02 / DATA EXPERIMENTS
Data Quality Lab
Ready to exploreHow do simple cleaning steps change a dataset’s quality?
Updated: 27 September 2026A fictional sample dataset. No files are uploaded and no other project records are changed.
Steps run in order: trim → deduplicate → exclude. Identical rows match across all fields.
Before cleaning
| ID | Name | Score |
|---|---|---|
| S01 | " Demo A " | "91" |
| S02 | "Demo B" | "78" |
| S02 | "Demo B" | "78" |
| S03 | Missing | "84" |
| S04 | "Demo D" | Missing |
| S05 | "Demo E" | "120" |
| S06 | "Demo F" | " 66 " |
After cleaning
| ID | Name | Score |
|---|---|---|
| S01 | " Demo A " | "91" |
| S02 | "Demo B" | "78" |
| S02 | "Demo B" | "78" |
| S03 | Missing | "84" |
| S04 | "Demo D" | Missing |
| S05 | "Demo E" | "120" |
| S06 | "Demo F" | " 66 " |
Retained 7 of 7 records. Missing values and scores have not been guessed or replaced.
What can we observe?
A record may contain several issues, so issue counts should not be added as a record total. Compare each step and its effect on retained data. Exclusion reduces available data and does not guarantee correctness.
Experiment limits
An educational experiment using seven fictional records, not a comprehensive audit or an assessment of real people. There is no file upload or result storage.
Explore the related project ↗No name or email required. Your choice is remembered in this browser.
How is your choice remembered?