Accuracy of non-visual directional pointing with various manual input devices

In a self-testing vision screener, examinees use an input device for pointing the orientation of the targets, which are presented inside the vision screener. Examinees operate the input device without visual feedback. In the present study, the suitability of pointing devices was evaluated for conditions such as are present in a self-testing vision screener. The evaluation consisted of an experimental assessment of pointing accuracy and recording subjective ratings while using the various devices. Six commercially available computer input devices – a joystick, a gamepad, a trackball, two track pads and a PC mouse - were evaluated under visual conditions similar to those that would be present when using a self-testing vision screener.Pointing accuracy was found to vary significantly with the type of device (F(3.2, 93.1) = 3.937, p = 0.009) and the effect of the device on pointing accuracy was important (partial η² = 0.120). The most accurate pointing was achieved when participants used the joystick. Using the joystick, a mean of 96.8% (SD = 4.3%) of pointing trials resulted in the correct orientation. If only diagonal orientations are considered, the correct pointing rate increased to a mean of 99.5% (SD = 1.5%) when using the joystick.In terms of the subjective ranking, the gamepad and the joystick achieved the best and the second best ranks respectively, whereas the trackball was the least preferred device.Based on our findings, we recommend using a joystick as an input device in pointing tasks in order to minimize the effects of suboptimal visual feedback on motor performance. As for the particular case of testing visual acuity, various procedures are suggested. Thus, the effect of suboptimal visual feedback on the outcome of the acuity test is reduced.