Implications of Quantity of Touch Screen Buttons on Response Time
An independent research project
Background
As a human factors researcher who is passionate about surface transportation, the exponential increase in the use of touch screen controls in vehicle interiors piqued my interest. While touch screen controls provide benefits such as providing a compact, easily manipulatable system, touch screen controls lack benefits of traditional vehicle controls, such as tactile controls.
Introduction
Tactile feedback plays an important role in tasks such as driving, allowing drivers to devote their attentional resources to the road and rely on tactile feedback to use controls rather than dividing their attention. In contrast, using a touch screen while driving raises safety concerns of performing two visually demanding tasks at the same time (Wickens, 2008). Vehicle manufacturers that have begun adopting fully touch screen control interfaces, such as Tesla, seem to be leaning on the aesthetic and trendiness of the futuristic, buttonless interface design. Depending on the amount of attentional resources allocated, limitations in visual attention directly impact the performance of a given task (Fernandez-Duque & Johnson, 2002). Although many car manufacturers are following this trend, it is crucial to investigate the effects it may have on driver performance and road safety.
Furthermore, larger touch screens likely equate to a larger quantity of buttons on the touch screen. Based on this observation and the concern for driver performance, I conducted an independent research study that investigated the relationship between response time and the number of buttons on an interface where tactile feedback is not available to the user.
Objectives of the Study:
- Test the hypothesis: The greater the quantity of keys/buttons available, the longer it will take for one to respond (Hick, 1952).
- Validate or challenge the findings of a previous study that concluded visual demand is equivalent for tactile and touch screen controls (Bach et al., 2008).
Methods/Procedure
A within-subjects design was employed, where 15 participants’ response times were evaluated across three treatment conditions. For the low complexity condition, I occluded all but four keyboard keys using paper. For the medium and high complexity conditions, I occluded eight keys and 12 keys, respectively. Additionally, I programmed the study to be conducted using PsychoPy, a Python-based software package for running computerized experiments. This also allowed me to disable all key inputs except the non-occluded keys on the keyboard for each respective complexity condition, simulating a touch screen interface by requiring the participant to look at the keys to press the correct one. Photos demonstrating the occlusion of keyboard keys using paper are shown below:
(1) Low complexity (4 keys of the keyboard visible)
(2) Medium complexity (8 keys of the keyboard visible)
(3) High complexity (12 keys of the keyboard visible)
For each of the three treatments, participants were prompted to press the button on the keyboard that corresponded to the character displayed on the screen as demonstrated in the photo below.
Results
A one-way ANOVA was performed to compare the effect of three different levels of touch screen complexity (number of buttons) on response time (RT). The test revealed that there was not a statistically significant difference in RT between at least two groups (F(2, 42) = [.319], p = 0.729).
Limitations
- Small sample size
- Participants were familiar with QWERTY keyboard
- Random assortment of keys would ensure familiarity of QWERTY is not a confounding variable
This study was presented at the California State University of Long Beach’s Human Factors & Ergonomics Society Student Chapter (HFES) 19th Annual Research Conference (March 2024)
