Let's say I'm analysing my customers behaviour. I want to know if their wages is related to their purchases. I have two main questions here: If I decide to create groups and check the different outcomes, (1) how can I know if I should split the wages between 0 to 1000 usd, 1000 to 2000 usd, 2000 to 3000 and so on, or if I should split between 0 to 2500, 2500 to 5000 and so on? and (2) if group 1 has 2000 customers, group 2 has 5000 customers, etc., how can I know what's the minimum amount of individuals I need for each group?
Asked
Active
Viewed 16 times
2
-
If you have the original data, why do you want to bin? See https://stats.stackexchange.com/questions/68834/what-is-the-benefit-of-breaking-up-a-continuous-predictor-variable It would be better for you to tell us about your data and which Qs you want to ask from the data ... – kjetil b halvorsen Jan 30 '21 at 02:12