The cone and the lazy bubble

Abstract

We evaluated two cursor designs in the continuum between the traditional point cursor and the bubble cursor by Grossman and Balakrishnan. The lazy bubble cursor expanded to envelop the closest target when the ratio of the distances to the closest and the second closest target was less than 1:2. In addition to this lazy behavior the cone cursor had a tail that stayed on the last enveloped target until the next target was enveloped. In an experiment with 18 participants we found that the bubble cursor was faster than our cursors that had smaller target activation areas but the difference remained very small. Of the bubble cursor variants the lazy bubble exhibited higher error rate than the other two. Thus, the winners on the objective metrics were the bubble cursor and the cone cursor. The lazy bubble cursor and the bubble cursor were preferred in subjective ratings.

Additive Voronoi Cursor: Dynamic Effective Areas Using Additively Weighted Voronoi Diagrams

“…Starburst [3] used a different partitioning method, which adapted to clustered targets with non-uniform distributions. Several variations of the Bubble cursor have been proposed [16,20]. Recently, speed-dependent area-cursor techniques [8,28] were proposed.…”