Cognitive + Motor Effort Study

The recent decade’s surge in interaction with digital environments, characterized by increased screen time and multiple task options (e.g., email, web browsing, media), necessitates an examination of the decision-making processes governing task switching in these multitask environments. As human-centered design increasingly incorporates technology becomes progressively more immersive, interaction design moves towards more natural, familiar, gesture based commands, such as reaching and pointing. By quantifying effort as a measure of natural reaching distance, we can measure the influence of physical effort on task choice in a manner that’s particularly relevant to emerging technologies such as virtual reality. This study aims to explore how motor and cognitive efforts influence task selection decisions within digital platforms.

The Voluntary Task Switching (VTS) paradigm is designed to analyze volitional behavior in dynamic multitasking environments (Arrington, 2008). Subjects select which task to perform allowing for the consideration of top-down objectives and bottom-up processing such as stimulus availability on decision making. In our current research, we investigated the effect of motor effort on voluntary task selection by manipulating the physical reaching distance required to respond to the two tasks. Participants instructed to select tasks “randomly” should not be biased by stimulus-response positions if task selection is purely top-down. Therefore an impact of stimulus-response positions on task selection would indicate that voluntary behavior is influenced by the motor effort demanded by a task response.

Method:
20 right-hand-dominant subjects completed a VTS paradigm in which a number and letter appeared on the screen on every trial and participants had to select which task to perform with instructions to perform the tasks randomly. The stimuli for the even/odd task were the numbers 2–9 and for the consonant/vowel task were the uppercase letters A, B, C, E, I, L, U, and W. The stimuli appeared in eight possible locations, four on the left and four on the right of a central fixation point. Position 1 was located closest to the fixation point and 4 furthest away. For each trial, one stimulus would be in position 1, while on the other side of the fixation point, there was an equal probability of the stimulus being in any of the 4 positions. Stimulus identity and position were manipulated randomly. In addition to the target stimuli, the display screen contained four response boxes (labeled “con,” “vow,” “even,” and “odd”), one above and one below each target. As stimuli moved from position 1 to 4, the response boxes moved diagonally towards the corners of the display, maintaining vertical alignment, but increasing Y-axis distance from the stimulus. As such, position 1 required the shortest reaching distance to respond, and position 4 required the furthest. The stimuli remained on-screen until a response was collected via touchscreen display. Time between trials was 200 ms. Subjects repeated 14 blocks of 48 randomized trials.

Results:
The primary analysis of the task choice results looked at the probability of performing the task associated with a particular stimulus. Data were trimmed by removing the first trial of each block. We performed a 4 (position: 1, 2, 3, or 4) X 2 (side: left or right) repeated-measures ANOVA on the task choice data. The mean proportion of trials where a task was performed on a stimuli is provided here for each cell in the design, moving from the left to right of the screen: L4 = 0.34, L3 = 0.34, L2 = 0.41, L1 = 0.44, R1 = 0.56, R2 = 0.50, R3 = 0.44, R4 = 0.35. While the pattern of responding showed a bias toward performing the task associated with the stimulus on the right, the main effect of side was not significant, F(1, 18) = 2.2, p > .1, η2p = 0.11. There was a significant effect of position in which participants selected to perform the task associated with a stimulus with decreasing frequency as the stimulus moved further from fixation, F(3, 54) = 23.8, p < .01, η2p = 0.55. The effect of position entered into a significant interaction with side, such that the decrease in task selection as a function of distance was greater for stimuli on the right of the screen compared to those on the left, F(3, 54) = 4.5, p < .01, η2p = 0.20. This effect may be a result of a floor effect for stimuli on the left of the screen, which was the side opposite from their dominant hand with which responses were being made.

Conclusions:
Our research in the VTS paradigm suggests that decisions regarding selection and transition between tasks in a multitask environment appear to be sensitive to manipulations in motor effort demand. Despite the instruction to perform tasks randomly, a discernable pattern emerged, where subjects predominantly opted for tasks requiring less motor effort, evident in their more frequent selection of tasks associated with stimuli positioned closer to the fixation point. These results complement the findings of Arrington (2008) in which stimulus availability in terms of stimulus onset times biased task choice toward the more available stimulus. Our findings indicate a nuanced interplay between top-down objectives and bottom-up processing, challenging the notion that voluntary task selection is purely driven by cognitive strategies. Overall, these results point towards a significant effect of motor effort demand on task selection.

In light of these observations, our study contributes to a deeper understanding of the factors influencing voluntary task selection in dynamic multitask environments. It calls for an integrated approach to human factors engineering, such that physical aspects of task selection are given comparable priority to cognitive factors. These data could have significant implications for designing more effective and ergonomic multitasking environments.

• • Eastern Psychological Association (2024)
  • Institute for Data, Intelligent Systems, and Computation at Lehigh University (2024)
  • Lehigh University Honors Symposium (2024)