A marginally significant interaction of Congruency by Block was found in the by-participants analysis, F 1 (3.47, 117.91) = 2.31, p = .07, while this interaction was highly significant in the by-items analysis; F 2 (6, 99) = 3.20, p = .007, η p 2 = .16.
While there was no reliable Block 1 difference between the Training and No Training conditions, the Block 3 difference was highly significant in both sets of analyses, reflecting the substantial improvement across blocks when feedback training was provided, relative to the control condition, t 1 (31.27) = 3.20, p = .001, d = 1.14; t 2 (23) = 11.10, p < .001, dz = 2.27. 5 Finally, to investigate the magnitude of Block 1–Block 3 improvement to incongruent pairings in the Training condition, a final set of t -tests was conducted, t 1 (50) = 4.82 p < .001, dz = .67; t 2 (23) = 10.97 p < .001, dz = 2.24.