Sorting and filtering can make a spreadsheet easier to use, but the records must remain intact. Before changing the view or order, identify the complete dataset, its headings and the fields that belong together in each row.
Practise on fictional data or an authorised reversible copy. A successful-looking sort can still corrupt the meaning if one column moves independently of its related information.
Define the complete record range
Check the header row, first and last records, blank rows, totals and any separate notes. Confirm whether the application’s table or filter range includes every relevant field and row.
Microsoft’s sorting guidance explains sorting ranges and tables, including related-data considerations. Use the actual application’s current documentation and pay attention to any selection warning.
Do not assume a visually empty row marks the end of the data. Conversely, a totals row should not silently be treated as another customer or booking record.
A fictional sorting exercise
Create three invented records with an identifier, name and booking date: R01 for Lee, R02 for Sam and R03 for Alex. Give them different dates, retaining a copy of the original associations.
Sort the whole record range by date. Check that each identifier still has the same name and date as in the source. The order should change; the relationships should not.
If only the date column moved, the apparent records may now be wrong. Stop using the result and recover through the approved reversible process rather than trying to reconstruct associations from memory.
Understand filtering as a view
A filter shows a subset according to defined conditions. Hidden rows usually remain part of the dataset, and the visible list may not represent every record.
Before copying, deleting, counting or reporting, check how the application treats filtered and hidden rows for the action involved. Do not infer a total from what happens to be visible on screen.
Clear or document filters before handing the file to another person where appropriate. A colleague may otherwise believe the visible subset is the complete list.
Use a before-and-after check
Record or verify: For an administration application, this practice exercise can illustrate the records-handling skills in our administration roles guide, provided you explain how you checked that each row remained intact.
- Source: approved dataset and recoverable version.
- Range: all record columns and relevant rows.
- Headers: correctly identified, with totals or notes handled separately.
- Keys: field and direction used for sorting or filtering.
- Associations: sample identifiers still linked to the correct values.
- Completeness: expected record count and any hidden subset.
- Handover: current view and changes explained to the next user.
For important work, choose checks suited to the risk and dataset rather than relying only on a small visual sample.
Resolve inconsistent data first
Mixed dates, numbers stored as text and inconsistent category labels can produce unexpected order or filters. Investigate the underlying values and source meaning through an authorised cleanup process.
After the change, check the actual output needed for the task and preserve a traceable version where required. Reliable sorting and filtering improve access to records while keeping the record relationships and scope understandable.
