Fig. 1

ROC curves for the A young and older groups in the control condition, B control and experimental conditions for the young group, and C control and experimental conditions for the older group. The data were binned into 4 levels of confidence: high confidence (ratings of 6 and 7), medium–high confidence (rating of 5), medium–low confidence (rating of 4), low confidence (ratings of 3, 2 and 1). The size of each data point represents the relative frequency of the responses at each level of confidence. The region used for calculating pAUC for each condition is shown by the shaded regions. The dashed curves represent signal detection model fits (equal variance independent observation model). The error bars are 68% bootstrap confidence intervals, which are the bootstrap equivalent of ± 1 standard error. The black dashed line indicates chance performance