Why the original observation matters
A student corrects a spreadsheet value after noticing a transcription error. The corrected result may be valid, but a reader should still understand what changed and why. An audit trail preserves the story of the evidence.
Keep original observations
Record measurements, dates, units and relevant conditions before analysis. Avoid overwriting raw notes with polished results.
Document corrections
Explain the previous value, corrected value and reason when a meaningful change is made.
Separate raw and processed data
Averages, graphs and derived variables should be traceable to the underlying observations.
Record methods and versions
Changes to calculations or experimental procedures can alter interpretation. Keep relevant context.
Use proportionate tools
A simple dated research notebook may be sufficient for a school project. Complexity should serve learning, not bureaucracy.
Protect privacy
Where projects involve people, record only appropriate information and follow school guidance.
Break-test: a group member changes results without explanation
Can the team reconstruct the original data and assess whether the conclusion remains valid?
Connected knowledge
See Research Data Validation, Research Integrity and MOE Syllabuses.