What affects team behavior? Preliminary linguistic analysis of communications in the Jazz repository

There is a growing belief that understanding and addressing the human processes employed during software development is likely to provide substantially more value to industry than yet more recommendations for the implementation of various methods and tools. To this end, considerable research effort has been dedicated to studying human issues as represented in software artifacts, due to its relatively unobtrusive nature. We have followed this line of research and have conducted a preliminary study of team behaviors using data mining techniques and linguistic analysis. Our data source, the IBM Rational Jazz repository, was mined and data from three different project areas were extracted. Communications in these projects were then analyzed using the LIWC linguistic analysis tool. We found that although there are some variations in language use among teams working on project areas dedicated to different software outcomes, project type and the mix of (and number of) individuals involved did not affect team behaviors as evident in their communications. These assessments are initial conjectures, however; we plan further exploratory analysis to validate these results. We explain these findings and discuss their implications for software engineering practice.