Temporal Aspects of Usability

Papers from a Workshop

Chris Johnson and Phil Gray

Temporal properties of interaction were largely neglected in the first generation of interface design notations. GOMS, TAKD and production rules were all developed to identify operator tasks and analyse traces of interaction These approaches, typically, abstracted away from temporal issues. This reflected the contemporary focus upon interaction with stand-alone office systems, such as word-processors and spread-sheets. Similarly, early experimental work in HCI focused upon the impact of modes and errors on task completion times. It tended to ignore the changes in interaction that occur as users gain new skills through the prolonged use of complex systems.

Recent years have, however, seen the increasing application of multi-user, computer supported co-operative work tools, such as conferencing systems and shared text-editors. They have also seen the growth of distributed information sources using Internet facilities. Both of these trends have increased the importance of temporal properties of interaction. Tasks may be complicated or abandoned if users cannot rely upon timely responses from their colleagues. Remote site failures can lead to frustration and error when using information retrieval systems, such as Web browsers.

In response to these recent developments, a number of HCI researchers have turned their attention to temporal aspects of usability. This work is characterised by its diversity. Some groups have extended the application of formal development techniques from real-time and safety-critical systems. Others have developed innovative experimental methods that can be used to identify long term trends as well as short term tactics during interaction with complex systems. This diversity creates a number of problems for research in this area. In particular, it can be difficult to keep track of relevant research in other disciplines.

In July 1995, a workshop on `Temporal Aspects Of Usability' was organised at the University of Glasgow. This was intended to provide a common forum for researchers in this area. The following contributions were submitted as position papers for this workshop. They fall into three broad themes.

The first strand is represented by Thomas' paper and the contribution from O'Donnell and Draper. These articles show that empirical evidence can be used to identify the temporal strategies that users exploit during interaction.

The second strand is represented by the papers from Palanque and Bastide and from Dix and Abowd. These articles focus upon the problems of developing appropriate models for temporal properties of interaction. This work provides a bridge between the real-time observations of empirical investigations and the abstract temporal constructs of formal specifications.

Finally, the papers by Abowd and Ton and by Fields, Wright and Harrison and by Mezzanotte and Paterno' illustrate the ways in which temporal specification notations can be constructively applied to support the development of human-computer interfaces.

Acknowledgments

The workshop from which this collection was drawn received support from the UK Joint Councils Initiative project "Temporal Aspects of Usability" no.9201233. It formed part of Glasgow University's Computing Science Research Festival and was organised in collaboration with the British HCI group.

Issue
Article
Vol.28 No.2, April 1996
Article
Issue