Fig. 5: Comparing learning across experiments.
From: Visual perceptual learning of feature conjunctions leverages non-linear mixed selectivity
a Conjunction discrimination performance is significantly better after conjunction learning with global feedback (Experiment 1) than separate feature learning (Experiment 3) (p = 0.0367, d = −0.7277; left bars). In contrast, there is no statistically significant difference in conjunction discrimination performance after training with feature-specific feedback (Experiment 2) versus separate feature training (p = 0.1441, d = −0.5767). Brief conjunction discrimination training (identical to Experiment 1) does not significantly affect conjunction discrimination performance after separate feature learning (p = 1, d = 0.1997; right bars, CT: Conjunction training, ST: Separate feature training). * denotes p < 0.05, n.s. denotes non-significant. b Orientation LIs. Orientation is learned faster and better when trained in conjunction with color (experiment 1) than in the presence of task irrelevant color (experiment 4; F(1,27) = 8.2049, p = 0.008, η2 = 0.2331). CLGF Conjunction Learning with Global Feedback, CLFF Conjunction Learning with Feature-specific Feedback, SFL Separate Feature Learning, OL Orientation Learning. Error bars represent the standard error of the mean.
