AIC Model comparisons. ΔAIC values for each of the history models ( M 1 3 , color coded) vs. the no-history model ( M 0 ), applied to the data from (B) experiment 2, per condition (marked by one- or two-letter codes, explained in Table 1 ; × marks cross-modal conditions). The bars represent group means and gray circles represent the individual participants’ ΔAIC values, for each experiment and condition. Error bars indicate standard errors of the mean. ΔAIC magnitudes between 2 and 6, between 6 and 10, and > 10 are considered positive, strong, and very strong evidence, respectively, for the model with the lower AIC value. These boundaries (ΔAIC = ±2, ±6, and ±10) are marked by vertical dashed lines. Similar model comparisons using BIC are presented in Additional file 1 : Fig. S3
