Relative task performance: control minus conditioned LS mean time in the green arm (positive = learning; authors' value) [Proportion / Ratio]