Page 169 - Demo

Eye-tracking reveals bias to flanges in orang-utans1677Figure 2. Results of the replication of Experiment 1 in Experiment 2 depicting (A) predicted probability of fixating first on the flanged male stimulus and (B) predicted proportion of time spent fixating on the flanged male stimulus. We report the overall prediction and the predictions for each participant. Grey areas represent the posterior predictions for each iteration of the model, black dots indicate the median posterior prediction, and black lines indicate the 95% credible interval.With regard to the total fixation duration, we found that the interaction between morph of the left stimulus and morph of the right stimulus predicted looking time bias towards the left stimulus (Supplementary Table 4; Figure 3B; see Appendix G for model stability checks). If a flanged male and unflanged male stimulus were paired, the bias deviated robustly from 0.5 (flanged left-unflanged right: b = 0.588 [0.030], 89% CrI [0.540; 0.636], pd = 1.000; unflanged left-flanged right: b = 0.427 [0.028], 89% CrI [0.382; 0.473], pd = 0.996). However, this was not the case when two males of the same morph were shown (flanged-flanged: b = 0.521 [0.041], 89% CrI [0.455; 0.585], pd = 0.702; unflanged-unflanged: b = 0.505 [0.039], 89% CrI [0.440; 0.567], pd = 0.556). The pairwise contrasts confirmed this pattern, although not all of them were robust (Supplementary Table 5).Tom Roth.indd 167 08-01-2024 10:42