An Experimental Study for Complete-IOCO Theory

Context: Conduction of empirical studies in the software testing area has grown in recent years. Besides, there is a need to evaluate the cost and effectiveness of existing techniques. Objective: This paper presents an experimental study to evaluate the applicability cost of the test generation using the Complete-IOCO method. This method automatically generates test suites based on specification models from Mealy Input/Output Transition Systems (Mealy IOTS). Method: A large number of specification IOTSs were randomly generated and the Complete-IOCO method was used to generate a test suite for each specification IOTS. Results: The results show that the number of stable states and the number of inputs are strongly related to the number of test cases and the average size of the test suites. These results were compared with a case study performed previously with real specifications, evidence that the size and the number of test cases of the randomly generated IOTS test suites are similar to the real specifications. Conclusion: Thus, the randomly generated specification IOTSs can aid in the test generation methods evaluation and comparison.