Fig. 6: Symmetry of representational change within pairs across learning.

A Linear mixed effects modeling revealed that target words in semantically related pairs showed less representational change (i.e. a higher correlation between their initial representation and final representation) across learning than did cue words when correctly recalled at Day 2 (t₍₃₇₂₎ = 3.546, p = 0.002, d = 0.18, 95% CI = [0.039 0.137]). This effect was not observed when comparing cues and targets in unrelated pairs. Green indicates semantically related pairs, yellow indicates unrelated pairs; solid bars indicate cue words, striped bars indicate target words. B Related pairs of words show a significantly different (more negative) asymmetry in representational change than unrelated pairs (t₍₄₅₂₎ = 2.414, p = 0.016, η_p² = 0.01, 95% CI = [0.006 0.064]). All displayed correlation values are Fisher r-to-z transformed. Open circles reflect means of individual participants (N = 80), with connecting lines showing within-subject differences across conditions. Error bars reflect standard error of the mean. Symbols reflect statistically significant differences using two-tailed tests across conditions using Holm-Bonferroni corrections for multiple comparisons (~p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001).