Worksheets and research resources · Page 35

Harmonize Measurement Units Without Losing Provenance

Create a traceable conversion map and verify units, scale factors, and summary statistics before combining data.

Record what happened, which protocol rule applied, and how each deviation was assessed without automatically excluding participants.

Write explicit membership and grouping rules so participant counts remain traceable across efficacy, safety, and sensitivity analyses.

Document the paired population before deriving an SD of change.

Record the design and evidence behind a cluster approximation; this is not a power calculation.

Define acceptable disagreement before inspecting paired results; no patient measurements belong in this blank planning file.

Complete a separate copy for each outcome and time point. Use aggregate counts only. Arithmetic balance does not validate the analysis population.

Plan and validate joins with aggregate diagnostics. Keep real identifiers and row-level records in the approved data environment.

Clean inconsistent labels, transform tabular data, and inspect reproducible operation histories with free, open-source desktop software.

Find research data, software, and supporting materials in a public repository. Check each record’s license and file-access conditions before reuse.

Create a free persistent researcher identifier and maintain connections to your research outputs across name or affiliation changes.

Explore a free, open catalog connecting scholarly works, researchers, and institutions. Check source records before relying on automated metadata.