2.3. Data analysis
Separate analyses of variance (ANOVAs) were performed
on the data collected in 2010 and 2011 for each environment
(drought stress and well-watered) to generate entry means
adjusted for block effects according to an alpha lattice design.
Replications, years and incomplete blocks were considered as
random effects while experimental hybrids were considered
fixed effects. Hybrids were then analyzed as a randomized
complete block design (RCBD) combined over the two years
because the lattice design did not have significant advantage
over RCBD. All analyses were performed with PROC GLM in
SAS [11] using a RANDOM statement with TEST option.
Person's correlation coefficients between grain yield and other