Pablo Martínez-López

@martinezlopezp.bsky.social

90 Followers 135 Following

PhD student in Experimental Psychology at @univmalaga.bsky.social https://martinez-lopezpablo.github.io

A) Overview of Exp. 2 design. All participants followed the same procedure: habit formation was induced and tested using one set of stimuli, and approximately 4 months later, participants repeated the procedure using a different stimulus set. Thus, both habit formation and expression had to remain stable to yield consistent individual differences across tests. Mappings were classified as original (Mapping A) or revised (Mapping B, in which two associations switched responses to test habit behavior), and as involving the first (1) or second (2) stimulus set, yielding Mappings 1A, 1B, 2A, and 2B. (B) Mean number of trials to reach the learning criterion (5 hits in a row per association) for the trained A mappings. (C) Data from the training stage. Each point represents one block. The upper panel shows mean RTs for correct responses, and the lower panel shows error rates. Error bars denote standard errors. (D) Data from the criterion blocks for each revised mapping. (E) Performance as a function of preparation time, showing responses for consistent associations (first panel), remapped associations under Mapping 1B (second panel), and remapped associations under Mapping 2B (third panel). (F) Habit formation was confirmed by comparing habit errors vs. other errors with ~300–600 ms preparation time. Practice effects were observed as habit and other errors decreased in the second assessment. (G) Correlation in habitual behavior across stimulus sets over the 4-month interval. (H) Correlation between improvement in RTs for correct responses during training and the proportion of habitual errors in the corresponding forced-response test. The upper panel shows the association between RT improvement during training on Mapping 1A and habitual errors under Mapping 1B; the lower panel shows the corresponding association for Mappings 2A and 2B. (I) Correlation in RT improvement across stimulus sets.
A) Scatterplot illustrating the correlation between the proportion of habit responses in the forced-response task and RT switch cost (in seconds) in the outcome-devaluation task (n = 47). B) Scatterplot depicting the relationship between RT switch cost (in seconds) in the forced-response and outcome-devaluation tasks (n = 47). C) Scatterplot showing the correlation between habit response and RT switch cost in the forced-response task (n = 50). In each figure, a marginal plot shows a histogram with a density plot overlapped for each measure.
Forced-response task. D) LMM predictions for goal-directed response RTs across stimulus types, training conditions, and stimulus onsets (i.e., excluding preparation times below ­ tmin). Shaded areas represent the 95% CI. The dashed lines illustrate the temporal moment in which each tone was displayed during the trial, whereas the solid line represents the imposed RT. E) Habit errors across preparation time windows for each training condition. Bars depict participants' mean proportion of habit errors, while error bars indicate the SEM. Dots show GLMM predictions, with error bars depicting the 95% CI. F) GAMM predictions for habit errors across training conditions and preparation times. Shaded areas denote the 95% CI. G) The partial derivative of the GAMM predictions for habit errors across training conditions and preparation times. Positive values indicate positive slopes (increasing likelihood of habit errors), negative values indicate negative slopes (decreasing likelihood of habit errors), and values near zero indicate a flattening of the function. Shaded areas represent the 95% CI