Dataset Guide · Example edition
Make the data understandable before analysis.
This fictional dataset records handover tasks in a small usability exercise. The field examples describe a schema; they do not contain real participant information.
| Field | Type | Meaning | Example |
|---|---|---|---|
| record_id | Text | Unique observation identifier | R-001 |
| variant | Text | Assigned document layout | Index |
| completed | Boolean | Task met the defined check | true |
| duration_seconds | Integer | Elapsed task time | 95 |
| issue_note | Text | Non-identifying observation | Opened old draft first |
Unit of observation
One row represents one task attempt. A participant who performs several tasks may have several rows, so row count is not automatically participant count.
Quality checks
Require unique record identifiers, known variant values, nonnegative durations, and a consistent completion definition. Record missing values explicitly instead of substituting zero.
Responsible reuse
Keep identifying data out of the analysis extract. Document collection dates, consent, transformations, and permitted uses in the actual dataset record before sharing it.