An Experiment Investigating

Procedural Techniques for Perspective-Based Reading

And

Error Abstraction

This is only a first pass at creating a lab package on the web for this experiment. In the future we intend to make this site more complete, including background and motivation for the experiment, and better organized. Currently we have available only our experimental artifacts and some lessons learned, so that this experiment may be replicated elsewhere.

This experiment investigates the effects of two techniques for detecting defects in requirements documents: a procedural form of Perspective-Based Reading (PBR2) and error abstraction. A plan of study which provides a short description of the motivation for the study and our research questions is available. Our full experimental design is below.

Click on each section title to go to a list of the documents needed for that session.

Session

Baseline

Individual Treatment

Team Treatment

1A

Ad Hoc Review

_

_

1B

_

Error Abstraction

_

1C

_

_

Team Meeting

2A

PBR2 Review

_

_

2B

_

Error Abstraction

_

2C

_

_

Team Meeting

Variations on the Design

The experiment is designed to be easily shortened if only one of the techniques is to be investigated: If PBR2 is of interest, 1A,1C,2A, and 2C may suffice, while if error abstraction is the only technique to be investigated, 1A,1B, and 1C is an option.

We also developed an Object-Oriented design perspective with appropriate data collection forms that was not used in our run of the experiment. You may be interested in using this perspective in addition to or in place of some of the others.

Feedback

Send mail to the author:
fshull@cs.umd.edu

This material is based upon work supported by the National Science Foundation under Grant No. 9706151.

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).