Exploring the Influence of Fault Type, Fault Position and Gender on the Testing Process using Code Review Technique

Abstract

In the software process, one of the oldest tasks is coding. Despite the body of knowledge developed, there are still variables such as the type and position of the fault, which require a more in-depth and specific analysis. This study reports a controlled experiment, in which four exploratory analyses are developed, considering the factors: fault type, fault position and gender of the reviewer. The experiment was developed in an academic setting with 54 students from a Mexican institution. From the first two analyses it was concluded that types of faults influence the detection process, since the tester detects more faults by commission than by omission, but also detects more computational faults compared to interface, data and cosmetic. Regarding their position, faults in the first half of the code are more visible than those in the second half. Finally, regarding tester gender, no significant differences were observed in case of effectiveness.