Hi

My independent variable consists of two variables. The first variable ( A) is a discrete number, which is obtained from the difference of two other discrete numbers, and the other variable (B) is a ratio.

Previous literature calculated the independent variable in the following way:

independent variable = standardize value of A + standardize value of B

Another published paper calculated the independent variable in the following way:

independent variable =Log(A) * B ( Interaction term of Log(A )and B)

Both indicate that a higher value of the independent variable is a higher influence on the dependent variable.

My confusion is that why will I prefer one over the other?

In the paper, I did not find any sound logic. What is the advantage of using standardization? Why should I use interaction if I use so?

I also check the correlation between the two measures. They are highly positively correlated (0.90).

Could anyone help me with sound logic (including some references if any to support whatever I do)?