Log-in | Contact Jeff | Email Updates

Question 185:

1

Answer:

No answer provided yet.

Most likely, your x variables are correlated too highly with each other (called collinearity). In regression, you ideally want the correlation between your x-variables to be low to none, but each correlate highly with your Y predictor.

Check the individual correlations for all 3 combinations of your x-predictors to see how highly they are correlated, then remove the predictors in your equation which have the highest correlation. Its a bit of trial and error on which variables to include, but use your understanding of where the data came from to pick the better predictor. Do this until you have significant predictors.

If your sample size is low, it's also more difficult to have many significant (p<.05) predictors in a regression equation, if you can increase your sample size and re-run the model, check the predictors then. Regardless of the sample size, highly correlated x's are undesirable, so always check for them.

Not what you were looking for or need help?

Ask a new Question

Browse All 869 Questions

Search All Questions: