Fig. 4. Replicating the main effects across five distinct data sets.
Experiment 1 (n = 235) was pre-registered on Open Science Framework (https://osf.io/chsvw) following a Pilot study (n = 93). Experiments 2 to 4 include different experimental manipulations of the Deliberation phase (depicted by yellow and cyan colors). We manipulated the number of times each deliberation pair was presented (Experiment 2, memory manipulation, n = 96), the magnitude of potential gains (Experiment 3, motivation manipulation, n = 95), and the high-level association of pairs by assigning some of them as belonging to the same painter (Experiment 4, binding manipulation, n = 93). a The tendency to select rewarded paintings for chosen pairs (Schosen+), but unrewarded ones for unchosen pairs (Sunchosen0) was replicated across conditions. b Memory of the deliberation pairs was related to inverse decision bias (difference in decision tendency of chosen and unchosen pairs) in all experiments and conditions. Points denote trial-averaged data of individual participants. In panel (a), error bars denote standard error of the mean. In panel (b), betas represent the memory slope coefficients in a Bayesian linear regression predicting inverse decision bias as a function of memory accuracy, and for both betas and model fits we show median and 95% highest-density intervals of the corresponding posterior distribution. S+ denotes a rewarded stimulus (for unchosen stimuli, this is the outcome of their chosen counterpart). Source data are provided as a Source Data file.