Submitted to Human-Computer Interaction Journal, January, 1996
User-Oriented Design Descriptions:
Training Software Engineers in the PUM Instruction Language
Ann E. Blandford 1
School of Computing Science, Middlesex University, Bounds Green Road, London, N11 2NQ, U.K. Email: A.Blandford@mdx.ac.uk
Simon J. Buckingham Shum 2
Knowledge Media Institute, The Open University, Milton Keynes, MK7 6AA, U.K. Email: S.Buckingham.Shum@open.ac.uk
Richard M. Young
MRC Applied Psychology Unit, 15 Chaucer Road, Cambridge, CB2 2EF, U.K.
A novel approach to designing or analysing systems only becomes useful when it is usable by practitioners in the field, and not just by its originators. Design techniques often fail to make the transition from research to practice because insufficient attention is paid to understanding and communicating the skills required to use the technique. This paper reports on work to train software engineering students to use a user-centred language for describing and analysing interface designs called the ?Programmable User Models Instruction Language?, or IL. Various types of data, including video, students? IL descriptions and brief usability reports were collected and analysed to establish what the outcomes of this training were. We conclude that students were able to grasp most of the important concepts in the IL. There is a measurable, if small, shift from the pre-test to the post-test analyses, indicating some influence from the IL construction process. However, the training and practice were too brief for the subjects to internalise enough of the semantics and underlying theory to use the IL really effectively. This study raises generic issues pertaining to the transfer of theoretically-based HCI design techniques to practitioners. The discussion focuses on the level of theoretical training required, the nature and encapsulation of craft skill, and the different degrees of formality with which a formalism can be deployed.
When a novel HCI technique is developed, one of the important issues to be addressed is how the technique can be transferred from the research to the practitioner community. Inevitably, as the work on Cognitive Walkthroughs has shown over the past few years (Lewis, Polson, Wharton & Rieman, 1990; Polson and Lewis, 1990; Wharton, Rieman, Lewis & Polson, 1994), this is unlikely to be a single event in which a technique, having been developed to maturity, is then handed over to the practitioners to be instantly taken up and exploited. Transfer involves both developing effective training material and establishing that the potential benefits to practitioners outweigh the initial costs of learning the technique (Buckingham Shum & Hammond, 1994b; Buckingham Shum, 1995). There is an iterative cycle of activity in which developers learn to take a practitioner-centred view on their technique, and adapt it in the light of experience, while practitioners learn about the technique and its scope and how to apply it.
In this paper, we report on an investigation that took place fairly early in that cycle of interaction between developers and practitioners. Two of the authors (Young and Blandford) are the developers
1 formerly at MRC Applied Psychology Unit, 15 Chaucer Road, Cambridge, CB2 2EF, U.K. 2 formerly at Human-Computer Interaction Group, Dept. Psychology, University of York, York, YO1 5DD, U.K.