[R-sig-ME] Tukey after lme does not match

Alen Hajnal Alen.Hajnal at usm.edu
Fri Jan 6 01:22:22 CET 2012


Dear R users:
I have a simple lme model based on the following data:
sub	trial	angle	RMSy
1	1	30	3.745084
1	2	0	7.520667
1	3	90	11.17038
1	4	15	7.581526
1	5	60	11.17822
1	6	75	8.440891
1	7	45	13.19024
1	8	15	9.822035
1	9	60	6.002665
1	10	75	4.393961
1	11	0	7.436676

When I run the model the results show that 0vs45, 0vs60, 0vs75, and 0vs90 are all significant main effects:

> m.base01<-lme(RMSy ~ angle+trial, data=data,  random=~ trial|sub,method='ML')
> summary(m.base01)
Fixed effects: RMSy ~ angle + trial 
                	Value 	Std.Error  	DF     t-value     p-value
(Intercept)  5.236020 	0.6312637 	267  8.294505  0.0000
angle15     -0.687669 	0.4140277 	267 -1.660925  0.0979
angle30     -0.571092 	0.4129365 	267 -1.383001  0.1678
angle45     -0.984597 	0.4139330 	267 -2.378638  0.0181
angle60     -0.874718 	0.4135615 	267 -2.115085  0.0353
angle75     -1.389835 	0.4113411 	267 -3.378788  0.0008
angle90     -1.620493 	0.4133209 	267 -3.920666  0.0001
trial       -0.009372 	0.0299782 	267 -0.312620  0.7548

However, when I try to run a Tukey posthoc test, I get different results (notice that the estimates are the same as the main effect values above): ONLY 0vs90 is significant at p<.05 :

> summary(glht(m.base01, linfct=mcp(angle = "Tukey")))

         Simultaneous Tests for General Linear Hypotheses

Multiple Comparisons of Means: Tukey Contrasts


Fit: lme.formula(fixed = RMSy ~ angle + trial, data = data, random = ~trial | 
    sub, method = "ML")

Linear Hypotheses:
             Estimate Std. Error z value Pr(>|z|)   
15 - 0 == 0   -0.6877     0.4082  -1.684  0.62652   
30 - 0 == 0   -0.5711     0.4072  -1.403  0.80077   
45 - 0 == 0   -0.9846     0.4081  -2.412  0.19329   
60 - 0 == 0   -0.8747     0.4078  -2.145  0.32592   
75 - 0 == 0   -1.3898     0.4056  -3.427  0.01070 * 
90 - 0 == 0   -1.6205     0.4075  -3.976  0.00138 **
30 - 15 == 0   0.1166     0.4079   0.286  0.99996   
45 - 15 == 0  -0.2969     0.4078  -0.728  0.99090   
60 - 15 == 0  -0.1870     0.4078  -0.459  0.99931   
75 - 15 == 0  -0.7022     0.4050  -1.734  0.59342   
90 - 15 == 0  -0.9328     0.4073  -2.290  0.24842   
45 - 30 == 0  -0.4135     0.4082  -1.013  0.95124   
60 - 30 == 0  -0.3036     0.4074  -0.745  0.98971   
75 - 30 == 0  -0.8187     0.4052  -2.021  0.40123   
90 - 30 == 0  -1.0494     0.4076  -2.574  0.13377   
60 - 45 == 0   0.1099     0.4077   0.269  0.99997   
75 - 45 == 0  -0.4052     0.4059  -0.998  0.95449   
90 - 45 == 0  -0.6359     0.4078  -1.559  0.70835   
75 - 60 == 0  -0.5151     0.4050  -1.272  0.86496   
90 - 60 == 0  -0.7458     0.4074  -1.831  0.52730   
90 - 75 == 0  -0.2307     0.4049  -0.570  0.99763  

Why does the Tukey not match the lme results?
Any help would be much appreciated!
Thanks, Alen




----------
Alen Hajnal, PhD.
Department of Psychology
The University of Southern Mississippi
118 College Drive #5025
Hattiesburg, MS 39406
USA
Tel. +1 (601) 266-4617
alen.hajnal @ usm.edu
http://ocean.otr.usm.edu/~w785427/lab.html



More information about the R-sig-mixed-models mailing list