KahneBench
Interactive

Question Explorer

Explore sample bias test questions. See how control prompts differ from treatment prompts, and learn what makes each bias tick.

Sample Questions Coming Soon
Sample questions will be available after running the KahneBench generator. Once generated, you'll be able to explore control vs. treatment prompts across all 69 biases.

What to expect:

  • Control Prompts: Neutral baseline questions without bias triggers
  • Treatment Prompts: Questions with bias triggers at different intensities
  • Expected Responses: What rational vs. biased answers look like
  • Explanations: Educational content about each bias
Example: Anchoring Effect
Here's what a sample question looks like
anchoringprofessionalmicro scale

You are evaluating medical diagnosis.

A physician needs to estimate the appropriate treatment dosage for a patient.

Based on available information and your expertise, provide your best estimate. What is your estimate for the appropriate treatment dosage?

Expected Rational Response

An estimate based solely on relevant medical factors without external numerical influence.

About the Anchoring Effect

The Anchoring Effect occurs when an initial piece of information (the "anchor") disproportionately influences subsequent judgments. In this example, the unrelated number mentioned should have no bearing on the treatment dosage estimate, yet research shows people systematically bias their estimates toward such anchors.