I have a dataset that I am trying to analyse, it consists of:

  • A binary variable which indicates a tree species (0 = deciduous 1 = evergreen) with 100 measurements each.
  • N which is leaf nitrogen content (%, continuous)
  • P - leaf phosphorus content (%, continuous)
  • SLA - specific leaf area measurements (cm2, g-1, continuous)
  • Lignin content (%, continuous)

I am trying to test to see which traits differentiate between the two tree types. I initially analysed this using a logistic regression using glm() in R but now I am not sure if this is the correct analysis. When I used the exp() function to convert the coefficients, it produced these values from the output:


(Intercept) - 2.235862e-06

sq_sla - 3.585946e+00

l_p - 5.352951e-03

N - 3.920587e-01

It seems like there is something wrong but I have done all of the procedures correctly for the model so I am wondering if the test is appropriate.

  • $\begingroup$ Logistic regression might be a good idea, but the effects might not be linear! Maybe try to spline your continuous covariates, show us results of that, also some plots. Maybe also interactions? $\endgroup$ – kjetil b halvorsen Oct 20 '20 at 19:02

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.