research-claim-map
$
npx mdskill add lyndonkl/claude/research-claim-mapTriangulates sources to rate evidence and prevent bias.
- Verifies claims before decisions or fact-checking statements.
- Uses search tools to gather supporting and contradicting evidence.
- Rates evidence quality as primary, secondary, or tertiary.
- Delivers confidence-rated conclusions with identified limitations.
SKILL.md
.github/skills/research-claim-mapView on GitHub ↗
---
name: research-claim-map
description: Systematically evaluates claims by triangulating sources, rating evidence quality (primary/secondary/tertiary), assessing source credibility, and reaching confidence-rated conclusions to prevent confirmation bias and reliance on unreliable sources. Use when verifying claims before decisions, fact-checking statements, conducting due diligence, evaluating conflicting evidence, or when user mentions "fact-check", "verify this", "is this true", "evaluate sources", "conflicting evidence", or "due diligence".
---
# Research Claim Map
## Table of Contents
1. [Workflow](#workflow)
2. [Evidence Quality Framework](#evidence-quality-framework)
3. [Source Credibility Assessment](#source-credibility-assessment)
4. [Common Patterns](#common-patterns)
5. [Guardrails](#guardrails)
6. [Quick Reference](#quick-reference)
## Workflow
Copy this checklist and track your progress:
```
Research Claim Map Progress:
- [ ] Step 1: Define the claim precisely
- [ ] Step 2: Gather and categorize evidence
- [ ] Step 3: Rate evidence quality and source credibility
- [ ] Step 4: Identify limitations and gaps
- [ ] Step 5: Draw evidence-based conclusion
```
**Step 1: Define the claim precisely**
Restate the claim as a specific, testable assertion. Avoid vague language - use numbers, dates, and clear terms. See [Common Patterns](#common-patterns) for claim reformulation examples.
**Step 2: Gather and categorize evidence**
Collect sources supporting and contradicting the claim. Organize into "Evidence For" and "Evidence Against". For straightforward verification → Use [resources/template.md](resources/template.md). For complex multi-source investigations → Study [resources/methodology.md](resources/methodology.md).
**Step 3: Rate evidence quality and source credibility**
Apply [Evidence Quality Framework](#evidence-quality-framework) to rate each source (primary/secondary/tertiary). Apply [Source Credibility Assessment](#source-credibility-assessment) to evaluate expertise, bias, and track record.
**Step 4: Identify limitations and gaps**
Document what's unknown, what assumptions were made, and where evidence is weak or missing. See [resources/methodology.md](resources/methodology.md) for gap analysis techniques.
**Step 5: Draw evidence-based conclusion**
Synthesize findings into confidence level (0-100%) and actionable recommendation (believe/skeptical/reject claim). Self-check using `resources/evaluators/rubric_research_claim_map.json` before delivering. Minimum standard: Average score ≥ 3.5.
