Observations on using empirical studies on developing a knowledge-based software engineering tool

There exist a wide variety of techniques for performing empirical studies which researchers in human-computer interaction have adapted from fields of cognitive psychology, sociology and anthropology. An analysis of several of these techniques is presented through an approach that balances empirical study with tool development. The analysis is based on, and illustrated with, a several-year experience of consulting in a scientific software environment and in building an evaluating a prototype knowledge-based tool to capture aspects of that experience. Guidelines for applying specific techniques and cautions about potential pitfalls are discussed. Many additional examples of using the techniques are cited from the literature.<>