Apps

🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!

Variance Neglect

Also Known As: Mean-only reasoning Distributional neglect
Aspect ID: variance_neglect

Definition

Variance neglect is the tendency to focus on mean expected outcomes while ignoring variability, spread, and tail risk. Two distributions with identical means can have dramatically different risk profiles. Variance neglect leads to systematic underestimation of risk and misallocation of resources, especially in policy and investment decisions.

Examples

Two medical treatments have the same expected recovery time (30 days). Treatment A has low variance: nearly all patients recover between 25 and 35 days. Treatment B has high variance: 50% recover in 5 days, 50% require 55 days or more. For a patient who cannot tolerate delay, these treatments are not equivalent.

An investment advisor compares two portfolios, both with an average annual return of 7%. Portfolio A steadily returns between 5–9% each year. Portfolio B alternates between +40% and −20% years. A retiree who needs to withdraw funds annually could be ruined by Portfolio B's variance even though the long-run average looks identical.

A city planner evaluates two bus routes that both average 20-minute journey times. Route A consistently takes 18–22 minutes. Route B averages 20 minutes but ranges from 5 to 55 minutes depending on traffic. Focusing only on the mean, the planner treats them as equivalent — ignoring that commuters on Route B routinely miss connections and appointments.

Verification Steps
Verification Steps
Binary yes/no questions that an AI must answer to detect a reasoning pattern in a text.
Each of the 452 aspects has verification steps — simple yes/no questions designed to systematically detect whether a pattern appears in a text. For ad hominem: "Does the argument attack a person rather than their claim?" For false dichotomy: "Are only two options presented when more exist?" This ensures consistent, reproducible analysis.

Binary (yes/no) questions an LLM must answer to identify this aspect:

  1. 1

    Does the argument focus exclusively on mean outcomes while ignoring variance?

    Type: binary
  2. 2

    Are distributions with identical means but different variances treated as equivalent?

    Type: binary
  3. 3

    Is the probability of tail outcomes (extreme events) being ignored in the analysis?

    Type: binary
  4. 4

    Would a risk-averse agent treat the two options differently despite equal means?

    Type: binary
Deep Dive
The expandable detail section on each aspect page with examples, psychology, and counter-strategies.
The Deep Dive section provides in-depth information about each aspect: a real-world example showing the pattern in action, an explanation of why it works psychologically, practical advice on how to counter it, alternative names, and links to related aspects.