what is the best test

#1
Hello!

I'm not the best at statistics. I understand the theory but I am still uncertain about choosing the appropriate test.

I am doing a study on antibiotic cure rates for a disease in cows. I wanted to see if there is an association between the colony number and the outcome (success to the treatment or not). I have put the colonies in categories: 1 if its 19 colonies or less and 2 if its 20 or more colonies. My outcomes are either 0 (success) or 1 (disease). I am comparing two variables to see if there is an association. I was thinking t-test but that calculates the difference between means and since I have categories and not continuous data I dont feel like this would give me much. The chi2 goodness of fit measures proportions, so I feel like that too isn't good for me. Can I use ANOVA?

QUESTION: what test will show me if there is an association between my two groups?:confused:
 

rogojel

TS Contributor
#2
I would keep the colony numbers . Do you have a single outcome ( like succes or failure) per colony or do you have more detailed information (like percentage cured)? Some type of regression seems to me to be the right tool, either logistic or simple.

regards