Test effort estimation in regression testing

Abstract

Software test effort estimation has always been a challenge for the software practitioners, because it consumes approximately half of the overall development costs of any software project. In order to provide effective software maintenance it is necessary to carry out the regression testing of the software. Hence, this research work aims to propose a measure for the estimation of the software test effort in regression testing. Since, the effort required developing or test software shall depend on various major contributing factors like, therefore, the proposed measure first estimates the change type of any software, make test cases for any software, then calculate execution complexity of any software and tester rank. In general, the regression testing takes more time and cost to perform it. Therefore, the effort estimation in regression testing is utmost required in order to compute man-hour for any software. In order to analyze the validity of the proposed test effort estimation measure, the measure is compared for various ranges of problem from small, mid and large size program to real life software projects. The result obtained shows that, the proposed test measure is a comprehensive one and compares well with other prevalent measures proposed in the past.