This lab package is currently under construction. The experiment involved a third replication of the perspective-based requirements reading and an extension of these concepts to the reading of UML-based object oriented designs. Currently we have created a list of lessons learned and assembled below the documents used in the experiments.
The requirements experiment was similar to the previous experiments, but abbreviated. Rather than have the entire class perform a control treatment and then the test treatment, we divided the class into two groups. The first group used one of the perspectives from the previous experiment (the Use Case perspective), and the second group used a checklist. We based our checklist on the Volere Template. After the individual reading and error abstraction, the individuals met as teams. Some teams were made of 3 use case students, some of 3 checklist students, and some were mixed. The teams merged the individual fault lists and individual error lists.
Documents used [links to be added soon]:
After the requirements review, the students turned in their team fault lists. We reviewed their fault lists, and removed from the requirements document any defects that any student team had discovered. Even after this, some seeded defects remained in the document as no team had found them. Each team proceeded to generate a UML-based design from these "fixed" requirements. Since Use Cases are an input to the design process that we taught the students, the students used the use cases that they had generated during the requirements review. However, some groups were comprised of three checklist readers; to account for this we provided each group (not just the checklist-only groups) with use cases that one of the students had written during the requirements review.
Documents used [links to be added soon]
After the students turned in their designs, we selected the best two and handed them back to the class for the design reading. We discuss the design reading experiment in great depth in the "Detecting Defects in Object Oriented Designs" paper (link will be provided shortly). Each individual on each team chose a perspective, and each team was required to have all three perspectives covered. After reading the document for defects, the individuals met as a group and merged their defect lists.
Documents used [links to be added soon]
If you have any feedback on this package, please send mail to the author fred@cs.umd.edu.
Any opinions, findings and conclusions or recomendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).
Last Modified on Sun Apr 11 23:23:39 US/Eastern 1999
Michael "Fred" Fredericks (fred@cs.umd.edu)