I have two different regression models developed from two different datasets as follows:

Y1 = B0*X1+B1 , R2 = 0.16

Y2 = C0*X2+C1 , R2 = 0.34

Y1 and Y2 are representing the same measure but they have different values

X1 and X1 are different features.

How I can compute the importance (contribution) independent variable in each model?

  • 1
    $\begingroup$ In case X1 and X2 are two different features, I suggest you create a single regression model containing both features X1 and X2. The p-values give you then an indication on the contribution of each feature. If X1 and X2 are the same features, the result of the model is due to the split of the dataset. $\endgroup$
    – Molitoris
    Commented Jun 23, 2021 at 6:58
  • $\begingroup$ X1 and X2 are different features from different datasets and cannot be entered in the same model. $\endgroup$ Commented Jun 23, 2021 at 7:04
  • 1
    $\begingroup$ 1) Think about what you mean by 'importance' and look at other questions in the tag that you have used. 2) Be careful about using p-values to quantify this (I would not use them). 3) Having 2 predictors, models and datasets does not change anything about your question, which is basically "How do I quantify importance in any regression?" $\endgroup$
    – mkt
    Commented Jun 24, 2021 at 6:45
  • 1
    $\begingroup$ e.g. stats.stackexchange.com/questions/422769/… $\endgroup$
    – mkt
    Commented Jun 24, 2021 at 6:46
  • $\begingroup$ @Reinstate Mocnica Thank you 😊 $\endgroup$ Commented Jun 24, 2021 at 10:34


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.