Neil Thompson is a UK-based independent testing consultant and manager, serving blue-chip clients through www.TiSCL.com, either directly or with other consultancies. Co-author with Paul Gerrard of 2002 book "Risk-based e-business testing".

Neil has worked for 30 years in information systems, first with a computer manufacturer, then two leading software houses, an international user organisation, followed by ten years as a management consultant with global firms and nine years with his own company.

Has a wide and international perspective on the IS business, through diverse roles including programmer, systems analyst, team leader for enhancements and maintenance, database administrator, and project manager.

Member of British Computer Society (and its specialist interest groups in Software Testing and Configuration Management), also a Certified Management Consultant and IEEE member. Neil spoke at the first EuroSTAR in 1993, then again in 1994, 1999, 2002 (with Paul Gerrard), 2004 (best paper award) and 2006. Also spoke at STAREast in 2003, attempting to build a bridge between “best practice(s)” and the Context-Driven School.

Neil is still old enough to appreciate The Bothy Band and Horslips.

Holistic Test Analysis and Design

Neil Thompson

Thompson information Systems Consulting Ltd.

(presentation prepared in collaboration with Mike Smith, Testing Solutions Group)

To test professionally and understand software risks fully, we need to know what our tests cover. Counting test cases is not enough—that’s like sizing business requirements by counting program modules. Neil Thompson presents a test analysis and design method that integrates four key entities into a holistic approach: test items, testable features, test basis documents, and product risks. Testing standards and textbooks have anaesthetized many into the delusion that test cases are a known quantity and all we need to derive them is a few basic techniques. Neil argues that we need a broader approach, incorporating exploratory techniques and new thinking into our testing. We should consider:

§         distinguishing logical and physical coverage;

§         information traceability;

§         how coverage should play a part in governance scorecards; and

§         a measurement framework for management to understand testing better.