Data cleaning

Introduction

Data cleaning lets you selectively remove or restore records from your data download and analysis. This is useful for removing test data or excluding records that appear to be from low-quality respondents, such as those who completed the survey in an unrealistically short time.

The data cleaning dialog is shown. Called out are the four numbered steps listed above.

Removing records

Records can be removed using either system-generated Record IDs or author-defined Unique pass-in IDs. Both are available in your data download.

Two methods are available:

  • Remove: paste a list of IDs to remove from the dataset.
  • Keep: paste a list of IDs to keep, removing all others.

Select your method and ID type from the dropdowns, then paste or type a list of IDs — one per line — and click Save.

After saving, the section displays the number of records removed. To view removed records, download the spreadsheet. To remove additional records, click Define records again.

Restoring records

To restore all removed records, click Restore all records and confirm in the dialog.