Hello. I am looking for a multiple comparisons test to follow up a
repeated mesures ANOVA i have conducted. Im not an expert in
statistics or in R but i have managed to produce this:
"START"> fit5.lme<-lme(logfaa~segment,random=~+1|fish)
> anova(fit5.lme)
numDF denDF F-value p-value
(Intercept) 1 8 14.553052 0.0051
segment 4 8 4.198819 0.0402
> summary(fit3.lme)
Linear mixed-effects model fit by REML
Data: NULL
AIC BIC logLik
34.53912 36.65722 -10.26956
Random effects:
Formula: ~+1 | fish
(Intercept) Residual
StdDev: 7.073225e-06 0.5134332
Fixed effects: logufr ~ segment
Value Std.Error DF t-value p-value
(Intercept) -1.2899434 0.2964308 8 -4.351584 0.0024
segmentS2 1.0399190 0.4192164 8 2.480625 0.0381
segmentS3 1.1002549 0.4192164 8 2.624551 0.0304
segmentS4 0.3346369 0.4192164 8 0.798244 0.4478
segmentS5 -0.1543765 0.4192164 8 -0.368250 0.7222
Correlation:
(Intr) sgmnS2 sgmnS3 sgmnS4
segmentS2 -0.707
segmentS3 -0.707 0.500
segmentS4 -0.707 0.500 0.500
segmentS5 -0.707 0.500 0.500 0.500
Standardized Within-Group Residuals:
Min Q1 Med Q3 Max
-1.54451640 -0.23806358 -0.01628289 0.20255062 1.56079929
Number of Observations: 15
Number of Groups: 3
"END"
Anyone know how i can follow this up to find out which segment(s)
causes the significant difference?
Snor