Statistical test and power

Thank you for your help! I am having trouble understanding what statistical test I can use, in order to figure out a power analysis.

I want to prove that an animal is capable of vocally matching a stimulus. So I have a pre-recording of the animals call, a artifical call played to the animal, and then the animal's responses.

There are several ways calls can vary, and there a 7 variables I will be changing (but only one at a time). When varying each of these 7 variables they are varied on a continious scale (any numerical interval for example between 150 to 1,500).

So I want to show that the call that is played to the animal, the animal matches its response. It seemed like a t-test would be appropriate? The hypothesis would then be that the played sound and animals response the same?


TS Contributor
Hi mandalions9,

To me, this sounds more like an experimental design with seven treatments/factors and one continuous response. Unless a colleague sees something I'm missing, I think you should analyze this experiment with an ANOVA. Try also looking for General Linear Models, those are closely related. Good luck