Exploring Methods to Improve Pen-Based Menu Selection for Younger and Older Adults

A Comparison of Touchscreen and Mouse for Real-World and Abstract Tasks with Older Adults

Computer technology is increasingly being used to facilitate the timely identification of cognitive impairment in older adults. Our Cognitive Testing on Computer (C-TOC) project aims to develop a self-administered online test for older adults to take at their home. Due to the freedom of devices they can use, it is important to investigate whether different input devices can impact test performance. We compared touchscreen and mouse input on both abstract and real-world pointing and dragging tasks: classic Fitts’s Law tasks and tasks drawn from C-TOC. The abstract and real-world tasks were designed to require equivalent motor skills. Our research goals were to determine (1) if performance on computerized cognitive tasks are affected by input device, and (2) if performance differences due to input device can be explained by those observed on Fitts’s Law tasks. Sixteen older adults completed both types of tasks using a touchscreen and a mouse. We found that input device affected speed on three out of four cognitive tasks while only affecting accuracy on one task. Secondarily, our results suggest that Fitts’s Law results of differences in mouse and touch cannot be used to predict device differences in the performance on C-TOC tests. As an additional research goal, we looked into the movement patterns of one real-world dragging task—the C-TOC Pattern Construction task—to see if they could provide richer performance measures, beyond speed and accuracy. Such measures could compensate for the lack of a clinician observer who is typically present in comparable paper-based cognitive tests. We found that older adults naturally adopted different movement patterns on the two devices: they tended to make shorter moves and a greater number of moves on a touchscreen than with a mouse. Altogether, our results suggest that careful device-based performance calibration will be needed in computerized tests.

Empirical evaluation of circling interface with head-mounted mouse emulator users

PURPOSE

To evaluate the performance of the circling interface, which is an alternative interaction method for selecting and manipulating on-screen objects based on circling the target, rather than pointing and clicking.

METHOD

We conducted empirical evaluations with actual head-mounted mouse emulator users from two different groups: individuals with spinal cord injury (SCI) and individuals with cerebral palsy (CP), comparing each group's performance and satisfaction level on pointing tasks with the circling interface to performance on the same tasks when using dwell-clicking software.

RESULTS

Across all operations, for both subjects with SCI and with CP, the circling interface showed faster performance than the dwell-clicking interface. For the single-click operation, the circling interface showed slower performance than dwell selection, but for both double-click and drag-and-drop operations, the circling interface produced faster performance. Subjects with CP required much longer time to complete the tasks compared to subjects with SCI. If errors caused by circling on an area with no target and unintentional circling caused by jerky movements and an abnormally tiny circle are automatically corrected by the circling interface, their performance accuracy with the circling interface outperformed existing solutions without a steep learning curve.

CONCLUSIONS

Circling interface can be used in conjunction with existing techniques and this kind of combined approach achieve more effective mouse use for some individuals with pointing problems. It is also expected to be useful for both computer access and augmentative communication software. Implications for Rehabilitation A circling interface will improve clinical practice by providing an alternative pointing method that does not require physically activating mouse buttons and is more efficient than dwell-clicking. Being used in conjunction with existing techniques, some individuals who are head mouse users can achieve more effective mouse use. The Circling interface can also work with AAC devices.

Effects of Target Expansion on Selection Performance in Older Computer Users

Point and click interactions using a mouse are an integral part of computer use for current desktop systems. Compared with younger users though, older adults experience greater difficulties performing cursor positioning tasks, and this can present limitations to using a computer easily and effectively. Target expansion is a technique for improving pointing performance where the target grows dynamically as the cursor approaches. This has the advantage that targets conserve screen real estate in their unexpanded state, yet can still provide the benefits of a larger area to click on. This article presents two studies of target expansion with older and younger participants, involving multidirectional point-select tasks with a computer mouse. Study 1 compares static versus expanding targets, and Study 2 compares static targets with three alternative techniques for expansion. Results show that expansion can improve times by up to 14%, and reduce error rates by up to 50%. Additionally, expanding targets are beneficial even when the expansion happens late in the movement, that is, after the cursor has reached the expanded target area or even after it has reached the original target area. The participants’ subjective feedback on the target expansion are generally favorable, and this lends further support for the technique.