Theoretische und experimentelle Analysen zum Fitts´schen Gesetz bei der Interaktion mit großflächigen berührungssensitiven Bildschirmen

• Theoretical and experimental analyses on Fitts´ Law on large-scale touch screens

Vetter, Sebastian Timm; Ziefle, Martina (Thesis advisor)

Aachen : Publikationsserver der RWTH Aachen University (2012)
Dissertation / PhD Thesis

Aachen, Techn. Hochsch., Diss., 2012

Abstract

The aim of this thesis is to develop a theoretically founded and experimentally proven refinement of Fitts´ Law for the interaction with large touch screens. Fitts´ Law defines the channel capacity of the human motor system while performing target-oriented pointing movements. Nowadays Fitts´ Law is mainly used in the field of Human-Computer-Interaction for example to analyze graphical user-interfaces according to the size and position of interaction elements. These interaction elements typically carry textual information and are of rectangular shape. Their position on the screen is variable. Applying Fitts´ Law in the field of software ergonomics raises the questions of how to define the width for rectangular target objects and how the angle between starting position and target object affects movement time. Although numerous experimental studies focused on these questions, the state of knowledge is still uncertain and contradictory. Most studies suffer from methodological weaknesses because they do not separate between effects of the angle and effects induced by target width on movement time. Furthermore the current state of research shows numerous poorly validated extension of Fitts´ Law. Therefore, in this thesis, the influence of target width and the angle is analyzed separately in two experimental studies. The experiments were conducted on a large touch screen. Against the background of an increasing number of older computer users an age diverse sample (N=100) was investigated. In the first study different target width parameters were investigated (e.g. the target width in the direction of motion W´, the shorter side length Wmin, the sum of the side length W+H and the area of the target object W*H). In order to eliminate confounding effects of the angle, rectangular target objects were displayed at a fixed position and rotated around their centre. This experimental setup results in a systematical variation of only the target width in the direction of motion W´, while other target width parameters e.g. the shorter side-length (Wmin) remains constant. The results show that the movement time is dependent on the target width in and orthogonal to the direction of movement. The distribution of endpoint coordinates confirms that the target width in direction of movement corresponds to the correction of the movement length and target height to the correction of movement direction. In the second study the influence of the angle was investigated. In order to eliminate confounding effects of the target width circular target objects were used. The targets were displayed at constant amplitudes at different angular positions. Additionally, the movement trajectories were recorded with an infra red motion tracking system. The results of the movement time data show a sinusoidal distribution depending on the angle which can be explained by the inertial anisotropy of the hand-arm system. The distribution of movement endpoints confirms the independence of the correction of movement length and movement direction. The mean length of movement trajectories also shows a sinusoidal shape which supports the assumption of a biomechanical cause. The mean velocity profiles show that the movements are executed in a ballistic and a correction phase. The results were integrated in a model and validated in an additional experimental study. In the experimental task rectangular target objects of different side-length ratio were displayed at a random position on the screen. On the basis of the movement time data the empirical goodness of fit was determined for eight extensions of Fitts´ Law, known from literature. The results show highest predictive power for the derived model. Furthermore, the results clearly suggest an age-differentiated modeling of Fitts´ Law because the model parameters differ significantly. Possible fields of application of the derived Fitts´ Law are the ergonomic arrangement of interaction elements on graphical user interfaces on large touch screens. Because of the similarity between target directed movements and reaching movements in manual assembly tasks the derived Fitts´ Law can be used to analyze and ergonomically optimize the spatial arrangement of material boxes.

Institutions

• Chair of Communication Studies [735410]