Employing individual measures of baseline glucocorticoids as population-level conservation biomarkers: considering within-individual variation in a breeding passerine.

The final number of clusters was validated based on two characteristics obtained from a discriminant function analysis, with cluster identity as the dependent variable and the original habitat variables as independent variables ( Leimeister, 2010 ), as follows: (i) a highly significant Wilks's λ (Wilks's λ = 0.046; P < 0.0001) indicating that >95% of the total variance in the discriminant scores was explained by differences between groups (clusters); and (ii) investigation of the number of errors the discriminant function analysis produced; two clusters produced the lowest number of classification errors (1%).

also in 132,142 other papers