Inferring productivity factor for use case point method

Abstract

This paper aims to investigate and infer new range for productivity factor for Use Case Points method. Productivity factor is a ratio of man hours per use case point. In this paper, we use linear regression as an inference tool for productivity factor. This value is inferred from three different datasets. Then, we compared these inferred values with standard value for productivity factor. This study use MMRE (Mean Magnitude of Relative Error) measure and standard statistical methods for reliability assessment. The MMRE is also used for all comparisons. The experimental results show that the range for productivity factor is much narrow than recommended values with the mean of 15.