Skip to content

Investigate unusual cases without deleting inconvenient evidence

Decide whether an extreme public account is a data error, a different condition or a legitimate exception before summarizing the reviewed material.

Guides

What should happen to a strikingly unusual research case?

A public discussion about electric kettle durability may contain one account of immediate failure among several accounts of long use. That difference could reflect a defective unit, a different model or incomplete context. An outlier review should investigate the distinction without deleting the case merely because it complicates a smooth summary of the evidence.

Example research prompt
Review the unusually early kettle failure in the supplied set. Check explicit model and task details, preserve the case and show how it affects the descriptive summary.
01 / Guides

Identify the unusual feature

State what is unusual relative to the reviewed set, such as reported failure on the first use. Include the original claim and any explicit model details. Exclude assumptions that the author is mistaken simply because other accounts differ; rarity within a selected corpus is not evidence that a report is false.

02 / Guides

Check comparability and errors

Inspect whether the case concerns the same model, task and time basis as the other accounts. Correct a clear transcription error while retaining an audit trail. If a key detail is missing, mark comparability uncertain rather than inventing a value or assigning the case to whichever group minimizes its effect on the summary.

03 / Guides

Review the mechanism separately

Ask whether the account suggests a distinct condition worth investigating, such as an explicitly described electrical fault. Keep the claim attributed and avoid diagnosing a cause beyond the text. A vivid single incident can motivate follow-up, but it does not establish a typical experience or justify a broad safety conclusion without appropriate evidence.

04 / Guides

Deliver inclusive summaries

Present the main pattern with the unusual case and explain any justified exclusion from a particular comparison. If the result changes substantially when the case is omitted, show that sensitivity rather than hiding it. Leave unresolved differences visible and recommend the next evidence needed to distinguish error, condition difference and genuine exception.

Keep the boundaries clear.

  • An unusual account is a reason to inspect comparability, not an automatic reason to discard it.
  • Examples describe review methods, not live tests or guaranteed platform coverage.

Start with a read-only check: ask your agent to report actual source coverage before collecting records or exporting an audience.

Primary references

Documentation reviewed 2026-11-13. Product capabilities, client availability and plans can change. Third-party marks identify their products, not an affiliation.

Know what social media shows about you.
Start with a question.

Get started
montage