Raw Evidence and Synthesis Separation

These examples and illustrative results are independently authored teaching materials, not measured model results.

Use case

Two stakeholders describe one feature’s audience differently. Teaching A says ‘Start with startups; one administrator can configure it.’ B says ‘Start with enterprises; procurement and security teams participate.’ Statements do not automatically establish a settled product position.

Mechanism

Store authorized attributable statements with question/time, then write synthesis separately. Cite sources for commonalities, disagreements, candidates and decisions still needed. Do not edit originals to manufacture agreement. Handle sensitive material under the record policy rather than indefinitely retaining private data for completeness.

Bad example

Replace A/B with 'The audience is enterprise' and delete the startup statement. Treat the confident synthesis as agreed consensus.

Good example

Create Raw and Synthesis sections from A/B. Raw preserves attribution/meaning. Synthesis marks audience unresolved, lists startup and enterprise candidates, and asks which the first version prioritizes and what setup boundary it serves. Cite A/B for interpretations; do not present candidates as decisions or rewrite originals to remove disagreement.

Why the change matters

Mixing evidence and interpretation can turn an inference into something a person supposedly said. Separation lets readers inspect disagreement and see how a later positioning decision changes synthesis.

Observable expectation

An illustrative synthesis retains the audience difference and asks a focused priority question. Claims such as consensus or enterprise target need actual decision evidence.

Revising synthesis leaves Raw meaning/attribution intact. Unsupported conclusions cannot masquerade as interview quotes. No real participant data is included.

Limits

Statements may be mistaken, biased or individual opinions; attribution is not truth. Privacy/storage scope comes first. Candidate counts and Raw/Synthesis organization are workflow choices rather than universal research formats.