Planning for Test Data Preparation as a Best Practice

By Paula Grose

After working in and managing testing efforts on and off for the past 18 years, I have identified a best practice that I use in my testing projects and I recommend it as a benefit to other testing projects, as well.

This best practice is test data preparation, which is the process of preparing the data to correlate to a particular test condition.

Oftentimes, preparing data for testing is a big effort that people underestimate and overlook. When you test the components of a new system, it’s not as simple as just identifying your test conditions and then executing the test—there are certain factors you should take into account as you prepare your test environment. This includes what existing processes, if any, are in place to allow for the identification or creation of test data that will match to a test condition.

A test case may consist of multiple test conditions. For each test condition, you must determine all the test data needs. This includes:

• Input data
• Reference data
• Data needed from other systems to ensure synchronization between systems
• Data needed to ensure each test will achieve its expected result

Planning for test data preparation can greatly reduce the time required to prepare the data. At the overall planning stage for testing, there are many assessments that should be conducted, including:

• Type of testing that will be required
• What testing tools are already available
• Which testing tools may need to be acquired

If, at this point, there are no existing processes that allow for easy selection and manipulation of data, you should seek to put those processes in place. Most organizations have a data guru who is capable of putting processes in place for this effort—or at least can assist with the development of these processes.

The goal is to provide a mechanism that will allow the selection of data based on defined criteria. After you do this, you can perform an evaluation as to whether the existing data meets the need—or identify any changes that must be made. If changes are required, the process must facilitate these changes and provide for the loading/reloading of data once changes are made.

One word of caution concerning changing existing data: You must be certain that the existing data is not set up for another purpose. Otherwise, you may be stepping on someone else’s test condition and cause their tests to fail. If you don’t know for sure, it is always better to make a copy of the data before any changes are made.

About the Author: Paula Grose worked for the State of California for 33 years, beginning her work in IT as a Data Processing Technician and over time, performing all aspects of the Systems Development Life Cycle. I started in executing a nightly production process and progressed from there. As a consultant, Paula has performed IV&V and IPOC duties focusing on business processes, testing, interfaces, and data conversion. She currently leads the Data Management Team for one of KAI Partners’ government sector clients. In her spare time, she is an avid golfer and enjoys spending time with friends, and playing cards and games.