Comparing Intensity and Location-Based Tactile Change Detection Performance

Abstract

Transmitting information in the tactile channel is a promising solution in complex domains such as aviation and healthcare because it can offload the often oversaturated visual and auditory channels. While the tactile modality is a promising means of presenting information, the phenomenon of change blindness has been repeatedly demonstrated in the tactile modality. To deepen our understanding of the limitations of tactile information processing it is critical to fully understand how change blindness affects the sense of touch. This study directly compared location-based (i.e., when there is a change in the location of the tactile cue) and intensity-based (i.e., when there is a change in tactile cue intensity) change detection to examine whether changes in location and intensity can be considered equally susceptible to change blindness.

Objectives

Transmitting information in the tactile channel is a promising solution in complex domains such as aviation and healthcare because it can offload the often oversaturated visual and auditory channels. As such, the use of vibrotactile displays can potentially increase mental bandwidth and improve multitasking performance. While the tactile modality is a promising means of presenting information, the phenomenon of change blindness (i.e., the failure to detect changes between stimuli) has been repeatedly demonstrated in the tactile modality. Change blindness is thought to result from the failure to accurately store and compare a set of stimuli in the working memory, possibly due to limited attentional resources, capacity, or faulty stimulus encoding.

Prior work on change blindness has focused mainly on vision and audition, with limited work on how the phenomenon affects touch. To deepen our understanding of the limitations of tactile information processing it is critical to fully understand how change blindness affects the sense of touch. Several studies have examined tactile change detection under various parameters, including the number of body locations stimulated, cue complexity, movement, distractions between stimuli, and vibrational intensity. However, no study to-date has directly compared location-based (i.e., when there is a change in the location of the tactile cue) and intensity-based (i.e., when there is a change in tactile cue intensity) change detection. The present study uses a simple stimulus apparatus to examine whether changes in location and intensity can be considered equally susceptible to change blindness.

Approach

Participants were presented with two one-shot change detection tasks using 10 mechanical tactors attached to the fingertips of the left and right hands with hypoallergenic adhesive rings. The experiment consisted of 160 trials separated into two blocks (block 1 = location changes, block 2 = intensity changes). Each trial consisted of a pair of tactile stimuli presented for 500 ms each, separated by a 500 ms blank interstimulus interval. Tactile stimuli could appear on any two fingers of the same hand and were applied at a consistent 100 Hz frequency. For location changes, intensity was held constant at 0.3 .WAV sound file amplitude while the location of one of the two fingers could change. For intensity changes, three levels of change trial intensity based on .WAV sound file amplitude were possible (0.1 = low intensity change, 0.3 = baseline/no change, 0.9 = high intensity change) while finger locations were held constant. Participants indicated whether they felt a change in tactile cues using foot pedals (left foot pedal = change, right foot pedal = no change). Participant response times were limited to a maximum of 5 s for each trial and trials without a response were removed from further analysis. Trials within each block were presented in random order with equal probability. An equal number of trials were presented to the left and right hands, with changes present in 50% of the trials.

Findings

A series of repeated measures one-way ANOVAs were performed to analyze the effect of change type (location or intensity) on sensitivity (d’), response bias (c), and reaction time. The results indicate that sensitivity (d’) is significantly different between changes in location and intensity, with changes in location being more discriminable than changes in intensity (p < .001). Response bias (c) was also significantly different between change types (p < .001). Subjects adopted a more conservative response strategy for changes in tactile intensity. There was a significant main effect of change type on reaction time (p < .001). Post hoc tests were conducted on the main effect of change type and subject to a Bonferroni adjustment. Results showed that reaction time was significantly higher for location changes than intensity changes (p < .001). On average, response times were 130 ms slower for location-based changes than intensity-based changes. These findings may reflect that changes in intensity are faster and potentially easier for humans to process compared to changes in location.

Takeaways

The findings of this work can inform the future design of vibrotactile displays in a variety of domains. In tactile display design, this study shows that tactile location changes take longer to process and react to than tactile intensity changes. Changes in tactile location were responded to more liberally than changes in tactile intensity. Designers should weigh the consequences of false positives and false negatives when deciding which type of tactile cue to deploy based on the response strategies observed. A significant difference in average reaction times was demonstrated that could have severe implications for operators of vibrotactile displays, especially in safety - critical domains. Tactile location changes took, on average, 130 ms longer to react to than tactile intensity changes. Increases in tactile cue complexity—for example, using combinations of changes in intensity and location—ideally should be avoided to minimize a further increase in reaction time.