Join Reconciliation Checklist
Use this before and after combining datasets with concat, merge, or join.
Before joining
- [ ] I know what one row represents in the left dataset.
- [ ] I know what one row represents in the right dataset.
- [ ] I know which column or columns are the join keys.
- [ ] I checked whether keys are unique in each dataset.
- [ ] I checked for missing join keys.
- [ ] I checked for duplicate join keys.
- [ ] I know which join type I need: inner, left, right, or outer.
- [ ] I wrote why this join type is appropriate.
Key checks
| Dataset | Key column(s) | Row count | Unique keys | Missing keys | Duplicate keys |
|---|---|---|---|---|---|
| Left | |||||
| Right |
Join decision
Join type:
Reason:
Expected result:
Possible risk:
After joining
| Check | Expected | Actual | Pass? |
|---|---|---|---|
| Row count | |||
| Important columns retained | |||
| Missing matches reviewed | |||
| Duplicate rows reviewed | |||
| Suffix columns understood |
Common warning signs
- The joined table has many more rows than expected.
- The joined table has fewer rows than expected.
- Important columns have many new missing values.
- Duplicate keys create repeated records.
- Columns with similar names now have confusing suffixes.
- The join changes the meaning of one row.
Final explanation
Write two or three sentences explaining why the joined table is trustworthy enough for the next analysis step.