The evidence behind context testing

The same question can draw on different sources.

The Disclosure Effect informs why CommsRx tests the context around an audience’s question.

What the study found

Added patient or caregiver context crossed the study’s source-change threshold in 105 of 357 eligible question–system comparisons (29.4%).

120 authored health questions; three search-enabled API configurations; three repetitions per condition.
2,149 captured responses from 2,160 planned records.
Collection: 13 September 2026. Revised study record: 14 September 2026. Public technical summary: 21 September 2026.

What that means for communications

Test the question, the context and the configured system together. Then use communications judgment to decide which evidence gaps and source priorities matter. This is the approach the research informs; it is not a demonstrated CommsRx performance result.

Read the limits with the result

This descriptive preview is not peer reviewed. Its prespecified decision-stage hypothesis and broad higher-authority-source hypothesis were not supported. It did not test medical accuracy, patient behavior, commercial outcomes or a communications intervention. Added-context bundles varied together, so individual causes were not isolated.

The technical record describes corrections, analysis deviations and an unresolved archive-checksum discrepancy. The result is not a guaranteed ranking, visibility or revenue improvement.

Public sources

Research by Signal Fidelity Group, developer of CommsRx. Commercial interpretations are disclosed as SFG’s interpretation. This page links public records; private raw research files are not included. Summary prepared 9 October 2026.