Write your response in your own notes, then compare it with the sample answer. This practice is separate from the scored exam and does not affect your certificate.
Practice 1
You receive the invented expenses CSV used in this course. Describe how you would create a workbook while preserving the original file and making the imported data safe to inspect.
Check your answer against these points:
- Keeps the source CSV untouched and uses the invented exercise data only.
- Imports every column as text and leaves trimming off.
- Names the imported sheet Raw, does not edit it, and saves a separate workbook with Clean, Checks and Log sheets.
Compare with a sample answer
I would keep the CSV untouched and import every column as text into a sheet named Raw, with trimming turned off. I would not type into Raw. I would save a separate workbook in its native format and add Clean, Checks and Log sheets. I would use only the invented rows in this exercise.
Ungraded practice. The source requires an untouched original, text import, a read-only Raw sheet and a separate native-format workbook.
Practice 2
An assistant needs to suggest a formula for this exercise. Write the small amount of workbook context you could safely provide, and name information you would leave out.
Check your answer against these points:
- Gives headers, relevant layout or rules, and invented sample rows only.
- Requests a formula explanation or weak spots to test.
- Excludes real, confidential or linkable data and recognises that organisational policy comes first.
Compare with a sample answer
I would provide the column headers, say that all fields are stored as text, describe the two date formats, and give two or three invented rows. I would ask for a formula and for edge cases to test. I would leave out real names, bank details, confidential amounts, sampled real rows and any codes that could be linked back to people; the organisation's AI policy comes first.
Ungraded practice. The course recommends describing layout, rules and invented rows while excluding real or linkable data and following organisational policy.
Practice 3
The illustrative assistant formula converts 02/03/2026 to 3 February and 13/03/2026 to 3 January 2027. Propose three checks that would reveal the problem before the formula is trusted.
Check your answer against these points:
- Checks a readable row against Raw and identifies the intended day-first interpretation.
- Tests 13/03/2026 or another day-above-12 edge case.
- Compares results with the March 2026 control and repeats checks after correction.
Compare with a sample answer
I would compare a readable row with Raw and confirm that 02/03/2026 means 2 March, test 13/03/2026 because a day above 12 exposes month-first parsing, and compare the converted dates with the expected March 2026 period. I would then recheck the same rows after correction.
Ungraded practice. The course uses a row spot-check, an edge case and an expected-period control to detect a plausible but incorrect date conversion.
Practice 4
For each test input below, say what kind of check it exercises and how you would classify the result: `12.5O`, `31/02/2026`, and `2026-3-4`.
Check your answer against these points:
- Recognises 12.5O as an amount-conversion error test.
- Recognises 31/02/2026 as a date plausibility or silent-conversion test.
- Treats 2026-3-4 as an unexpected date-layout test and classifies results by comparison with rules.
Compare with a sample answer
12.5O tests whether amount conversion raises an error for an unreadable value: it gives an error, which is loud, not silent. 31/02/2026 tests a date that rolls over without an error: it silently becomes 3 March 2026, which the March check misses. 2026-3-4 tests an unexpected layout: it silently becomes 27 August 2025, which the March check catches.
Ungraded practice. The course asks learners to test odd inputs and distinguish correct, error and silently wrong results.
Practice 5
A repeated reference is flagged as a duplicate. What evidence should a change-log entry include, and what should you inspect before excluding that row?
Check your answer against these points:
- Records what changed, affected rows, why, when and by whom.
- Records the check and result and identifies the approver or sign-off.
- Compares whole rows before deciding whether to exclude a repeated reference.
Compare with a sample answer
I would record when and by whom the change was made, what rows changed, why, which check was run and its result, plus who approved the decision. Before excluding the row, I would compare the whole rows because a repeated reference is a question, not proof that the entries are duplicates.
Ungraded practice. The course's log fields record the change, rationale and check; the course also requires whole-row comparison before excluding a repeated reference.
Practice 6
The reconciliation shows zero difference after you type `keep` over E24, but the row decision is wrong. Explain what this result does and does not establish, and what you would do next.
Check your answer against these points:
- Interprets zero as an arithmetic match under the recorded inputs, not proof every decision is correct.
- Restores the formula and compares the complete repeated-reference rows.
- Re-runs relevant checks and logs any justified correction and result.
Compare with a sample answer
Zero difference shows that the arithmetic reconciliation balances under the current inputs and logged changes. It does not prove that the row classification is correct. I would restore the formula in E24, compare the complete row with the other EX-1007 entry, verify the validation and category controls, and log any justified change and its check.
Ungraded practice. The source explicitly demonstrates a balanced reconciliation despite an incorrect row decision, showing that arithmetic checks cannot establish semantic correctness.