If an archive or bucket is in a specific order (ex., chronological), or is too numerous for human analysis, it is possible to select a random sample during the dataset creation process.
To select a random sample, follow these steps:
1. On your DiscoverText Dashboard, select the project of your choice from Recent Projects, or choose it from the Project drop-down menu on the Navigation Sidebar.
2. In the Documents section on the Navigation Sidebar, click the + sign next to Data Archives to expand the list of your archives.
3. Right-click on the archive of your choice, and choose create dataset.
4. A pop-up will prompt you to create a name for your dataset. Enter a name for your dataset and click OK.
5. You will be sent to the Create Dataset page, where you should check the box Sample Dataset?.
6. An option named Sample size will appear, so you can select the size of your random sample.
7. Check the Advanced options of your choice, and click Create Dataset to finish.