Foundations for designing and evaluating user interfaces based on the crossing paradigm

Abstract

Traditional graphical user interfaces have been designed with the desktop mouse in mind, a device well characterized by Fitts' law. Yet in recent years, hand-held devices and tablet personal computers using a pen (or fingers) as the primary mean of interaction have become more and more popular. These new interaction modalities have pushed the traditional focus on pointing to its limit. In this paper we explore whether a different paradigm-goal crossing-based on pen strokes-may substitute or complement pointing as another fundamental interaction method. First we describe a study in which we establish that goal crossing is dependent on an index of difficulty analogous to Fitts' law, and that in some settings, goal crossing completion time is shorter or comparable to pointing performance under the same index of difficulty. We then demonstrate the expressiveness of the crossing-based interaction paradigm by implementing CrossY, an application which only uses crossing for selecting commands. CrossY demonstrates that crossing-based interactions can be more expressive than the standard point and click approach. We also show how crossing-based interactions encourage the fluid composition of commands. Finally after observing that users' performance could be influenced by the general direction of travel, we report on the results of a study characterizing this effect. These latter results led us to propose a general guideline for dialog box interaction. Together, these results provide the foundation for the design of effective crossing-based interactions.

Visual aids in 3D point selection experiments

“…Their extremely small depth extent also makes them difficult to select with touch-based techniques; it is unlikely that participants are able to reliably intersect a flat disc target. For discs, a crossing paradigm [2] is likely more appropriate, but would prevent comparison with other pointing studies.…”

Design‐based research on the use of a tangible user interface for geometry teaching in an inclusive classroom

“…Mouse and keyboard are the traditional input devices for interaction with computers, enabling wide possibilities for human interaction, but limitations of the mouse as an input device require adaptation of motor skills for specific problem solving. The keyboard and mouse are not always adequate substitutions for a pen and paper (Apitz, Guimbretière & Zhai, ). The computer mouse is used to access multiple graphical user interface (GUI) controls in sequence (Fitzmaurice & Buxton, ; Tuddenham, ).…”

Manipulating multiple sliders by crossing

“…Crossing-based interfaces let users invoke commands by crossing (or painting over) widgets instead of clicking them [3]. Crossing is used to drag and drop icons in background windows [13] and to successively select elements in hierarchical menus [2].…”

“…Finally, the most efficient crossing gestures are continuous and performed orthogonally to the crossed object [1,3]. However, crossing requires performing a steering task (e. g., to cross multiple checkboxes), and steering in a narrow tunnel (dragging along a precise path) is a slow motor task [1,3].…”

“…However, crossing requires performing a steering task (e. g., to cross multiple checkboxes), and steering in a narrow tunnel (dragging along a precise path) is a slow motor task [1,3].…”