Hello All

#1
Hello to All,
I'm newby in such field, ... sorry with my question.

I'm trying to use the regression analysis of only 2 numbers.
It seems to be excluded of any statistics logic.
But what's happening if the third number would be the average of both known numbers?
Does the new calculation with 3 terms, one being the average of other two, will improve the results?
Thank you in advance for support!
 
#2
Hi welcome. I'm not sure I understand your question. What do you mean with '2 numbers'? Do you mean 2 (independent) variables? If you really mean 2 numbers, then regression makes no sense. Regression deals with variables/vectors of numbers, not with single numbers.

If you mean variables, then no, averaging two variables and adding that to your model will not improve the results. The problem when one of your variables can be expressed as a linear combination of other variables, is that you can't reliably determine the values of your regression coefficients. There are too many solutions for the regression equation.