Paper

Khalvati K, Kiani R, Rao RPN. Bayesian inference with incomplete knowledge explains perceptual confidence and its deviations from accuracy. Nature Communications, 2021.

Notes

Source paper for the macaque_rdm_confidence vertical slice's accuracy summary. Underlying behavioral data is from the Kiani-Shadlen 2009 random-dot motion confidence/wagering task with two monkeys (M1, M2). Each accuracy curve is reported separately for trials with no sure-target option ("accuracy_no_sure_target") and trials where the sure-target option was offered but the direction was chosen instead. The no-sure-target rows also have target-coded psychometric and motion-duration chronometric proxy findings for per-monkey baseline model fits; these do not recover raw signed direction or saccade latency.

Curves on shared axes

Findings extracted from this paper, plotted alongside one another.

Compare to the rest of the literature on /findings.

Findings

Finding Curve Species Source level Stratification Points Fits
finding.khalvati-kiani-rao-2021.accuracy.accuracy-no-sure-target.m1 accuracy_by_strength non-human-primate figure-source-data accuracy_no_sure_target 6
  • accuracy-rate-null AIC=43082.3
  • chance-floor-accuracy-logistic AIC=36378.8
finding.khalvati-kiani-rao-2021.accuracy.accuracy-no-sure-target.m2 accuracy_by_strength non-human-primate figure-source-data accuracy_no_sure_target 5
  • accuracy-rate-null AIC=31691.4
  • chance-floor-accuracy-logistic AIC=28646.5
finding.khalvati-kiani-rao-2021.accuracy.accuracy-sure-available-direction-chosen.m1 accuracy_by_strength non-human-primate figure-source-data accuracy_sure_available_direction_chosen 6
  • accuracy-rate-null AIC=21362.6
  • chance-floor-accuracy-logistic AIC=16588.3
finding.khalvati-kiani-rao-2021.accuracy.accuracy-sure-available-direction-chosen.m2 accuracy_by_strength non-human-primate figure-source-data accuracy_sure_available_direction_chosen 5
  • accuracy-rate-null AIC=17464.6
  • chance-floor-accuracy-logistic AIC=15175.6
finding.khalvati-kiani-rao-2021.chronometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m1 chronometric non-human-primate figure-source-data accuracy_no_sure_target 6
  • chronometric-constant-rt AIC=2.0
  • chronometric-hyperbolic-rt AIC=6.0
  • ddm-vanilla AIC=36238.7
finding.khalvati-kiani-rao-2021.chronometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m2 chronometric non-human-primate figure-source-data accuracy_no_sure_target 5
  • chronometric-constant-rt AIC=3.2
  • chronometric-hyperbolic-rt AIC=6.8
  • ddm-vanilla AIC=28686.5
finding.khalvati-kiani-rao-2021.psychometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m1 psychometric non-human-primate figure-source-data accuracy_no_sure_target 6
  • ddm-vanilla AIC=36238.7
  • logistic-4param AIC=36236.9
  • sdt-2afc AIC=36264.9
finding.khalvati-kiani-rao-2021.psychometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m2 psychometric non-human-primate figure-source-data accuracy_no_sure_target 5
  • ddm-vanilla AIC=28686.5
  • logistic-4param AIC=28654.3
  • sdt-2afc AIC=28664.7

Fit parameters

Per-finding fits committed alongside this paper. Parameters, log-likelihood, AIC, and BIC for each variant. Click the finding id for residuals, model-selection winner, and provenance.