A neat spreadsheet can still tell the wrong story
A student measures plant growth for two weeks and later discovers a missing observation. Filling the gap with a guessed number would make the chart look complete but undermine the investigation. Data integrity means preserving what was actually observed and distinguishing it from corrections or estimates.
Keep raw observations separate
Record original measurements, units, dates and conditions. Later calculations and graphs should be traceable to those observations.
Use consistent identifiers
Label specimens, groups and measurement periods so observations are not accidentally assigned to the wrong subject.
Handle missing data honestly
A blank entry should not silently become zero or an invented value. Explain why the data are missing and how this affects interpretation.
Check transcription and calculations
Simple checks can reveal swapped digits, wrong units and spreadsheet formula errors.
Document legitimate corrections
When an error is confirmed, preserve enough context to explain what changed and why. Correction is not the same as falsification.
Protect participants
Surveys and projects involving people should follow school guidance on consent, confidentiality and appropriate data collection.
Break-test: the graph looks too perfect
Can students show the underlying measurements and explain whether any values were removed, rounded or estimated?
Connected knowledge
See School Research Integrity, Data Collection Design and MOE Syllabuses.