MacKenzie, I. S. (1991). Fitts' law as a performance model in human-computer interaction. Doctoral dissertation. University of Toronto: Toronto, Ontario, Canada.

Fitts' Law as a Performance Model in
Human-Computer Interaction

I. Scott MacKenzie

Abstract

1. Introduction

1.1 Background
1.2 A Role for Human Performance Models
1.3 Information Theory Foundation
1.4 Equation by Parts

2. Detailed Analysis

2.1 The Original Experiments
2.2 Problems Emerge
2.3 Variations of Fitts' Law
2.4 Effective Target Width
2.5 Re-analysis of Fitts' Data
2.6 Effective Target Amplitude
2.7 Dimensions of Movement Tasks
2.8 Targets and Angles
2.9 Competing Models
2.9.1 The Linear Speed-Accuracy Tradeoff
2.9.2 Power Functions

3. Computer Input

3.1 Device Differences
3.1.1 Rotary vs. Linear Motion
3.1.2 Displacement vs. Force Sensing
3.1.3 Absolute vs. Relative Positioning
3.1.4 Dimensions Sensed
3.1.5 Control-Display Gain
3.1.6 Spatial Transformation
3.1.7 Interactions Between Muscle Groups and Limb Groups
3.2 Task Differences

4. Literature Review and Surveys

4.1 The Generality of Fitts' Law
4.2 Re-analysis of Published Data
4.3 Human Factors and Human-Computer Interaction
4.3.1 Card, English, and Burr, 1978
4.3.2 Drury, 1975
4.3.3 Epps, 1986
4.3.4 Jagacinski and Monk, 1985
4.3.5 Kantowitz and Elvers, 1988
4.3.6 Ware and Mikaelian, 1987
4.4 Across-Study Comparison of Performance Measures
4.5 Sources of Variation
4.5.1 Device Differences
4.5.2 Task Differences
4.5.3 Selection Technique
4.5.4 Range of Conditions and Choice of Model
4.5.5 Approach Angle and Target Width
4.5.6 Error Handling
4.5.7 Learning Effects
4.5.8 Summary
4.6 Fitts' Law in Dragging Tasks

5. Methodology

5.1 Hypotheses
5.2 Experiment 1
5.2.1 Subjects
5.2.2 Apparatus
5.2.3 Procedure
5.2.4 Design
5.3 Experiment 2
5.3.1 Subjects
5.3.2 Apparatus
5.3.3 Procedure
5.3.4 Design
5.4 Experiment 3
5.4.1 Subjects
5.4.2 Apparatus
5.4.3 Procedure
5.4.4 Design

6. Results and Discussion

6.1 Plan for Analysis
6.2 Experiment 1
6.2.1 Adjustment of Data
6.2.2 Distribution Characteristics
6.2.3 Summary Tables
6.2.4 Design for Device and Task Effects
6.2.5 Movement Time
6.2.6 Errors
6.2.7 Index of Performance
6.2.8 Fit of the Model
6.2.9 Comparison of Methods for Calculating IP
6.2.10 Model Comparisons
6.3 Experiment 2
6.3.1 Adjustment of Data
6.3.2 Approach Angle
6.3.3 Fit of the Model
6.3.4 We in Two Dimensions
6.3.5 Model Comparisons
6.3.6 Parametric Analysis
6.3.7 The W' Model
6.4 Experiment 3
6.4.1 Target Width in Dragging Tasks

7. Summary and Conclusions

7.1 The Shannon Formulation
7.2 Target Width
7.3 Normalizing for Spatial Variability
7.4 Device and Task Differences
7.5 Serial vs. Discrete Tasks
7.6 Error Rate as a Parameter
7.7 Using Fitts' Law: An Example
7.8 Future Directions

References

Appendix A

Appendix B