Hi everyone,

I'm doing research on business owners, who have started multiple businesses. I had a sample of 2500 businesses and looked at the business owners' success rate from their first to last business. That gave me two groups: 1) business owners who were SUCCESFUL in their first business 2) business owners who were UNSUCCESSFUL in their first business.

I then used several key variables to determine what kind of changes they made from their first to their last business. My result was basically that previously successful business owners make MORE changes to their last business compared to previously unsuccessful ones.

So now I am wondering whether there is some correlations test I could apply to show that my results are significant. I'm honestly a bit lost since it's been 7 years that I've last taken a statistics class. I was thinking something like the ANOVA test...so I was looking up tutorials on how to do it (I use Python programming) but I got a bit lost.

I only have 4 things I want to input:
A) #of previously successful business owners B) #their percentage of change
C) #of previously unsuccessful business owners D) #their percentage of change

All the ANOVA examples I looked at used a ton of variables. Is there no simple way to do this? I was thinking of using something along these lines, but I'm unsure how to change it to use my inputs (A-D): http://hamelg.blogspot.nl/2015/11/py...art-16_23.html

Or maybe it's not even the right correlations test? Is there another test that seems more applicable for my situation?

I'd be happy to know I'm even on the right track with ANOVA...so I could take a few tutorials on that and move on with my life.