POLIBITS, Vol. 62, pp. 5-12, 2020.
Abstract: Task cohesion is a key component of team performance. This paper explored the use of collaboration and content-based measures to examine task cohesion within global software development teams. The study aimed to predict the perception of task cohesion among teams involving students from two different countries. The study applied collaboration from previous work and also proposed new metrics such as Reply Similarity and Reply Rate. In addition, a machine learning classifier is used to derive content measures by categorizing teams’ message interactions as social, planning, or work. Correlation analyses are conducted to examine whether collaboration and metrics are predictive of task cohesion. The analyses are conducted at the individual and group levels and used the culture factor as a control variable since cohesion has been found previously affected by location. The research findings suggest that content-based measures were more effective in predicting individual-level cohesion while collaboration-based metrics were more effective at the group-level.
Keywords: Component, formatting, style, styling, insert
PDF: Collaboration and Content-Based Measures to Predict Task Cohesion in Global Software Development Teams
PDF: Collaboration and Content-Based Measures to Predict Task Cohesion in Global Software Development Teams
https://doi.org/10.17562/PB-62-1
See table of contents of POLIBITS 62.