Finding a value to show there's significant differnece between two population means

#1
I'm working with two data sets and want to compare the difference of means between two particular rankings between two different years. My boss wants me to find a number to prove there is significant difference between these means. I am wondering if there is a statistics equation to find this number?
I've already found the standard error and suggested we use that range to prove there is a difference but he insists that there is a way to find another number to represent "significant difference".
I understand this seems very vague, I'm not a stats major (though I have talked with one and they are also not too sure what my boss is looking for!)

Thanks for any advice!
 

gianmarco

TS Contributor
#2
Re: Finding a value to show there's significant differnece between two population mea

I assume that your samples are independent. There are two ways to test the difference:
a) parametric test: t-test
b) non parametric: Mann-Withney test

The choice depends on some features of your data (e.g. normal or non-normal distributed).
There are many earlier threads on these tests, so you could try and search on your own.


Hope this helps
Regards
Gm