268 TC. Lethbridge et al.
4.3.1 Different perceptions of the problemAcademics without much experience in industry may have very different notions about what software engineering involves and what are the real problems.
On the other hand, industry managers tend to vary widely in the software engineering knowledge they possess. This can lead to difficulty communicating, and misunderstandings about the problem that is to be tackled. This issue is very much related to classic difficulties
in requirements analysis where, due to inadequate communication and preconceived ideas, customers have one perception of the problem and software engineers another.
4.3.2 Failure to staff project with sufficient numbers of skilled researchersEmpirical research has not customarily been widely performed in the software engineering community, and for some people lacks a certain respect or is considered to be soft. The Mitel-CSER project has certainly suffered from this phenomenon we have on occasion tried to convince graduate students to become interested in such studies and have found that they don’t see it as real engineering.
Empirical studies of usability, as performed by human factors experts, are seen to be part of an entirely different culture.
For these reasons, it is hard for the project leaders to attract researchers (graduate students, postdoctoral researchers and faculty) who have expertise and interest. Hopefully this book will make a difference.
In addition to having questionable
interestingness or respect, empirical projects also often generate profuse volumes of data, which is very time-consuming to analyze. This acts as a deterrent to software engineering researchers who are used to solving engineering problems.
In the Mitel-CSER project, we attempted to use administrative assistants to transcribe tapes in interviews, however this failed because the interviews used so much technical jargon that the transcribers could not adequately understand them.
Share with your friends: 