## Evidence Quality Framework
**Rating scale:**
**Primary Evidence (Strongest):**
- Direct observation or measurement
- Original data or records
- First-hand accounts from participants
- Raw datasets, transaction logs
- Example: Sales database showing 10,000 customer IDs
**Secondary Evidence (Medium):**
- Analysis or interpretation of primary sources
- Expert synthesis of multiple primary sources
- Peer-reviewed research papers
- Verified news reporting with primary source citations
- Example: Industry analyst report analyzing public filings
**Tertiary Evidence (Weakest):**
- Summaries of secondary sources
- Textbooks, encyclopedias, Wikipedia
- Press releases, marketing materials
- Anecdotal reports without verification
- Example: Company blog post claiming customer count
**Non-Evidence (Unreliable):**
- Unverified social media posts
- Anonymous claims
- "Experts say" without attribution
- Circular references (A cites B, B cites A)
- Example: Viral tweet with no source
## Source Credibility Assessment
**Evaluate each source on:**
**Expertise (Does source have relevant knowledge?):**
- High: Domain expert with credentials, track record
- Medium: Knowledgeable but not specialist
- Low: No demonstrated expertise
**Independence (Is source biased or conflicted?):**
- High: Independent, no financial/personal stake
- Medium: Some potential bias, disclosed
- Low: Direct financial interest, undisclosed conflicts
**Track Record (Has source been accurate before?):**
- High: Consistent accuracy, corrections when wrong
- Medium: Mixed record or unknown history
- Low: History of errors, retractions, unreliability
**Methodology (How did source obtain information?):**
- High: Transparent, replicable, rigorous
- Medium: Some methodology disclosed
- Low: Opaque, unverifiable, cherry-picked
## Common Patterns
**Pattern 1: Vendor Claim Verification**
- **Claim type**: Product performance, customer count, ROI
- **Approach**: Seek independent verification (analysts, customers), test claims yourself
- **Red flags**: Only vendor sources, vague metrics, "up to X%" ranges
**Pattern 2: Academic Literature Review**
- **Claim type**: Research findings, causal claims
- **Approach**: Check for replication studies, meta-analyses, competing explanations
- **Red flags**: Single study, small sample, conflicts of interest, p-hacking
**Pattern 3: News Fact-Checking**
- **Claim type**: Events, statistics, quotes
- **Approach**: Trace to primary source, check multiple outlets, verify context
- **Red flags**: Anonymous sources, circular reporting, sensational framing
**Pattern 4: Statistical Claims**
- **Claim type**: Percentages, trends, correlations
- **Approach**: Check methodology, sample size, base rates, confidence intervals
- **Red flags**: Cherry-picked timeframes, denominator unclear, correlation ≠ causation
## Guardrails
**Avoid common biases:**
- **Confirmation bias**: Actively seek evidence against your hypothesis
- **Authority bias**: Don't accept claims just because source is prestigious
- **Recency bias**: Older evidence can be more reliable than latest claims
- **Availability bias**: Vivid anecdotes ≠ representative data
**Quality standards:**
- Rate confidence numerically (0-100%), not vague terms ("probably", "likely")
- Document all assumptions explicitly
- Distinguish "no evidence found" from "evidence of absence"
- Update conclusions as new evidence emerges
- Flag when evidence quality is insufficient for confident conclusion
**Ethical considerations:**
- Respect source privacy and attribution
- Avoid cherry-picking evidence to support desired conclusion
- Acknowledge limitations and uncertainties
- Correct errors promptly when found
## Quick Reference
**Resources:**
- **Quick verification**: [resources/template.md](resources/template.md)
- **Complex investigations**: [resources/methodology.md](resources/methodology.md)
- **Quality rubric**: `resources/evaluators/rubric_research_claim_map.json`
**Evidence hierarchy**: Primary > Secondary > Tertiary
**Credibility factors**: Expertise + Independence + Track Record + Methodology
**Confidence calibration**:
- 90-100%: Near certain, multiple primary sources, high credibility
- 70-89%: Confident, strong secondary sources, some limitations
- 50-69%: Uncertain, conflicting evidence or weak sources
- 30-49%: Skeptical, more evidence against than for
- 0-29%: Likely false, strong evidence against
More from lyndonkl/claude
- abstraction-concrete-examplesBuilds structured abstraction ladders that translate high-level principles into concrete, actionable examples across 3-5 levels. Bridges communication gaps, reveals hidden assumptions, and tests whether abstract ideas work in practice. Use when explaining concepts at different expertise levels, moving between abstract principles and concrete implementation, identifying edge cases by testing ideas against scenarios, designing layered documentation, decomposing complex problems into actionable steps, or bridging strategy-execution gaps.
- academic-letter-architectGuides the creation of evidence-based academic recommendation letters, reference letters, and award nominations that combine concrete examples, meaningful comparisons, and genuine enthusiasm. Use when writing recommendation letters for students, postdocs, or colleagues, or when user mentions recommendation letter, reference, nomination, letter of support, endorsement, or needs help with strong advocacy and comparative statements.
