Long Term Human-Computer Interaction

Although machine interfaces have been made much easier for novices to learn, still very little is known about how users progress from novice to expert performance. This volume is based upon the results of one of the largest continuous field studies ever performed in human-computer interactiona seven year study involving 4,000 students at Sydney University. The results will be valuable to software developers and researchers.

1. Setting the Scene.- 1.1 Evidence for Continuous Change.- 1.1.1 The Challenges of Prolonged Use.- 1.2 A Changing Focus in HCl.- 1.3 Framework for Long Term Interaction.- 1.3.1 Practice.- 1.3.2 Transfer.- 1.3.3 Exploration.- 1.3.4 Vocabulary.- 1.3.5 Flexibility.- 1.3.6 Process Dynamics.- 1.4 The Costs of Longitudinal Studies.- 1.5 The Sydney Study.- 1.6 Outline of the Book.- 2. The Long Haul From Novice to Expert.- 2.1 Skill Acquisition.- 2.1.1 The Power Law of Practice.- 2.1.2 Transfer.- 2.1.3 World Class Performance.- 2.2 The Problems of Novices.- 2.2.1 Active Learning.- 2.2.2 The Minimalist Paradigm.- 2.2.3 Learning by Exploration.- 2.2.4 Discussion.- 2.3 Experienced Performance.- 2.3.1 The Search for Optimality.- 2.3.2 GOMS Models.- 2.4 Longitudinal Studies.- 2.4.1 Rates of Working.- 2.4.2 Knowledge and Methods.- 2.4.3 Thinking Times.- 2.5 Conclusion.- 3. Data Collection.- 3.1 Approaches to Longitudinal Data Collection in HCI.- 3.1.1 The Importance of an Appropriate Methodology.- 3.2 The White Rat of HCI.- 3.2.1 A Retrospective View of Text Editor Technology.- 3.2.2 The sam Editor.- 3.3 Monitoring at Sydney.- 3.3.1 The Research Group’s Goals for the Study.- 3.3.2 Design Goals.- 3.3.3 The Monitoring Language.- 3.3.4 Examples.- 3.3.5 Data Management.- 3.3.6 Sources of Error.- 3.3.7 Analysis Tools.- 3.4 Cohorts.- 3.4.1 CS1 Cohorts.- 3.4.2 Three Year Undergraduate Cohorts.- 3.4.3 Four Year and Other Cohorts.- 3.5 Drop Out Rates.- 3.6 Summary.- 4. Learning and Transfer.- 4.1 The First Lab.- 4.2 Task Descriptions.- 4.2.1 The quit Task.- 4.2.2 The insert Task.- 4.2.3 The keyboard write Task.- 4.2.4 Method of Calculating Learning Curves.- 4.3 Initial Learning in 1991.- 4.3.1 The Effect of Previous Word Processing Experience.- 4.3.2 The 1991 CS2 Cohort 66.- 4.3.3 Performance on the Time Dimension.- 4.3.4 Discussion.- 4.4 Initial Learning in 1992 and 1993.- 4.4.1 Changes to the Teaching.- 4.4.2 Performance Improvements.- 4.4.3 Improved Learning by the 92-95 Honours Cohort.- 4.4.4 Discussion.- 4.5 Typing Speeds.- 4.5.1 Changes Over 2 Years.- 4.6 Performance Improvements in Editing.- 4.7 Conclusions.- 5. A Fresh Look at Vocabulary.- 5.1 Previous Work.- 5.2 Frequencies in the Sydney Study.- 5.2.1 The Effects of Sample Size on Observed Vocabulary.- 5.3 The Type Token Ratio.- 5.4 Trends in TTR.- 5.5 Transitions in the TTR.- 5.5.1 U-shaped Learning.- 5.6 Other Behaviours.- 5.7 Discussion.- 5.8 Conclusions.- 6. Exploration.- 6.1 Previous Field Studies of Exploration.- 6.1.1 Styles and Stages.- 6.1.2 Models.- 6.2 Operational Definition of Exploration.- 6.3 Early Exploration.- 6.3.1 Quantity of Early Exploration.- 6.3.2 The Top 10 Measure.- 6.3.3 Cohort Analysis.- 6.3.4 Discussion.- 6.4 Continuous Exploration.- 6.4.1 Exploration Rate.- 6.4.2 Exploration Curves.- 6.4.3 Advent Surfaces.- 6.4.4 Cohort Comparison.- 6.4.5 Summary.- 6.5 Serendipity.- 6.5.1 Serendipitous Digrams.- 6.6 The Importance of Chance.- 6.7 Conclusions.- 7. Crossovers.- 7.1 A Classification of Temporal Effects.- 7.1.1 System Delays.- 7.1.2 Task Gain.- 7.1.3 Learning Lag.- 7.2 Longitudinal Changes in sam Command Use.- 7.2.1 Population View.- 7.2.2 Individual View.- 7.3 File Write Methods.- 7.3.1 GOMS Model.- 7.3.2 Keystroke Level Model.- 7.4 Visualisation Technique.- 7.4.1 Measurement of Time.- 7.4.2 Command Preferences.- 7.4.3 User Numbers.- 7.5 File Write Crossover Results.- 7.5.1 The Range of Behaviours.- 7.5.2 Comparison of Cohorts.- 7.5.3 The Effect of Activity Rate.- 7.6 Discussion.- 7.6.1 Methodology.- 7.7 The Zone of Exploration Model.- 7.7.1 Example.- 7.7.2 Oscillation.- 7.7.3 Zone of Proximal Development.- 7.7.4 Type Token Ratio Constraints.- 7.8 Conclusion.- 8. Conclusions.- 8.1 Methodologies.- 8.1.1 Continuous Monitoring.- 8.1.2 The Focus on Tasks.- 8.1.3 Treatment of Time.- 8.1.4 Retrospective Data.- 8.1.5 Cohorts.- 8.1.6 Cross-Sectional Populations.- 8.2 Process Dynamics.- 8.2.1 Vocabulary.- 8.2.2 Exploration.- 8.2.3 Crossovers.- 8.3 Implications for Design.- 8.3.1 Adaptive Interface Constraints.- 8.3.2 Support for Exploration.- 8.3.3 Support for Serendipity.- 8.4 Training, Education and Job Design.- 8.4.1 A Technological Underclass.- References.- A. Command Codes and Ranks.- A.1 Command Ranks for the 91–93 Cohort.- A.2 Honours Year Command Ranks.- B. A Model of the Command Distribution.- B.1 Introduction.- B.2 Application to Observed Vocabulary.- Author Index.