[R] question about Tukey test
Ivan Piffer
ivan.piffer at ismaa.it
Tue May 20 17:00:09 CEST 2003
I'm using SAS for my statistical purpose and I'm triyng to jump in the R'
world.
How can i implement Tukey linear test in R?
In sas-code my lines are :
PROC GLM;
BY PARCELLA;
CLASS TESI;
MODEL PROD=TESI;
MEANS TESI/TUKEY lines;
IN R:
ab<-type.convert(a$tesi)
a<-cbind(a,ab)
cc<-anova(lm(prod~ab,data=a))
cc2<-aov(prod~ab,data=a)
tt<-TukeyHSD(cc2,ordered=TRUE)
but how can I put the option lines that i find in sas (MEANS TESI/TUKEY
lines) in R language?
My output in sas is this and I wanna have the same in R.
Thank you...
Ivan
General Linear Models
Procedure
Tukey's Studentized Range (HSD) Test
for variable: PROD
NOTE: This test controls the type I
experimentwise error rate, but generally has a higher type
II error rate than REGWQ.
Alpha= 0.05 df= 98 MSE=
12.43238
Critical Value of Studentized
Range= 4.486
Minimum Significant
Difference= 4.5893
WARNING: Cell sizes are not
equal.
Harmonic Mean of cell
sizes= 11.88
Means with the same letter are not
significantly different.
Tukey Grouping Mean
N TESI
A 11.788
12 R_GALA_TENROY
A
B A 8.385
12 MONDIAL_GALA
B A
B A 8.124
12 R_GALA_TENFOR
B
B 7.134
12 GALAXY
B
B C 6.301
12 ROYAL_GALA
B C
B C D 4.874
12 OBROGALA
B C D
B C D 3.864
11 BROOKFIELD
C D
C D 1.713
12 REDGALA95
D
D 1.238
12 SCHNIGA
More information about the R-help
mailing list