Skip to Main Content
Touchscreen technology has become pervasive in the consumer product arena over the last decade, offering some distinct advantages such as software reconfigurable interfaces and the removal of space consuming mice and keyboards. However, there are significant drawbacks to these devices that have limited their adoption by some users. Most notably, standard touchscreens demand the user's visual attention and require them to look at the input device to avoid pressing the wrong button. This issue is particularly important for mobile, capacitive sensing, nonstylus devices, such as the iPhone where small button sizes can generate high error rates. While previous work has shown the benefits of augmenting such interfaces with audio or vibrotactile feedback, only positive feedback (confirmation of button presses) has been considered. In this paper, we present a simple prototype interface that provides negative vibrotactile feedback. By negative, we mean feedback is generated when an inactive or ambiguous part of the screen, such as the area between two buttons, is touched. First, we present a usability study comparing positive and negative vibrotactile feedback for a benchmark numerical data entry task. The difference in performance is not statistically significant, implying negative feedback provides comparable benefits. Next, based on the experimenter's observations and the users comments, we introduce a multimodal feedback strategy-combining complementary positive audio and negative vibrotactile signals. User tests on a text entry experiment show that, with multimodal feedback, users exhibit a (statistically significant) 24 percent reduction in corrective key presses, as compared to positive audio feedback alone. Exit survey comments indicate that users favor multimodal feedback.