QW2001 Paper 7T2 Mr. James Lyndsay (Workroom Productions) The Importance of Data In Functional Testing

Key Points

• Plan the data for maintenance and flexibility
• Know your data, and make its structure and content transparent
• Use the data to improve understanding throughout testing and the business

Presentation Abstract

A system is programmed by its data. Functional testing can suffer if data is poor. This presentation gives an understanding of the ways that data work fits into the overall test effort, and gives an overview of the ways that good data can be used to improve functional testing.

There are three kinds of test data;
* Environmental data tells the system about its technical environment.
* Setup data tells the system about the business rules.
* Input data is information input by day-to-day system functions. Some input data is fixed and available at the start of the test. Some is consumable, and forms the test input.

The presentation deals with how to recognise these types and their common problems during pre-test, testing and go-live.

Data can be loaded into the system manually, or by tools. The presentation discusses the advantages and pitfalls of various methods and suggests partitioning strategies to allow reliability and flexibility in the same dataset. The frequency and timing of data loading is also discussed.

Data maintenance is a substantial task, often comparable in size with test script maintenance. The presentation discusses common problems and possible solutions. A key solution is that good data content can help reduce the workload of data maintenance.

Good data can allow testing to carry on in areas not covered by the initial scripts and requirements. Naming conventions can help data to be accurate, and can make test results easier to interpret. A good test data structure promotes a common understanding and helps avoid mistakes. Accurate and appropriate content reduces the number of test process errors.

Data can help the business focus when requirements are vague. User involvement in data descriptions allows early insight into possible problems. The presentation discusses the pitfalls and advantages of sourcing data from the business.

Before winding up, a brief mention is made of operational profiles, non-functional testing, and data verification before/during live operation.

About the Author

James Lyndsay is an independent test consultant with ten years experience. Specialising in test strategy, he has worked in a range of businesses from banking and telecomms to the web, and pays keen attention to the way that his clients' focus is shifting away from functional testing.