METRIC IN PREDICTING ERROR IN SOFTWARE: SYSTEMATIC LITERATURE REVIEW

Abstract

Early detection in software of errors can save time, cost, and reduce software complexity. The goal of writing this Systematic Literature Review (SLR) was to identify software metrics and their application in predicting software errors. The method used in this writing was the Systematic Literature Review (SLR) using Publish and Perish software. The results of this Systematic Literature Review showed that Object-oriented metrics were used in more than a majority of the selected articles compared to size metrics, complexity metrics, and process metrics, with Chidamber and Kemerer (CK) being the most frequently used. Object-oriented metrics were reported to be more successful in detecting errors compared to size and complexity metrics. On the other hand, process metrics appeared to be better at predicting errors when combined with size metrics.