Is there a hypothesis test that two entropies are different?

12 Views Asked by At

I have 2 equal sized groups of people and each group selected A, B, or C for 6 questions. The first group was given no guideline and the second group was given a guideline. I have a hypothesis that the guideline caused people to answer the questions a specific way, and thus the answers in the second group would be more "concentrated" and the answers in the first group will be evenly spread out among A, B, and C. I think entropy would be a good measure of this concentration. My idea for a hypothesis test is to take the mean entropy across the 6 questions for each group. Is there a test that the entropies are "different" enough to be statistically significant?