Statistical test for comparing percentage not mean

#1
Hi,

I am wondering what test should I use to compare two percentage value from two different observations. Let's say I have a damage of crop 15% versus 2% in control, is there a ways to statistically test the significance between the two values? Not sure if this make sense statistically!!! Your input is much appreciated. Thanks a lot.
 

Karabiner

TS Contributor
#2
You have 2 variables (group and damage yes/no), and you need their respective frequencies (e.g. "from n=200 crops examined, 30 were damaged and 170 were not damaged; from n=100 controls, 2 were damaged and 98 were not damaged"). Then you can perform a crosstabulation with Chi² test.

With kind regards

Karabiner
 

hlsmith

Not a robit
#3
Ag-Bob, provide more information. The shortness of your post leaves us guessing at your true study question. Karabiner's recommendation may be correct or not, we can tell without more information!