Conservatism Bias¶

Capture the tendency of people to revise a probability judgment less than Bayes' rule prescribes when evidence arrives — the reported posterior landing short of the correct one, anchored too close to the prior — measured as a signed gap against an explicit Bayesian benchmark.

Core Idea¶

Conservatism bias is the systematic tendency of reasoners to revise probability judgments less than Bayes' rule prescribes: the reported posterior lies between the prior and the correct Bayesian value, too close to the prior. Edwards (1968) found evidence that should shift a 0.50 prior to ~0.97 typically produced only ~0.70. The mechanism is anchor-and-insufficient-adjustment, strongest for abstract, sequential, weakly narrative evidence. It is opposite in sign to representativeness-driven over-updating, and it compounds across an evidence stream.

Scope of Application¶

Conservatism bias lives in one domain — human probabilistic belief updating against a Bayesian benchmark — the contexts below being application settings of that single substrate, not structurally distinct systems.

Judgment-under-uncertainty research — the canonical bookbag-and-poker-chip paradigm.
Behavioral finance — post-earnings-announcement drift, surprises incorporated gradually.
Forecasting and intelligence analysis — slow revision when signals contradict priors.
Clinical diagnosis — diagnostic momentum, an impression held past the likelihood ratios.

Clarity¶

The clarifying move makes "under-revision" a measurable claim, at the price of an explicit normative benchmark: the bias is defined only against Bayes' rule, forcing the analyst to compute what the posterior should have been and read the judgment as a point short of it. This converts a vague worry that someone is "stubborn" into a signed, quantifiable gap located in the magnitude of the update — distinguishing it from confirmation bias — and, held against representativeness, sharpens the question from "is judgment biased?" to "which regime, and which way does the error point?"

Manages Complexity¶

Belief-revision errors otherwise present as an unruly catalog — stubbornness, sluggish markets, miscalibrated forecasts — each inviting an ad hoc account. Anchored to one Bayesian benchmark, conservatism collapses that catalog to a single signed scalar: the gap between observed and prescribed posterior, its sign reading off direction. Two parameters then predict and locate the error — the regime, which fixes the sign, and the per-step attenuation, whose compounding over a stream explains cumulative miscalibration a single judgment would not.

Abstract Reasoning¶

The bias licenses moves organized around a signed distance from the Bayesian benchmark: a diagnostic move computing prior, likelihood ratio, and implied posterior to localize the error in update magnitude; a regime-classification move inferring the sign before any number is computed; interventionist moves that force the full-size update and re-aggregate sequential ratios to expose compounded drift; and a boundary-drawing move fixing that the concept is meaningless outside a prior, a likelihood ratio, and a computable normative posterior.

Knowledge Transfer¶

Within human probabilistic belief updating the bias transfers as mechanism, intact — though its "domains" are application contexts of one substrate, not distinct systems. The signed-distance diagnosis and its corrective family (compute the prior and likelihood ratio, force the full update; reference-class forecasting; scoring-rule feedback) carry across judgment research, behavioral finance, intelligence analysis, and clinical diagnosis. Beyond the human reasoner the bias does not travel — a non-human system implements Bayes or departs from it by design, not psychology. What recurs is the parent pattern under-relaxation / inertia in updating (Kalman filters, regularized models, institutional wisdom), with Bayes as the benchmark; carry that, not the named bias.

Relationships to Other Abstractions¶

Current abstraction Conservatism Bias Domain-specific

Parents (3) — more general patterns this builds on

Conservatism Bias is a kind of Bias Prime

Conservatism bias is bias specialized to probability updates that remain systematically too near the prior and short of the Bayesian posterior.
Conservatism Bias is part of, typical Anchoring Prime

Conservatism bias typically contains anchoring when the prior acts as the starting value from which admitted likelihood evidence produces insufficient adjustment.
Conservatism Bias presupposes Bayesian Updating Prime

Conservatism bias presupposes Bayesian updating because its defining quantity is the shortfall between a reported posterior and the Bayes-prescribed one.

Neighborhood in Abstraction Space¶

Conservatism Bias sits in a moderately populated region (47^th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.

Family — Unclustered & Miscellaneous (309 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12