Abstract
Modern mobile devices provide a convenient platform for collecting research data in the field. But because the working of these devices is often cloaked behind multiple layers of proprietary system software, it can be difficult to assess the accuracy of the data they produce, particularly in the case of timing. We have been collecting data in a simple visual reaction time experiment, as part of a fatigue testing protocol known as the Psychomotor Vigilance Test (PVT) (Dinges et al., 1994). In this protocol, subjects run a 5-minute block consisting of a sequence of trials in which a visual stimulus appears after an unpredictable variable delay. The subject is required to tap the screen as soon as possible after the appearance of the stimulus. In order to validate the reaction times reported by our program, we had subjects perform the task while a high-speed video camera recorded both the display screen, and a side view of the finger (observed in a mirror). Simple image-processing methods were applied to determine the frames in which the stimulus appeared and disappeared, and in which the finger made and broke contact with the screen. The results demonstrate a systematic delay between the initial contact by the finger and the detection of the touch by the software, having a value of 80 +− 20 milliseconds.