Data Quality Decisions: Five Practical Questions

Check your reporting judgement

Choose one answer for each question, then select “Check my answers”. You can change answers and try again. The fictional examples focus on practical decisions before reporting. Answers are processed in your browser; this resource does not submit your choices or collect personal details.

The Government Data Quality Framework identifies several dimensions of quality, including completeness, uniqueness, consistency, timeliness, validity and accuracy. These questions apply selected concepts through original RDE scenarios. A field can be complete while still containing an incorrect value.

Text alternative and answer explanations

An enquiry table has 100 records. Twenty residence fields are blank. What should you do before a country chart?

1. Guess from the telephone code.
2. Show unknown residence and explain missing data.
3. Remove blank records without mentioning them.

Suggested answer: Show unknown residence and explain missing data.. A telephone country code does not establish residence. Keep missing residence visible and describe its effect on the chart.

The same submission identifier appears twice after an import. What is the next step?

1. Count both immediately.
2. Delete every record with that identifier.
3. Investigate the duplicate and apply an agreed rule.

Suggested answer: Investigate the duplicate and apply an agreed rule.. Check whether the import duplicated one submission. A documented rule protects the original record and makes the adjustment traceable.

One file uses 03/04/2026, but its date convention is not documented. What should you do?

1. Confirm the convention before combining files.
2. Assume all sources use your preferred convention.
3. Treat it as text and ignore the reporting period.

Suggested answer: Confirm the convention before combining files.. Confirm whether the source means 3 April or 4 March. An unresolved date convention can put a record in the wrong reporting period.

A required field contains a value in every record. Does that prove the values are correct?

1. Yes, because nothing is missing.
2. No, completeness and accuracy are different checks.
3. Yes, if the field is formatted consistently.

Suggested answer: No, completeness and accuracy are different checks.. A populated field can contain the wrong value. Completeness does not establish accuracy.

A weekly report uses a source that was last updated three months ago. What should you do?

1. Hide the update date.
2. Relabel it as this week’s source.
3. Check fitness for the intended use and disclose the period.

Suggested answer: Check fitness for the intended use and disclose the period.. The usefulness of data depends on its purpose and currency. Make the source period clear and investigate whether a current source is needed.

Use your result

The score counts how many answers match the suggested response in these five scenarios. It does not measure your overall competence or readiness for employment. Choose one explanation you found useful and apply the principle to a practice dataset.

Before sharing a real report, document the source, reporting period, intended use and known quality limitations. Use your organisation’s agreed controls for sensitive data.

Cover image: AI-generated illustrative scene.