Assessing Human Mobile Computing Performance by Fitts' Law
Thomas Alexander (FGAN - Research Institute for Communication, Information Processing and Ergonomics, Germany), Christopher Schlick (RWTH Aachen University, Germany), Alexander Sievert (German Sport University Cologne, Germany) and Dieter Leyk (German Sport University Cologne, Germany and Central Institute of the Federal Armed Forces Medical Services Koblenz, Germany)
Copyright: © 2008
This chapter describes the interdependence between locomotion while walking and human input performance in mobile Human-Computer-Interaction (HCI). For the analysis of the interdependence, appropriate performance measures, for example, subjective workload ratings or error rate, have to be applied. The way in which Fitts’ law can enhance the analysis is explained. In an experiment with n=18 participants, the general indices of performance (bits per second) were measured while standing and walking with constant speed (2, 3.5, 5 km/h). Results show a significant increase of the error rate and a significant decrease of the index of performance for increased walking speed. Subsequent regression analyses allow quantitative estimation of these effects. The results show a division of the interdependence in two parts, based on the difficulty of the input task; they define threshold values for accuracy of user input. These values can be applied for the implementation and design of future Graphical User Interfaces (GUI) for mobile devices.
Key Terms in this Chapter
Index of Performance: The index of performance is a measure for characterizing the speed of a visually controlled movement. Thus, it is also a measure of movement performance. It is calculated by the reciprocal value of the linear coefficient b of Fitts’ law. Another definition is the ratio of IDaverage to MTaverage. Both definitions allow a comparison of input performance under different movement conditions.
Subjective Workload: There are many definitions of the term Workload. One concise definition was given by O’Donnell & Eggemeier (1986): “… Workload refers to that portion of the operator’s limited capacity actually required to perform a particular task.” Subjective workload describes the effort invested by human operators into task performance. It can be assessed by subjective ratings.
Mobile Computing: In contrast to stationary computing, which comprises stationary working with a computer at the same location, and portable computing, which refers to stationary working with a computer at different locations, the term mobile computing describes working with a computer while moving. This leads to weight and size requirements and special demands for human-computer interaction. The user interface has to consider the characteristics of mobility, for example, disturbing, random external forces because of the movement, parallel processing of orientation tasks, and so forth, for an optimal user input performance.
Fitts’ Law: Based on Shannon’s theorem for information processing, Fitts’ law allows the estimation of required times for rapid movements between a starting point and a target area. Fitts introduced the index of difficulty (ID) as a characteristic measure for such movements. The ID is defined as: ID = log2 (2 A / W). The term A describes the amplitude of the movement, or the distance between starting point, and target area, and the term W, the width of the target area. The movement time MT is linearly dependent on the ID. It is: MT = a + b ID. The coefficients a and b are both regression coefficients. The coefficient [a] defines the intercept for ID=0 and the coefficient b the steepness of the relationship.