[R-sig-ME] how to report the results from lmer() in APA-style
ONKELINX, Thierry
Thierry.ONKELINX at inbo.be
Mon Feb 16 16:29:41 CET 2009
Dear Liliana,
Have at look at https://stat.ethz.ch/pipermail/r-help/2006-May/094765.html. Douglas Bates explaines in that post why you can't find p-values in the summary of lmer().
You could also have a look at RSiteSearch("lmer p-value").
HTH,
Thierry
----------------------------------------------------------------------------
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Research Institute for Nature and Forest
Cel biometrie, methodologie en kwaliteitszorg / Section biometrics, methodology and quality assurance
Gaverstraat 4
9500 Geraardsbergen
Belgium
tel. + 32 54/436 185
Thierry.Onkelinx at inbo.be
www.inbo.be
To call in the statistician after the experiment is done may be no more than asking him to perform a post-mortem examination: he may be able to say what the experiment died of.
~ Sir Ronald Aylmer Fisher
The plural of anecdote is not data.
~ Roger Brinner
The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data.
~ John Tukey
-----Oorspronkelijk bericht-----
Van: r-sig-mixed-models-bounces at r-project.org [mailto:r-sig-mixed-models-bounces at r-project.org] Namens Liliana Martinez
Verzonden: maandag 16 februari 2009 15:42
Aan: r-sig-mixed-models at r-project.org
Onderwerp: [R-sig-ME] how to report the results from lmer() in APA-style
Dear all,
I am trying to apply the lmer function in R 2.8.0. to some linguistic data, but I am at a loss when it comes to reporting the results (see below). The APA recommendations say that effects should be reported as follows:
F (df1, df2) = ... , p. = ...
The question is, where do I find all these things? So far I have learned through different sources that df1 and F can be found through using the anova() function (is this correct?), but where do I find df2 and p ?
I have even bigger problems when my dependent variable has a binomial distribution, because then the anova() and pvals.fnc() functions cannot be applied.
I wonder as well whether there is a commonly approved way of reporting the output of the 'print (xxx.lmer)' and 'xxx.pvals$fixed' commands? I can see that some of the levels of a factor are significantly different from the baseline, and this is of interests for me, but how shal I report it? Or should other tests be applied in order to find the difference between the levels? (and, if yes, what tests?)
Any help/ advice/ references will be greatly appreciated.
Best regards
Liliana
----------------------------------
print (all_v_a_va_vf_vp_vt.lmer , corr = F)
Linear mixed model fit by REML
Formula: rating ~ verb + angle + verb:angle + verb:type + verb:prec +Â Â Â Â Â verb:fol + (1 | subject)
  Data: rating_allbegend_no270
  AIC  BIC logLik deviance REMLdev
 11067 11225 -5507   10941  11015
Random effects:
 Groups  Name       Variance Std.Dev.
 subject (Intercept) 0.092792 0.30462
 Residual            1.691374 1.30053
Number of obs: 3240, groups: subject, 40
Fixed effects:
                              Estimate Std. Error t value
(Intercept)Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â 4.51944Â Â Â 0.13102Â Â 34.49
verbzaobikalia                -1.42685   0.15830  -9.01
verbzaviva                    -2.82315   0.15830 -17.83
angle180Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â -0.88611Â Â Â 0.09694Â Â -9.14
angle360Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â -2.84722Â Â Â 0.09694Â -29.37
verbzaobikalia:angle180Â Â Â Â Â Â Â -0.37778Â Â Â 0.13709Â Â -2.76
verbzaviva:angle180Â Â Â Â Â Â Â Â Â Â Â Â 1.76944Â Â Â 0.13709Â Â 12.91
verbzaobikalia:angle360Â Â Â Â Â Â Â Â 2.24444Â Â Â 0.13709Â Â 16.37
verbzaviva:angle360Â Â Â Â Â Â Â Â Â Â Â Â 4.95833Â Â Â 0.13709Â Â 36.17
verbobikalia:typeround         0.23333   0.12466   1.87
verbzaobikalia:typeround      -0.07407   0.12466  -0.59
verbzaviva:typeround           0.35370   0.12466   2.84
verbobikalia:precno_prec      -0.01111   0.09694  -0.11
