Evaluate competing hypotheses on complex research questions

Weigh rival explanations systematically against source evidence using confidence levels and counter-arguments.

You getHypotheses Defined, evidence Matrix and diagnostic Analysis

Prompt

Prompt
Act as a principal intelligence analyst. Using the Analysis of Competing Hypotheses methodology, evaluate the pasted research puzzle across at least three distinct explanatory hypotheses.

Rules:
- Formulate three mutually exclusive, plausible hypotheses to explain the phenomenon.
- Evaluate every piece of available evidence for consistency or inconsistency against each hypothesis.
- Assign a confidence rating (Low, Medium, High) to each hypothesis based solely on empirical evidence, not probability of occurrence.
- Emphasize diagnostic evidence that eliminates a hypothesis over evidence that merely confirms one.

Output format:
- Hypotheses Defined: H1, H2, and H3 stated in one concise sentence each.
- Evidence Matrix: Table with columns Evidence Item, H1 (Consistent/Inconsistent), H2, and H3.
- Diagnostic Analysis: Assessment of which hypothesis has the least disconfirming evidence.
- High-Impact Information Gaps: Two specific data points that would decisively disprove the leading hypothesis.

Here is the research problem and evidence:

Example output

Example for: unexplained traffic drop to high-performing product checkout pages

Hypotheses Defined
  • H1: An unannounced algorithmic search engine update reduced discoverability of specific landing page categories.
  • H2: A frontend script defect caused checkout page tracking tags to drop analytics events without losing real users.
  • H3: A new direct competitor launched an aggressive performance marketing campaign bidding on core brand keywords.
Evidence Matrix
Evidence ItemH1 (Algorithmic Update)H2 (Script Defect)H3 (Competitor Bidding)
Organic search sessions fell 38% overnightConsistentInconsistentInconsistent
Raw server logs show completed orders unchangedInconsistentConsistentInconsistent
Core brand impressions in search console remained flatInconsistentInconsistentInconsistent
Tag management deployment occurred 2 hours before the dropInconsistentConsistentInconsistent
Diagnostic Analysis

H2 (Script Defect) is the leading explanation because it contains the least disconfirming evidence. While H1 accounts for the drop in reported sessions, it directly conflicts with raw web server logs, which show stable conversion volumes and consistent payment gateway traffic. H2 fully reconciles the discrepancy between declining analytics pageviews and uninterrupted transactional volume.

High-Impact Information Gaps
  1. An end-to-end network payload trace on the client checkout flow to confirm whether the tracking script fires successfully on mobile browsers.
  2. Payment gateway transaction timestamps matched against Google Analytics session IDs during the drop window.
Written by AI from this prompt. Your results will vary.

Paste your data points, operational problem, or observed anomalies directly below the prompt in Gemini, ChatGPT, or Claude. To go further, ask the model to outline an experiment to collect the two identified information gaps.

Make it repeatable

Keep using this prompt without pasting it in every time.

Keep exploring

More in Research