Design of Programmer’s Skill Evaluation Metrics for Effective Team Selection

Abstract

Software projects are now developed in critical environments with hard restrictions and contesting limitations. For developing a product under deadline, cost, and quality constraints, extraordinary efforts are required. The team selection is the part of software planning to avoid the time, cost, and maintenance overheads. This paper has provided five metrics to measure the aspects of the programmer’s capabilities. The metrics are provided to measure the technical capability, experience, bug resistivity, coding capability, and learning interests of programmers. Each measure of programmer capability is based on multiple sub-features. These features and sub-features are evaluated based on the rating collected from the programmer, group members, and the team leader. Each team member submits his view to rate the programmer’s adaptation to specific features. The rating of each co-programmer is evaluated under the light of precise distinctive weight. The weights are assigned based on the feature and its dependency and knowledge of the team members. The evaluated feature weights are finally applied under high-level capability metrics to measure the programmer’s strength for that feature. After generating the individual capability measure, the aggregate operators are applied to conclude the capability of the programmer. At the final stage, the rule-based decision criteria are defined to distinguish the expert, skilled, and low-performance programmers. The experimental data are collected by conducting a survey on five teams of programmers with overall 30 programmers. The proposed metrics adaptive model can improve the decision criteria for the selection of team members for specific projects.