search

runehaubo / lmerTestR

Repository for the R-package lmerTest
48 stars 9 forks source link

Question about lmerTest #28

Closed kevinmchu closed 4 years ago

kevinmchu commented 4 years ago

I built a linear mixed effects model and would like to test the statistical significance of the fixed factor coefficients using the lmerTest package. However, I am not entirely sure how to test for significance, so I have a couple of questions.

1) Can I test for significance by passing the mixed effects object through the anova method? 2) Do I need to run the step method if I am testing for coefficient significance?

runehaubo commented 4 years ago

You can use anova() summary() drop1() and ranova() for random effects - there is no need for step in this case.

Cheers Rune

man. 9. sep. 2019 kl. 18.51 skrev kevinmchu notifications@github.com:

I built a linear mixed effects model and would like to test the statistical significance of the fixed factor coefficients using the lmerTest package. However, I am not entirely sure how to test for significance, so I have a couple of questions.

  1. Can I test for significance by passing the mixed effects object through the anova method?
  2. Do I need to run the step method if I am testing for coefficient significance?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/runehaubo/lmerTestR/issues/28?email_source=notifications&email_token=ABHZZLNQJVHCWT6SG4ACMP3QIZ5KDA5CNFSM4IU5UTL2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HKHI24A, or mute the thread https://github.com/notifications/unsubscribe-auth/ABHZZLPTRWAUR3EIDGACRLLQIZ5KDANCNFSM4IU5UTLQ .

kevinmchu commented 4 years ago

Thank you for your prompt reply. Do I need all those methods to test for statistical significance of fixed effects coefficients? Or can I pass the lmm object directly to the anova function?

runehaubo commented 4 years ago

You may only need the anova function depending on your situation - try out the functions and see if you get what you want. Further questions long these lines should be sent to the mixed-models mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

Cheers Rune

On Mon, 9 Sep 2019 at 20:05, kevinmchu notifications@github.com wrote:

Thank you for your prompt reply. Do I need all those methods to test for statistical significance of fixed effects coefficients? Or can I pass the lmm object directly to the anova function?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/runehaubo/lmerTestR/issues/28?email_source=notifications&email_token=ABHZZLNZS24ZV2STOCNXZDLQI2F7JA5CNFSM4IU5UTL2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD6IQP3Y#issuecomment-529598447, or mute the thread https://github.com/notifications/unsubscribe-auth/ABHZZLLCXIOTMIJSWI5RDCLQI2F7JANCNFSM4IU5UTLQ .