problem in analysis of augmented design

please help with brief explanation on how to better anlyse my data using augmented design

Honore Kam

Shaibu, you can use REML analysis in Genestat

posted on April 26, 2013 12:18PM
Marcos Malosetti

Dear Shaibu,
What type of design did you use? You may have a non-resolvable or a resolvable augemented design.

If it is a non-resolvable design, you only have replications of your check genotypes, and the experiment looks as a standard design with respect to the checks (for example an RCBD), with the addition of the tested unreplicated lines. The analysis in that case can be done using a mixed model with random block effects, which is the standard model for a RCBD experiment: y=mu + block(random) + genotypes (fixed) + error.

If the experiment was setup as a resolvable augmented design, it means that all tested lines are replicated, but checks are used inside each replication to account for incomplete block effects, the model is: y=mu + full replicate(fixed) + block within replicate (random) + error.

Hope it helps,

Regards,

Marcos

posted on May 7, 2013 10:05AM
Back to top