Agriculture Reference
In-Depth Information
oneway
trangerm trt
Transformed data:
Analysis of Variance
Source SS df MS F Prob > F
----------------------------------------------------------------
Between groups 18.1114002 23 .787452182 100.14 0.0000
Within groups .377453807 48 .007863621
----------------------------------------------------------------
Total 18.488854 71 .260406394
Bartlett's test for equal variances: chi2(23) = 9.6640
Prob>chi2 = 0.993
Compare the two ANOVA tables and notice how the chi
2
is no lon-
ger significant (p ≤ 0.05) with the transformed data indicating the
variances are equal. In both analyses the treatment (between groups)
effects are significant; however, the detected differences between the
treatments will be different. In Chapter 8, Post Hoc Tests, I covered
multiple range tests including Duncan's Multiple Range Test, which
we will use again here. The
pwcompare
command will give us simi-
lar results, but all comparisons are shown, whereas the duncan.do file
condenses the output making it easier to see the results. Load the do
file duncan.do. This program was originally written to analyze data
from a RCBD, so a couple of minor changes will be needed to use it
with a CRD. Find the following piece of code and make the follow-
ing changes; comment out the rep argument in the first line below as
well as e(df_2)+1 in the third line. Enter 3 as the value for the local
macro repl.
anova `depend' `indep' `rep'
// Calculates anova
local var = (e(rmse))^2
// Error mean square from ANOVA
local repl = e(df_2)+1
// Number of replications
When finished, it should look like this:
anova `depend' `indep' //`rep'
// Calculates anova
local var = (e(rmse))^2
// Error mean square from ANOVA
local repl = 3 //e(df_2)+1
// Number of replications
Run the duncan.do file and then enter the command
duncan
germ trt
Search WWH ::
Custom Search