Compare only when the evidence is compatible
The comparison workflow checks provenance, methodology and units before calculating differences. Compatible units can be reconciled when a valid conversion exists, while incompatible method inputs can block the comparison or mark it as requiring caution. A baseline can then be selected to show deltas and whether observed differences sit within the recorded scatter.
Keep blocked comparisons visible instead of forcing a chart
When two results should not be compared directly, NotyResults can explain the mismatch rather than presenting a misleading difference chart. This is important when calibration, units, analysis method or other provenance differs between samples.
Turn reviewed results into reusable scientific text
The result workspace can prepare a figure caption, methods paragraph or concise summary from the saved analysis context. AI-assisted explanation can also be requested for an individual result or a comparison, but the underlying values and comparability status remain available for review.
A short workflow
- Open the relevant project or all-project result list and select the saved analysis you need.
- Review metrics, statistics, scope and provenance before interpreting the result.
- For comparisons, select compatible results and read any unit or methodology warnings before using the calculated delta.
- Export or copy the reviewed result summary and keep the saved analysis record as the traceable source.
Output you can keep
Export supported result views as HTML, Markdown or PDF. Use the generated caption, methods text or summary as a draft tied to the saved result, then edit it to match the scientific context and reporting requirements.
Comparison limits
- A valid numerical conversion does not make two different experimental methods scientifically equivalent.
- Results marked not comparable should not be forced into a single quantitative conclusion.
- AI explanation is secondary to the recorded measurements, method context and domain knowledge.
- If the source analysis changes, regenerate or reopen the saved result before reporting an older comparison.