Consider an e-commerce company, has partnered with two different local retailers for a same product. Retail ‘A’ receives an average rating of ‘a1’ stars while retailed ‘B’ receives ‘B’ stars. Which one of the two is better? Of course the one that has higher rating. However, the question to answer here is, ‘is the difference statistically significant’? To further elaborate, it might be possible that the consumers receiving goods from a particular retailer might just be in general less appreciating the product itself.
Following tests a data analyst can conduct while comparing such insights:
-
Paired t-test: If the user groups is similar and from the same population
-
Independent t-test: If the user groups belong to different population