Figure 8
Model Fits to Pain Dictator Game Choices.
(a) BIC values for the alternative models, either with status quo bias (light blue bars, right bar of each pair) or without (dark blue bars, left bar of each pair). The social discounting model provided the most parsimonious fit, indicated by the grey dot. (b) Maximum likelihood choice probabilities of the social discounting model, formed by taking the mean of the predicted choice probabilities across participants, are indicated by gray circles and are overlaid with the observed choice probabilities as displayed in Fig. 4, indicated by blue triangles for choices to take on pain and blue circles for choices to give pain. As in Fig. 4, a single data point represents a mean choice probability across 47 participants and error bars are 95% confidence intervals around the mean, generated by re-sampling the data 100,000 times with replacement.
