To provide guidance around the testing process.
Data QA Checklist
- Validate data against source using crosstabs. Check data history.
- Check calculations for accuracy when filters or dashboard actions are applied.
- Do CASE and IF calculations account for all possible scenarios or is data excluded or incorrectly classified?
- For example, if my case statement is: CASE [State] WHEN ‘Columbus’ THEN ‘Ohio’ WHEN ‘Cincinnati’ THEN ‘Ohio’ ELSE ‘Not Ohio’ END, I will classify any new cities as “Not Ohio”. It would be better to write it as: CASE [State] WHEN ‘Columbus’ THEN ‘Ohio’ WHEN ‘Cincinnati’ THEN ‘Ohio’ ELSE ‘Undefined’ END.
- Do CASE and IF calculations account for all possible scenarios or is data excluded or incorrectly classified?
- Ensure dates are not hard coded – in either filters or the data source.
- Cycle through all filters and actions to check visualizations are loading as expected. Items to check with filters include, is data dropping out that should not be dropping out? Can you account for a user filtering out all the data?
- If you cannot account for a user filtering out all the data, it may be helpful to add “No Data Available”. A tutorial is available on Tableau Server.
Check dashboard performance on Desktop and Server.
- Resolve any timing issues with data refresh and ensuring timing can meeting stakeholder expectations.
- Hide unused fields.
- Add metadata – including workbook description and field definitions.
Visual QA Checklist
- Color should match the Ohio State University brand guidelines. Templates may be found on Tableau Server.
- Check all numbers for consistent formatting (e.g., consistent rounding, bold/not bold, and size).
- Check all tooltips for consistent formatting.
- Review all text for relevant placement and readability. Adjust placement and sizes if needed.
- Check all chart formats for consistency (e.g., consistency in removal of gridlines, axes, tick marks, and row banding).
- Check for spelling. This includes checking on story points, headings, text descriptions, metadata, tooltips, filters, and footnotes.
- Ensure the dashboard is accessible. A more comprehensive guide is available.
- Add a ‘data as of’ date.
- Add contact information for the team in case of questions.
User Testing Checklist
- Show your dashboard to someone unfamiliar with the project. Ask them to explain it back to you in 30 seconds. If they cannot, simplify.
- Give users five questions to answer using the dashboard – without your help.
- Soft launch to a small group before it is released – watch them use it.