Aggregate missingness-pattern audit
Method review date: 2026-11-03

Instructions: Save a copy, complete blanks and record unresolved questions. Use fictional or aggregate values only; never patient identifiers. This planning aid does not certify an analysis.

DEFINE THE COMMON DATASET
Population and observation unit: ____
Dataset/query/version: ____
Variables in fixed order (up to 8 for linked tool): ____
Codes treated as missing and reason: ____
Structural not-applicable values considered separately? ____

TABULATE EXCLUSIVE PATTERNS
Use 1=missing and 0=observed in this worksheet/tool. Other software may reverse this.
Pattern | aggregate positive count | source check
____ | ____ | ____
____ | ____ | ____
____ | ____ | ____
Every observation contributes to exactly one pattern. Combine duplicate patterns; omit zero-count patterns.

RECONCILE OVERLAPPING DENOMINATORS
Sum of pattern counts/total: ____
All-zero pattern/complete across selected variables: ____
Any missing=total−complete: ____
Marginal missing for each variable: ____
Do not sum marginal missing counts to estimate people/records with any missing.
Available count for each variable pair (both zero): ____

CONNECT TO EACH ANALYSIS
Model/analysis required-variable set: ____
Complete count for that set: ____
Rows lost when adding a covariate: ____
Whether comparison of models also changes the sample: ____
Patterns do not establish MCAR/MAR/MNAR or justify complete-case analysis.
Reasons, assumptions, sensitivity analysis and independent check: ____

Practical guide: https://www.statmedanalysis.com/blog/missingness-pattern-audit
Project/version: ____  Completed by role/date: ____