Chapter 5: Random Sampling

Chapter 5: Random Sampling Transcript

In this section we’re going to look at how to pull a random sample from your data.

First, ensure the Customer Database is open in the database window and that Data is selected in the Properties window. Click the Analysis Tab, then select Random in the Sample Group. This is the Random Record Sampling Dialog box.

Enter the number of records you want to select – in this case, I want 10 - and accept the Random number seed value provided by IDEA. This number is used by IDEA as a starting point for the algorithm to calculate the random numbers. If the sample needs to be extended then entering the same random number seed but with a larger sample size produces the same original selection plus the required additional records.

Accept the default values in the Starting record number to select and the Ending record number to select fields. Leave the Allow duplicate records checkbox empty.

Click Fields to pick and choose which fields appear in the output database. Values highlighted in blue are selected. To remove an entry, click it.

Click OK to return to the Random Record Sampling dialog box. Provide a name for the new database you are creating. I’ll call mine Sample of Customers. Click OK to run the sampling task.

The new Sample of Customers database is now the active database and is shown as a child of the Customer Database. There is also a new field been added called SAM_RECNO which stands for Sample Record Number. This is the record number in the master database. In other words, record number 1 is record number x in the Customer database.

To close the Sample of Customers database, click the x in the database tab.