verbzaobikalia:precno_prec     0.20556   0.09694   2.12
verbzaviva:precno_prec        -0.24722   0.09694  -2.55
verbobikalia:precsmooth_prec  -0.19722   0.09694  -2.03
verbzaobikalia:precsmooth_prec 0.12778   0.09694   1.32
verbzaviva:precsmooth_prec    -0.15833   0.09694  -1.63
verbobikalia:folno_fol        -0.33333   0.09694  -3.44
verbzaobikalia:folno_fol       0.21389   0.09694   2.21
verbzaviva:folno_fol          -0.14722   0.09694  -1.52
verbobikalia:folsmooth_fol    -0.26667   0.09694  -2.75
verbzaobikalia:folsmooth_fol   0.31944   0.09694   3.30
verbzaviva:folsmooth_fol      -0.04167   0.09694  -0.43
> anova (all_v_a_va_vf_vp_vt.lmer )
Analysis of Variance Table
          Df Sum Sq Mean Sq F value
verb       2 131.07  65.53 38.7458
angle      2 136.09  68.05 40.2310
verb:angle 4 2489.53 622.38 367.9741
verb:type  3  30.62  10.21  6.0350
verb:prec  6  27.89   4.65  2.7478
verb:fol   6  45.62   7.60  4.4952
> all_v_a_va_vf_vp_vt.pvals
$fixed
                              Estimate MCMCmean HPD95lower HPD95upper pMCMC Pr(>|t|)
(Intercept)Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â 4.5194Â Â 4.5198Â Â Â Â 4.2530Â Â Â Â 4.7685 0.0001Â Â 0.0000
verbzaobikalia                 -1.4269 -1.4253   -1.7338   -1.1122 0.0001  0.0000
verbzaviva                     -2.8231 -2.8247   -3.1252   -2.5124 0.0001  0.0000
angle180Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â -0.8861Â -0.8862Â Â Â -1.0768Â Â Â -0.6987 0.0001Â Â 0.0000
angle360Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â -2.8472Â -2.8491Â Â Â -3.0388Â Â Â -2.6610 0.0001Â Â 0.0000
verbzaobikalia:angle180Â Â Â Â Â Â Â Â -0.3778Â -0.3766Â Â Â -0.6426Â Â Â -0.1051 0.0056Â Â 0.0059
verbzaviva:angle180Â Â Â Â Â Â Â Â Â Â Â Â Â 1.7694Â Â 1.7696Â Â Â Â 1.5117Â Â Â Â 2.0402 0.0001Â Â 0.0000
verbzaobikalia:angle360Â Â Â Â Â Â Â Â Â 2.2444Â Â 2.2465Â Â Â Â 1.9892Â Â Â Â 2.5295 0.0001Â Â 0.0000
verbzaviva:angle360Â Â Â Â Â Â Â Â Â Â Â Â Â 4.9583Â Â 4.9611Â Â Â Â 4.7059Â Â Â Â 5.2427 0.0001Â Â 0.0000
verbobikalia:typeround          0.2333  0.2339   -0.0069    0.4895 0.0666  0.0613
verbzaobikalia:typeround       -0.0741 -0.0757   -0.3212    0.1628 0.5384  0.5524
verbzaviva:typeround            0.3537  0.3527    0.0980    0.5983 0.0060  0.0046
verbobikalia:precno_prec       -0.0111 -0.0118   -0.2111    0.1717 0.9012  0.9088
verbzaobikalia:precno_prec      0.2056  0.2050    0.0145    0.3916 0.0344  0.0340
verbzaviva:precno_prec         -0.2472 -0.2456   -0.4378   -0.0615 0.0136  0.0108
verbobikalia:precsmooth_prec   -0.1972 -0.1969   -0.3790    0.0009 0.0412  0.0420
verbzaobikalia:precsmooth_prec  0.1278  0.1278   -0.0743    0.3078 0.1890  0.1875
verbzaviva:precsmooth_prec     -0.1583 -0.1569   -0.3441    0.0336 0.0952  0.1025
verbobikalia:folno_fol         -0.3333 -0.3344   -0.5240   -0.1438 0.0002  0.0006
verbzaobikalia:folno_fol        0.2139  0.2133    0.0254    0.4008 0.0282  0.0274
verbzaviva:folno_fol           -0.1472 -0.1467   -0.3379    0.0420 0.1314  0.1289
verbobikalia:folsmooth_fol     -0.2667 -0.2674   -0.4645   -0.0839 0.0058  0.0060
verbzaobikalia:folsmooth_fol    0.3194  0.3186    0.1252    0.5074 0.0006  0.0010
verbzaviva:folsmooth_fol       -0.0417 -0.0412   -0.2339    0.1437 0.6828  0.6673
$random
   Groups       Name Std.Dev. MCMCmedian MCMCmean HPD95lower HPD95upper
1Â subject (Intercept)Â Â 0.3046Â Â Â Â 0.2993Â Â 0.3027Â Â Â Â 0.2235Â Â Â Â 0.3868
2 Residual              1.3005    1.3011  1.3012    1.2694    1.3329
_________________________________________________________
[[alternative HTML version deleted]]
Dit bericht en eventuele bijlagen geven enkel de visie van de schrijver weer
en binden het INBO onder geen enkel beding, zolang dit bericht niet bevestigd is
door een geldig ondertekend document. The views expressed in this message
and any annex are purely those of the writer and may not be regarded as stating
an official position of INBO, as long as the message is not confirmed by a duly
signed document.
More information about the R-sig-mixed-models
mailing list