Universal topic learning · Practical workbook

Data Cleaning Plan Workbook

Tick the checklist and type your answers; they stay in this browser. Print or save as PDF with the button below. The review guide prints on a separate page.

Use fictional information to document the plan and its evidence. Apply your approved procedures for real work.

Practise with fictional information. Use your approved procedures for real work.

Before you begin

Define the task, purpose and boundaries.
Use approved tools and suitable information.
Complete each relevant planning field and mark unknowns.
Test normal, missing, conflicting and denied cases as relevant.
Record evidence, limitations and the review decision.
Name an owner and retest meaningful changes.

Your planning worksheet

Type your answers below. They are saved in this browser only; download or print them when you finish.

Raw file location and untouched copy

Question the clean data must answer

Row count in

Duplicate rule and rows removed

Standard formats for dates, names and units

Missing values and how they are marked

Change table reviewed (old to new)

Row count out and check

Cleaning log, reviewer and date

Practice scenario

A fictional sales export has 1,250 rows. 38 rows repeat an OrderID, one outlet name is spelt three ways and 12 rows have no date. Plan the clean-up before asking AI to do it.

Review guide

Work on a copy; the raw export stays unchanged.
Define a duplicate as the same OrderID before removing anything.
Mark the 12 missing dates as [date?] instead of filling them in.
1,250 in minus 38 removed = 1,212 out; the counts must tie.

Primary references

NIST Generative AI Risk Profile

Project-authored exercises. References support the underlying principles; this workbook does not prescribe company procedures.

Data Cleaning Plan Workbook

Before you begin

Your planning worksheet

Practice scenario

Primary references

Practice scenario: review guide