Title: Report on a Development Project Use of an Issue-Based
Authors: K.C. Burgess Yakemovic and E. Jeffrey Conklin
Citation: Proceedings of the Twenty-fifth International Conference on the System Sciences, January 1992
Reprinted in: Baecker, R.M., Readings in Groupware and Computer-Supported Cooperative Work 1993, pp. 566-579
This paper deals with the loss of vital information during meetings and group gatherings. Their unstructured nature might be the reason for this. The research problems faced by the authors were : how to get the rationale into some kind of a database and how to get it back out effectively. The IBIS structure of Issues, Positions and Arguments is suggested as one possible method. It involves three stages : some choice to be made, some set of alternatives and a commitment to one resolution.
To allow the method to be used with "off-the-shelf" software and hardware, Indented Text IBIS was developed. The root issues were what is required for the product hardware and software? This method was used by the entire team to capture the content of discussion that occurred at many meetings. In addition, two of the developers used it for a variety of individual analysis of tasks.
gIBIS is a SUN workstation based tool which provides a graphical interface for building and browsing IBIS networks. After 18 months of using itIBIS, the indented text files were converted to gIBIS format through a semi-automated process. The update was concentrated primarily on the design category material.
The observations revealed that loading data into the gIBIS provided new mechanisms for organizing the information which allowed issues to be located and updated. The main reasons for failure were identified as schedule pressure and failure to recognize important information the first time they were discussed. Team meetings were found to be more productive and so was the communication between organizations as also among project groups.
One of the interesting discussion points raised in the paper was the authors' belief that a tool for conversational structuring would be widely accepted only if they do not change the way people normally work. In one of the earlier, one other author had the opinion that designers should not restrict themselves to existing modes of work. Which approach will be more succesful is something that has not been figured out yet probably.