Simulated data show a double dissociation: early-trial reveal effects arise only from non-informed advice-taking, while outcome-dependent coherence repetition arises only from informed advice-taking¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r2 · q2
Currency · newest study 2024 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Empirical data showed a substantial early-trial reveal effect (median = 0.64) mimicked by non-informed but not informed simulated data, and an outcome × advice-presentation interaction (median = 0.05) mimicked by informed but not non-informed simulated data. → Maayan Pereg 2024 q2 i? An outcome × advice-presentation interaction in coherence repetition (median = 0.05) appeared in empirical data and was mimicked only by the informed advice-taking simulated dataset. → Maayan Pereg 2024 (2)

Evidence¶

Maayan Pereg 2024¶

Maayan Pereg, Uri Hertz, Ido Ben-Artzi and Nitzan Shahar. (2024). Disentangling the contribution of individual and social learning processes in human advice-taking behavior. npj Science of Learning 9:4. https://doi.org/10.1038/s41539-024-00214-0

q2 · i? · causal · r2

Simulation analyses using participants' estimated parameters with φ = 0 (informed only) or ω = 1 (non-informed only) compared against empirical data; the early-trial reveal effect (median 0.64) appeared in the non-informed artificial dataset (median 0.20) but not the informed one (median −0.03).

"For the empirical data, we found a substantial reveal effect (median = 0.64, CI89% between 0.555 to 0.73, pd ~100%"

Maayan Pereg 2024 (2)¶

Maayan Pereg, Uri Hertz, Ido Ben-Artzi and Nitzan Shahar. (2024). Disentangling the contribution of individual and social learning processes in human advice-taking behavior. npj Science of Learning 9:4. https://doi.org/10.1038/s41539-024-00214-0

q2 · i? · causal · r2

Bayesian logistic regression of coherence repeat on previous outcome × advice presentation in the empirical data showed a positive interaction (median 0.05), mimicked by the informed artificial dataset (median 0.03) but not the non-informed one (median 0.01, pd = 91.55%).

"For the empirical data, we found evidence in favor of a previous outcome X advice presentation paired interaction (median = 0.05, CI89% = 0.03–0.07, pd~100%"

Discussion¶

Related Claims¶

• Advice framed as social information holds greater causal influence than an identical non-social cue (a lottery wheel) — related
• A computational model combining informed and non-informed advice-taking with individual learning (Model 5) best predicts participants' choices — related
• Revealing a teacher's advice causally increases participants' tendency to choose the advised card in a reinforcement learning task — related