Naked Science Forum
General Science => General Science => Topic started by: scientizscht on 02/10/2020 22:06:49
-
Hello
I have data of a y,x1,x2 where x1 and x2 are independent variables.
How can I assess if the x1 or x2 are related to y and how strongly?
Also, is it possible to determine if there is a causation or simply correlation?
Thanks!
-
https://www.educba.com/3d-scatter-plot-in-excel will give you a quick insight into a 3-dimensional correlation, using the spreadsheet you probably already have (or can buy very cheaply).
Proof of causation requires that you hold x1 or x2 constant and see how y changes when you vary the other.
-
You can use a correlation function on the two variables separately, to see which has the greater correlation.
There are also multi-variable correlation functions.
But these work best if there is a linear relationship (or it can be transformed into a linear relationship).
-
Also, is it possible to determine if there is a causation or simply correlation?
Mathematically, no.