Background
The Psychology of Programming Interest Group is responsible for creating a community that is aware of usability issues in programming and allied activities. This community now has active members from USA, France, Israel, Italy, Belgium, Germany, Finland, Sweden, and other countries, and papers have dealt with all aspects of ‘programming’, from the intricate details of coding all the way up to the psychosocial aspects. The Annual Workshop is the chief venue for face-to-face interaction and fomenting links.
The style is deliberately complementary to the more formal style of many workshops and conferences, notably the workshops organised by Empirical Studies of Programming Foundation in USA. Our approach is low-key, with a strong emphasis on getting interested parties into the same place for discussion. We therefore welcome papers that are prospective just as warmly as papers present completed research, and we schedule cheap but lively social events (e.g. a pub quiz based on our research) that will help to create links between groups and individuals.
This Year: technical contents
This year we continued the approach. As ever, the papers presented a wide variety of topics. Invited papers addressed two of today’s big issues: re-use and comprehension (Detienne) and patterns in programming (Winder). Each of these topics combines cognitive psychological considerations with technical details of present-day languages and architectures. Pronouncements by the gurus of computer science and software engineering can be compared to the findings of cognitive psychologists, helping to transfer knowledge, modify over-strong opinion or increase the sophistication of argumentation and research. Regrettably, the third invited speaker (Spohrer) was unable to attend - a particularly unfortunate absence since the emphasis this year had been placed on the teaching and learning of programming.
Nevertheless the special interests of the host department, the Computer-Based Learning Unit, were well represented in a variety of papers treating several paradigms (conventional programming, declarative programming, visual programming, and learning other formalisms such as logic). Particularly interesting and challenging theses were proposed, from the ‘bricolage’ metaphor (Ben-Ari) to the latest developments of ‘Hank’, the Open University’s environment for knowledge representation by cognitive science students - and, at the other end of the scale, the development of a cognitive-based environment for teaching primary-schoolers an understanding of calculator usage.
Questions of evaluation were much aired this year. One of the available methods, Green’s ‘Cognitive Dimensions’ framework, has been tried by a number of groups independently, and this led to intense discussion between the groups with a prospect of a further single- issue meeting in the near future.
Also well aired were problems of comprehension, which reappear in many guises. A significant but unsolved issue is that of ‘mental representations’. Current interest in ‘patterns’ is only one of the possible forms of mental representation. Determinants of comprehensibility need still further study.
Another interesting development was the variety of formal but non-programming notations being investigated, including theorem provers, concept maps, calcuators, and representations for design rationale. We were at first surprised by this, but the development makes considerable sense: there are very few outlets for research on designing formal systems intended for practical use, and fewer still are the outlets in which relationships across many such systems can be compared. The technical committee was therefore pleased to see ‘programming’ being reconstrued in such an inclusive way.
Training and Development
This year, PPIG held its second Doctoral Consortium. We were pleased to have consortium attendees from USA, Israel and Germany as well as the UK. Papers covered a variety of topics and follow-up conversations have shown that attendees took away an increased awareness of the state of the art, bringing a wider variety of methodologies to bear on their chosen questions.
Knowledge Transfer and Dissemination
Dissemination is always a problem for academic workshops. As part of the complementarity that has grown up between ourselves and the Empirical Studies of Programming Series, we leave it to them to produce book-quality published proceedings: a valuable means of dissemination but with obvious consequences of greatly increasing the lead time and the formality of submissions. This year, as well as the usual limp-bound privately-produced proceedings, all papers have been archived on the PPIG web site, which we hope will start to increase the transfer between ourselves and the practitioner community. The dissemination issue brought some discussion at the workshop and plans have been made to archive some of the best past papers as well.
On this topic, it is perhaps worth noting that the technology of soft copy is still not resolved. Submissions in a variety of formats (dialects of Word, LaTeX, postscript, pdf, html) were brought to a common pdf format for archiving, but there was one significant exception, a paper containing a figure too large, and too detailed, to be scaled down to A4. In the future, not only in this area but in many others, we can anticipate not just large detailed graphics, but also the need for dynamic, interactive illustrations, to demonstrate exactly what a programming environment does, for example. We wonder whether EPSRC might investigate frameworks for the dissemination of scientific knowledge to industrial practitioners and provide a recommendations for small workshops and conferences.
Funding and Support
The Annual Workshop of Psychology of Programming Special Interest Group ‘99 was supported by grant no. GR/M56050 from the EPSRC, by the Psychology of Programming Special Interest Group, and by the Computer Based Learning Unit at Leeds University.
Dr Paul Brna and Dr TRG Green
Computer Based Learning Unit
Leeds University
Leeds LS2 9JT