Fig. 7: Model comparison for Study 3.

Points in the first and second rows show sampling rates. Human participant sampling data are reproduced from Fig. 2 (See also Table 3) and show that participants sample more (P < 0.05) in the 14-option condition (n = 65 participants) than the 10-option condition (n = 75 participants). In the first row, horizontal solid lines link samples data for humans (black points) and Ideal Observer (grey points) when BF₀₁ > 3 (at least moderate evidence for equal means) while dotted lines indicate when BF₁₀ > 3 (at least moderate evidence for different means). Neither humans nor the Ideal Observer showed BF₀₁ > 3 (at least moderate evidence for equal means). Statistical details for these pairwise tests can be found in Supplementary Table 1. Undersampling is presenting only for the longer sequence length. The second row shows BIC values, where lower values indicate better model fit. Black horizontal lines indicate when BF₀₁ > 3. When BF₁₀ > 3, the horizontal line is coloured the same as the point spread of the better model. Statistical details for these pairwise tests can be found in Table 7. The third row shows numbers of participants each model best fit. The second and third rows both show favourable evidence for Biased Prior as the best model. Model point spread data colours (See also legend in lower right panel): Cost to Sample (green), Cut Off (orange), Biased Prior (blue). Boxplots reflect first, second (median) and third quartiles, while whiskers reflect 1.5 interquartile range. Point spreads reflect individual participant mean values. Abbreviations: Subj = Models that make choices about subjective values; Obj = Models that makes choices about objective values.