- adr-architectureDocuments significant architectural and technical decisions with full context, alternatives considered, trade-offs analyzed, and consequences understood. Creates a decision trail that helps teams understand why decisions were made. Use when choosing between technology options, making infrastructure decisions, establishing standards, migrating systems, or when user mentions ADR, architecture decision, technical decision record, or decision documentation.
- adverse-selection-priorProduces a Bayesian prior probability that an offered transaction is +EV for the recipient, given that the counterparty chose to propose it. Applies Akerlof market-for-lemons logic -- if they offered it, they believe it is +EV for them, so the prior that it is +EV for us is materially below 50%. Reusable across trade evaluation, waiver drops (another team dropping a player is also adverse selection), job-offer analysis, M&A, and any "someone offered me this" situation. Use when you receive an unsolicited trade/offer/proposal, analyzing incoming trade prior, evaluating why a counterparty proposed a deal, or when user mentions adverse selection, market for lemons, why did they offer this, incoming trade prior, they proposed it, Bayesian adjustment on received offer.
- alignment-values-north-starCreates actionable alignment frameworks that give teams a shared North Star (direction), values (guardrails), and decision tenets (behavioral standards). Enables autonomous decision-making while maintaining organizational coherence. Use when starting new teams, scaling organizations, defining culture, establishing product vision, resolving misalignment, creating strategic clarity, or when user mentions North Star, team values, mission, principles, guardrails, decision framework, or cultural alignment.
- analogy-weight-checkFor every analogy in a substacker draft, verifies it carries mechanical weight — the analogy does real work explaining the mechanism, not merely decorates it. Cross-references analogy-catalog.md for novelty (is this analogy reused from a prior post?) and domain fit (biology > organizational > sports preferred; physics/military disfavored). Use whenever an analogy appears in the draft. Trigger keywords: analogy weight, decorative, mechanical weight, reused analogy, catalog check, metaphor check.
- answer-uncomfortable-questionTakes one strategic question about substacker ("should we launch paid?", "is this section dead?", "are we writing for the wrong audience?") and produces the mandatory evidence + reasoning + downside triad plus a recommendation. Used 3 times per Growth Strategist review. Trigger keywords: uncomfortable question, strategic question, evidence reasoning downside, triad.
- attribute-performanceFor each substacker post that materially over- or under-performs the rolling baseline (|z| ≥ 1.0), produces a plain-English attribution paragraph with calibrated confidence (high / medium / low / unexplained). Considers subject-line effect, topic zeitgeist, external share, day-of-week, length effect, and audience-notes signals. Labels unexplained outliers explicitly rather than fabricating a story. Use after compute-baseline when outlier posts exist. Trigger keywords: attribution, why did this post work, outlier explanation, performance analysis.
- auction-first-price-shadingComputes the optimal shaded bid for a first-price sealed-bid auction given a true private value, an estimate of the number of competing bidders N, and a value-distribution assumption. Implements the `(N-1)/N` equilibrium shading rule for uniform private values, adjusts for log-normal or empirical value distributions, layers a risk-aversion adjustment, and caps output against the bidder's remaining budget. Domain-neutral auction theory reusable across fantasy sports (baseball FAAB, NBA/NHL waiver auctions), prediction-market limit sizing, sealed procurement bids, and any blind-bid context. Use when user mentions "first-price auction bid", "sealed bid shading", "(N-1)/N", "FAAB bid amount", "auction shading", "optimal bid first-price", "bid for sealed-bid", "blind bid sizing", or when downstream logic needs a principled shade factor rather than an ad-hoc heuristic.
- auction-winners-curse-haircutApplies a Bayesian haircut to a bid valuation for common-value auctions where winning is itself evidence the bidder over-estimated. Takes a raw valuation, a value-type classification (common_value / private_value / mixed), the number of informed bidders N, and a signal-dispersion estimate, and returns an adjusted valuation. Domain-neutral and reusable across fantasy FAAB, prediction markets, M&A bids, ad-auction budgets, and any generic bidding context. Use when user mentions "winner's curse", "common value auction", "valuation haircut", "adverse valuation", "Bayesian bid adjustment", or "over-paying in auction".