Long Term Human-Computer Interaction: An Exploratory Perspective

This book is about longitudinal research in Human-Computer Interaction (HCI). Since the early 1980s much has been understood about the problems of novices learning new systems as well as the performance of expert users. However there is still a lot to discover about the transition from novice to expert and its implications for the design of systems. For instance: - How long does it take to become expert? - Does early experience of a system have any effect on subsequent flexibility? - How can flexibility and long term learning be promoted? - Are there strong constraints that should be taken into account in the design of adaptive systems? Longitudinal research in HCI has been rare for many reasons. There is always pressure to obtain results - the current climate of short-termism does nothing to promote a longer perspective. The field of HCI itself is changing fast, and there are often virgin technologies to explore which can be more ex­ citing and potentially profitable than research aimed at basic understanding. There is also a possibly mistaken view that longitudinal studies are always inherently expensive. The present volume grew out of a project at Sydney University. The very first ideas were discussed in 1989. In 1991 data logging started of a mainly undergraduate population using a Unix-based editor, sam(Pike 1987}. Due to good fortune the system continued unchanged into a second year and it became clear data collection had very low marginal costs.

Wordprocessors and other information technology are now used on a daily basis by a wide variety of people. Although interfaces have been made much easier for novices to learn, very little is still known about how users progress from novice to expert perf ormance. Improved understanding of long term learning and usage patterns have implications for HCI in terms of: design for prolonged flexibility; adaptive interfaces; user modelling; usability assessment; training; and data mining in web site management.

This volume is based upon one of the largest continuous field studies ever performed in Human-Computer Interaction - seven years of editor use by 4000 students at Sydney University. It assesses the results and argues that the process of change in long term HCI appears to be constrained by three pillars: (i) an essentially stationary measure of vocabulary use, (ii) continuous growth through exploration, and (iii) specialisation between methods via crossovers.

Long Term Human-Computer Interaction will be of interest to HCI researchers and designers, computer scientists, information specialists, cognitive scientists, academics and postgraduate students.