I have a GAM model with about 10 variables and 450 observations. I ran a simple GAM regression using R and I noticed that everything looks OK, except that one of the splines has many zeros in the p value (0.000006427491961) and when I calculate the confidence intervals for the estimate, I get something like this for the intercept: - estimate 0.06 - se 14 - 95%CI (40, 0.01)

Is this an issue of power? Also, I am not sure what happens if the intercept is significant, I have seen that is some cases in other analyses.

  • $\begingroup$ If you can publish the data (or simulated data), maybe someone can fit a GAM model using other software to see what happens. Otherwise, I cannot find the answers. $\endgroup$ – user158565 Oct 10 '18 at 1:09
  • $\begingroup$ Thanks! I was just wondering in general what does that mean, not necessarily related to GAM. For some reason, I thought it was an issue of power. $\endgroup$ – mikip Oct 14 '18 at 22:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.