Browse Prior Art Database

Test Classification and Prioritisation to Improve Test Efficiency

IP.com Disclosure Number: IPCOM000032605D
Original Publication Date: 2004-Nov-09
Included in the Prior Art Database: 2004-Nov-09
Document File: 2 page(s) / 64K

Publishing Venue

IBM

Abstract

Test cases for testing software are classified using a scheme which identifies the type of defects which they are likely to uncover. The order in which test cases are executed is based on the complexity of the type of defect; from the simplest to the most complex. By doing this it is possible to improve the efficiency of the testing process.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Test Classification and Prioritisation to Improve Test Efficiency

Disclosed is a method for ordering test cases for testing software which will improve the overall efficiency of software development during the testing phases of a project.

    Nearly all software development projects suffer from schedule constraints. Because of time pressure, development phases (development, functional testing, system testing) overlap to such a degree that system testing often ramps up before functional testing is complete. When this happens system test cases find defects that should have been found earlier. This has the following wasteful effects:

1. System testing time is wasted trying to setup and run complex tests which fall over very simple problems.
2. System testers are prevented from running any useful tests due to basic errors in one or more components.
3. The resultant diagnostics from such a failure are much more complex and therefore more time consuming to analyse and hence fix than simple functional tests.

    Test cases are written with sufficient granularity to identify the type of defect that they are likely to uncover; this is known as the trigger. The order in which these test cases are executed is based on the complexity of the trigger; from the simplest to the most complex.

    System test cases are principally composed of a series of simple tests from various components. System testing is not started until the test cases for most of the simplest triggers have successfully run.

    By executing functional test cases in the order described above, rather than traditionally by functional area, system test cases are less likely to find functional defects.

    This method was originally created for a waterfall development process with distinct phases: development, function test and sys...