Examining Change Sensitivity to Vibrotactile Beats in a Hand-Held Touchscreen Device
- Additional Document Info
- View All
Copyright 2017 by Human Factors and Ergonomics Society. While recent research has employed vibrotactile feedback as a means of communication, one novel form of vibrotactile feedback involves the generation of “beats”. They are amplitude-modulated vibratory signals that can be created by sending multiple sinusoidal signals at dissonant frequencies (Lim, Kyung, & Kwon, 2012; Yang et al., 2014). The resulting perception of a rising-and-falling amplitude signal (a single “beat”), can be characterized per unit time as beat frequency, which is a function of the difference between the two input signal frequencies. Although vibrotactile beat cues have potentials in better supporting multitasking contexts that are visually demanding, the fundamental psychophysical characteristics of absolute and difference sensitivities have not been well-studied. To build on the promising but sparse findings involving the application of vibrotactile beats, it is important to define the limits of human perceptual ability to differentiate vibrotactile beats at distinct beat frequencies. This study employed a psychophysical method of adjustment to examine the human ability to perceive change in vibrotactile beat presentations with a touchscreen tablet. Touchscreen interfaces may invoke heavy demands on visual processing resources and thus would benefit from employing nonvisual modality in user interaction. Specialized equipment and in-house software were used to present vibrotactile beats in a range of beat frequency, with the smallest change interval that the equipment can support. Experimental tasks involved fifteen participants interacting with the touchscreen in 1) distinct contexts, 2) adjusting the direction of finger movement to drive 3) changes of beat frequency within a range of beat frequency. Table 1 further describes the manipulation of independent variables. Any and all perceived changes were recorded. Participants “swiped” a fingertip along the vertical axis of the touchscreen to perceive and adjust vibrotactile beats, as shown in Figure 1.(Table presented) (Figure presented) The findings suggest that the smallest JND (Just Noticeable Difference) was observed when seated participants experienced an increasing sequence of beat frequency from the lowest. In contrast, the JND observed when the initial beat frequency changed to the highest was considerably larger. A statistical test confirmed this pattern; for seated participants, BEAT (whether the signal was increasing from the lowest or decreasing from the highest in the indicated direction) significantly affected the JND. The Weber-Fechner law corroborates these findings in that the initial beat frequency significantly affected the perception ability of vibrotactile beats. The significant BEAT influence indicates that the initial beat frequency that users first encounter during a haptic interaction should be smaller in order to support smaller JNDs and thus greater perceptual resolution. However, the perception of changes in vibrotactile beats was contingent upon the nature of touchscreen interaction. FINGER (the direction of finger movement) and CONTEXT (whether participants were seated or walking) interacted significantly, showing perceptual ability depended on task factors that impacted the speed of exploration, both of which have impacted the perception of vibrotactile beats. The difference due to these task factors is dramatic when working with the given the range of beat frequency. Therefore, the practicality of vibrotactile beats would be based on the context of use to which vibrotactile beats are applied. The results of this study will directly apply to a set of practical design guidelines for employing vibrotactile beats in hand-held touchscreen applications.
author list (cited authors)