EXPERIMENT

Experts and Dispersed Information

Error rates and confidence in expert versus novice judgement

An interactive companion to the EXPERTS project: how mistake rates and confidence diverge between expert and novice judgements under dispersed information.

Experts and Dispersed Information

Individual-level simulation

Top half — individuals. Each member estimates a regression slope β from n noisy points — more data ⇒ smaller SE ⇒ sharper estimate (EXPERT n=80 vs NOVICE n=4). Tune the parameters below; the group dynamics at the bottom show how individual estimates aggregate into a group decision.

Candidate states
Model
x design
press Go
NOVICE · avg error
--
nearest
--
Bayes-optimal
EXPERT · avg error
--
nearest
--
Bayes-optimal
Each cell: err (top) / conf (bottom)

Simulated group dynamics

Bottom half — groups. The same simulation, played out for a handful of true slopes. Each panel draws one fresh group — 1 EXPERT (n=80) plus NOVICES (n=4), group size set below — and shows the group decision that results from pooling their estimates. Press Go above for a new realisation.

scatter, β̂, SE, picks all regenerate on Go from a fresh base seed per (row, member) each Go press = new MC realisation
Group aggregation

A group — 1 EXPERT + NOVICES (size set below) — pools its members' slope estimates into one group decision. The behavioural tile below weights each member by wiτiρ (τ = precision = 1/SE²). Drag ρ to set how much the group trusts precision.

counter-precisionρ = −1 over-trusts noisy members optimal · precisionρ = 1 inverse-variance weighting
ρ = 0.00 equal weighting (⅓ each)
Group size
always 1 EXPERT + 2 NOVICES per group