LAB PROTOCOL · EVIDENCE-SELECTION BENCHMARK

Agent Source-Selection Bias: Evidence-Selection Lab

A 20-case protocol for testing whether an agent checks query diversity, source quality, negative evidence, stale pages, ranking bias, and deliberate disconfirmation before committing to an answer.

Protocol brief · July 23, 2026 · Editorial lab-design note

Why this benchmark matters

A fast answer can be wrong before the agent writes its first sentence. This protocol tests whether the investigation searches broadly enough, weighs source quality, notices missing or contradictory evidence, and actively tries to disconfirm its first plausible result.

Scope note: This protocol defines a test design, not a performance claim. Results depend on implementation, model configuration, permissions, source conditions, and the fixtures used.

Measurement matrix

Test surfaceWhat to observeEvidence-led assessment
Query diversityWhether materially different search formulations are attemptedScore coverage across synonyms, constraints, source types, and alternate interpretations.
Source qualityHow primary, current, and authoritative each result isRecord provenance, date, directness, and whether the source actually supports the claim.
Negative evidenceWhether absent, contradictory, or failed results are retainedRequire an explicit gap or conflict note instead of silently discarding inconvenient evidence.
Staleness detectionWhether outdated pages control the conclusionCompare publication and update dates, version context, and current upstream status.
DisconfirmationWhether the agent tests its leading hypothesisAdd a deliberate counter-search and score correction when the first answer is weakened.

Protocol steps

  1. Create 20 fixtures with a known answer, plausible distractor, stale page, and at least one negative or contradictory signal.
  2. Vary query wording, result ordering, source availability, and the position of the strongest evidence.
  3. Score source coverage, evidence quality, contradiction handling, correction rate, and unsupported certainty.
  4. Inspect the final rationale for omitted searches, unacknowledged gaps, and claims that outrun the collected evidence.
  5. Publish aggregate results with fixture versions, source timestamps, and clear limits on generalization.

Lab disclosure

This page was developed from a July 23, 2026 lab-bench brief supplied by the Content Site Consultant. It is a proposed evaluation protocol, not an independent benchmark result or product certification.