Executable Test Specifications Show No Negative Effect on Finding and Correcting Test Faults in Event Processing Application Development

This paper presents the results of an experiment with focus on comprehension of acceptance test cases in the field of event processing application development. We compare the usage of a semi-structured test specification format proposed by the IEEE and a structured table-oriented format called EPTDL, which has been proposed for automated acceptance testing in the domain of event processing. Our results show that there is no significant difference with respect to the effort used to reveal and correct faults in test case specifications as well as the correctness of the solutions. Considering the possibility to automate test execution our results encourage the usage of ATDD in event processing application development.