Fig. 4

Summary data from the example dataset and posterior predictive distributions from one- and two-mixture models. The red dots represent the observed means. The green line and ribbon represent the mean and 95% credible interval of the posterior predictive distributions from the one-mixture model. The blue line and ribbon represent the two-mixture model. Columns are divided according to participants’ mixture weights in the two-mixture model. The left column displays the data from participants whom the two-mixture model suggested were most strongly influenced by mixture 1 (i.e., had a mean mixture weight greater than 0.5). The right column displays the data from participants who the two-mixture model suggested were most strongly influenced by mixture 2 (i.e., had a mean mixture weight less than 0.5)