Handling the Trash
A final rule to remember as you explore smart data validation is the often-overused but ever-true phrase "Garbage in, garbage out." Systems today do a much better job than they used to of minimizing free-form text through the use of lookup tables and edit controls for validating data that users enter in GUIs. However, users are increasingly demanding the ability to accept and capture data through direct interfaces with systems outside the database (e.g., through Web services or third-party applications).The idea of applying validation rules to each of these new entry points has not been as widely accepted. Blindly accepting data from an external system potentially reduces data quality by putting control of your data's integrity in someone else's hands. Taking appropriate precautions, such as staging and screening incoming data to ensure it adheres to the same validation rules used by the rest of the system, is essential before integrating new data and existing data sets. Similarly, interface validation must also ensure a complete data transmission, meaning the receipt of all expected rows. Failing to validate completeness introduces the risk of consistency problems as well as the potential understatement of values.
The Challenge
By looking at the topic of smart data validation, we've explored several ways a system leverages many of its own physical constructs to increase the validity and value of data. So, are your organization's systems doing all they can to maintain the highest degree of data quality? If so, then congratulations! If not, then I challenge you to take the next step to increase your data's validity by applying some of the concepts presented here.Then, the next time you describe the data in your organization, your description might even include a characterization of its quality as well.
End of Article
Prev. page
1
2
[3]
next page -->
