Professional Detailing2 min read

Coating Trial Data: Separate Zero, Missing and Not Applicable

A recorded zero, a missing observation and a not-applicable field have different meanings. Keep them distinct so that averages, rates and completion checks do not silently treat absent information as a measured outcome.

Updated · NANOLAB Knowledge

The key distinction

A zero can be a valid measurement or count. Missing means the expected information was not obtained or recorded. Not applicable means the field does not belong to that case under the defined protocol. The data dictionary should specify how each status is entered and handled in analysis.

Practical checks

  1. Define permitted values and missing-data codes before collecting results.
  2. Preserve the reason for missing observations where known.
  3. Check calculations to ensure they treat each status according to the analysis plan.

Example

Illustrative example: an unmeasured cleaning time is entered as zero. An average then makes the process appear faster, even though no observation supports that value. A distinct missing status prevents the spreadsheet from turning incomplete work into an apparent performance benefit.

Scope and limits

Do not replace unknown results with favourable values to complete a dashboard. This guide does not prescribe a statistical missing-data method. The analyst should determine the appropriate treatment for the question, while the raw record preserves what was actually observed and what remains unknown.

Related guidance

Ask NANOLAB about the exact product and intended use. Use the current product documentation for formulation-specific instructions.