I use. The books do nor seem to deal with this type of situation.

Answer 1

The answer is not in math, but rather in the criteria of certainity. How certain do you wish your proof to be? 100Percent 90 percent, or 50 percent? Once the certainity is decided, thence you can reject the hypothesis or accept it using any one or more statistical models.

Answer 2

I thought that 95% was most common. My problem is finding statistical models that do not involve means and other nonpertinent functions.