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)?
Related Posts with Interaction or standardization
difference-in-difference analysis with a random effects modelDear all, So far I learned that DiD is used with a Fixed effects model. But I just read a paper of …
Bin by day on x-axis from clockI have a continuous clock variable, with a cumulative count over time. I want to plot the counts ov…
Planting multiple regressionHey, I hope you all safe and well. I'm a new to STATA and doing multiple regression modeling using x…
Planting multiple regressionHey, I hope you all safe and well. I'm a new to STATA and doing multiple regression modeling using x…
Planting multiple regressionHey, I hope you all safe and well. I'm a new to STATA and doing multiple regression modeling using x…
Subscribe to:
Post Comments (Atom)
0 Response to Interaction or standardization
Post a Comment