People are at least as good at optimizing reward rate under equivalent fixed-trial compared to fixed-time conditions

Figures

• 

  Fig. 1. Actual threshold values for each condition, compared to their optimal threshold. Circles represent median group-level posterior threshold for each block (starting at block 2) with error bars showing the 95% quantiles. This is compared with the group-averaged optimal threshold indicated by the coloured bands delineated by different shades of green. The lightest, centre band represents the 40–60% quantile, the middle shade represents the 20–40% and 60–80% quantiles, and the darkest shade the 10–20% and 80–90% quantiles. Thresholds above the optimal band indicate cautious behaviour, whilst those below the band indicate more urgent behaviour

• 

  Fig. 2. Prior-posterior distribution density plots of the group-level \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta _c$$\end{document}Δc parameter using the normal prior distribution. The prior distributions are in blue, and the posterior distributions are in red

• 

  Fig. 3. Experiment 3 comparisons of RR-Time/ None-Trial group combination. a Actual threshold values for each condition, compared to their optimal threshold. b Prior-posterior distribution density plot of the group-level \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta _c$$\end{document}Δc parameter using the normal prior distribution