Extraction of the Sample Data
The table below outlines the appropriate uses and limitations of extracting sample data.
Table 11: Appropriate Uses and Limitations of Extracting Sample Data| Use of the Random Sample | Examples |
|---|
| Conditions that allow for the random sample | - The volume of data is very large and extracting a full census would need an unreasonable amount of time
- The data are too large to fit onto the DVC’s portable media
|
| Uses of the random sample to confirm individual data elements | - To verify that calculations are performed correctly
- To make sure date ranges are correct
- To evaluate whether officials filter or categorize specific records properly
- To verify that any manual manipulation of the source and final stage data is exact
- To evaluate missing data and the impact on the calculation of derived data fields
- To verify that the organization is properly defining terms per CMS regulations, guidance, and Part C and Part D Reporting Requirements and Technical Specifications
|
| Limitations of the random sample | - Can’t be used to calculate total counts or sums for data elements, which need extraction of the full census
|