03-15-2017 08:08 AM
I have a test population of 928 customers and I have their spends. I also have a control population with 86k customers.
I want to make both these populations similar according to its spend.
How do I go about this?
Your help would be greatly apprecaited.
03-15-2017 02:16 PM
Random sampling is supposed to take care of that. How was your test sample selected? What do you want to match specifically: the mean spending, the variance, the whole statistical distribution?
03-16-2017 03:32 AM
My sample was selected based on a calling campaign. If they were pitched its a test population. The control is anybody not pitched the campaign. The problem with the control is that they might be pitched another campaign as it is quite different from the test population as compared in the average spend.
So, the aim is to make the control similar to the test population.
Ideally, we also have industry to which they were pitched the idea to. We also have individual spends of these ppl in that industry.
So, coming to your question..How do I want to make them similar: Possibly the spend overall, as well as industry spend.