11

I have problem with getting p value from my mixed model, library(lme4)

DWR<-lmer(DWRm2~Growth.stage+Se.application+Growth.stage:Se.application+(1|Block),data=Sub1)
summary(DWR)

before I used this model and I got p value with summary my model but now I can't get it, I have just t value

Nick Cox
  • 56,404
  • 8
  • 127
  • 185
Nashmin Ebrahimi
  • 129
  • 5
    library(lme4); help(pvalues) gives some discussion. – conjugateprior Oct 09 '14 at 11:21
  • @Smilig In spite of the similar titles the content seems to be completely different, since the other question seems to be focused on memisc not lmer; further, the answer there would be useless for this Q. – Glen_b Oct 09 '14 at 12:06

2 Answers2

28

I'm pasting the information from help("pvalues",package="lme4") here.

Users who need p-values have a variety of options. In the list below, the methods marked MC provide explicit model comparisons; CI denotes confidence intervals; and P denotes parameter-level or sequential tests of all effects in a model. The starred (*) suggestions provide finite-size corrections (important when the number of groups is <50); those marked (+) support GLMMs as well as LMMs.

  • likelihood ratio tests via anova (MC,+)

  • profile confidence intervals via profile.merMod and confint.merMod (CI,+)

  • parametric bootstrap confidence intervals and model comparisons via bootMer (or PBmodcomp in the pbkrtest package) (MC/CI,*,+)

  • for random effects, simulation tests via the RLRsim package (MC,*)

  • for fixed effects, F tests via Kenward-Roger approximation using KRmodcomp from the pbkrtest package (MC)

  • car::Anova and lmerTest::anova provide wrappers for pbkrtest: lmerTest::anova also provides t tests via the Satterthwaite approximation (P,*)

  • afex::mixed is another wrapper for pbkrtest and anova providing "Type 3" tests of all effects (P,*,+)

  • arm::sim, or bootMer, can be used to compute confidence intervals on predictions.

    When all else fails, don't forget to keep p-values in perspective.

Community
  • 1
Ben Bolker
  • 43,543
17

p-values in lme4 are deliberately not listed by default, see:

There are some "approximations" but better just forget about p-values in lmm (or generally forget about them because they "measure" mostly the sample size).

Ben Bolker
  • 43,543
Tim
  • 138,066
  • 5
    Could you briefly summarize the points made? Links can die, which would render an otherwise valuable answer almost useless. – Glen_b Oct 09 '14 at 10:39
  • 2
    (I do realize that might be quite a difficult task, but it would make your answer substantially more useful, particularly if both links were lost. The links should of course remain; the extensve discussion is highly valuable.) – Glen_b Oct 09 '14 at 11:32