To assess the precision of the two analysis methods, i.e., the
separate ANOVA and the BLUP based on a variance component
mixedmodel, we calculatedthe Pearsoncorrelationofthe genotype
prediction between locations and compared it between the two
analysis methods. The correlation between locations was not only
affected by differences in precision, but also by genotype–location
interaction. Clearly, the presence of genotype–location interaction
effects will attenuate the correlation relative to a situation with no
interaction. Since the operating interaction effects are the same for
both analysis methods, however,the correlations can serve as a relative
measure of precision fo