Learning about and from others' prudence, impatience or laziness: The computational bases of attitude alignment
Fig 4
Quantification of false-consensus (A) and influence biases (B), in terms of a comparison between the Same (yellow) and Different (purple) conditions. A: Top: average performance (+/- standard error on the mean) during the Prediction phase is plotted as a function of session stage (beginning/end) and condition type (Same/Different). Bottom: histogram of the ANOVA residuals (grey bars) and moment-matched Gaussian approximation (red line). B: Top: average difference in the number of low-cost choices between Decision phases 1 and 2 (+/- standard error on the mean) is plotted as a function of participants’ initial cost-susceptibility (low/high) and condition type (Same/Different). Note: in the Different condition, participants with an initial high (resp., low) cost-susceptibility have observed an artificial agent endowed with a high (resp., low) cost-susceptibility. Bottom: histogram of the ANOVA residuals (grey bars) and moment-matched Gaussian approximation (red line).