Hi, we have done a association study by SAS using proc GLM. Obtained results showed no significant association between observed genotypes and studied traits (model was non significant), but we saw significant differences between genotype means by Duncan and Tukey test. Can anyone help us in conclusion of this results?
Hi Ayoub,
It sometimes happens for Bonferroni or Tukey methods, but it is not true for Scheffe multiple comparison method.
For addressing a reference, one may suggest the following:
Page 90, the paragraph under the table 3.7
Applied Statistics, Analysis of Variance and Regression, Third Edition, R.Mickey, O. Dunn and V. Clark , Wiley , 2004
Hi Ayoub,
It happens sometimes in Bonferroni or Tukey methods, but it is not true for Scheffe multiple comparison. One reference is Page 90, first paragraph under table 3.7, "Applied Statistics, Analysis of Variance and Regression", Third Edition, Wily 2004 by R. Mickey, O. Dunn and V. Clark.
Sincerely,
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques