Assessing cross-modal interference in the detection response task

https://doi.org/10.1016/j.jmp.2020.102390Get rights and content

Highlights

  • Auditory and visual detection response tasks (DRT) were presented with tracking task.

  • DRT response times were assessed using shifted-Wald model.

  • Continuous tracking task data was transformed to discrete form to allow RT analysis.

  • DRT presence, but not modality of DRT stimulus, interfered with tracking task.

  • Higher tracking task load led to lower processing efficiency on both tasks.

Abstract

The detection response task (DRT) is a measure of workload that can assess the cognitive demands of real-world multitasking. It can be configured to present simple stimuli of several modalities, including auditory and visual signals. However, the concurrent presentation of the DRT stimuli alongside another task could cause dual-task interference, and the extent of this interference could be different based on the DRT’s configuration. It is necessary to consider the characteristics of the DRT stimulus, such as modality, to identify a minimally intrusive stimulus. Fifty participants completed a computer-based one-dimensional tracking task alongside a DRT. The DRT’s stimuli varied in their modality (visual/auditory), while the tracking task varied in its workload demand (low/high). DRT performance was modelled using a shifted-Wald model, while the tracking task was assessed using systems factorial technology (SFT), a non-parametric methodology capable of capturing a cognitive system’s workload capacity. To allow the latter’s use, we developed a method of transforming continuous tracking data into a discrete form akin to response times. Analysis of DRT data found little evidence that the DRT’s modality affected processing efficiency, while SFT analysis found limited-capacity processing on the tracking task across both DRT modalities. These findings suggest DRT modality had little effect on the level of interference between the two tasks.

Section snippets

Background

Multi-tasking has a deleterious effect on performance due to people’s limited capacity for processing information (Kahneman, 1973). Tasks that require more resources interfere to a greater extent than others, while automated tasks require no resources, and therefore have no impact on other tasks (Wickens, 2002). However, a unitary model of processing capacity where all tasks draw on a single pool of resources cannot explain findings relating to multi-modal multi-tasking—situations in which

Participants

Fifty undergraduates from the University of Newcastle (F = 25, M = 25) participated in the study. Mean age of participants was 22.4 years (SD = 5.6 years). Participants were remunerated with course credit. The study was approved by the University of Newcastle Human Research Ethics Committee.

Design

The task had two independent variables: primary task load, which was manipulated by changing the number of targets to be tracked and had three levels (High load — two targets to track simultaneously, LowLEFT

Results

Seven participants were excluded due to technical malfunction, low DRT accuracy (¡33% overall hit rate), and non-responses to primary tracking task trials.

Discussion

The DRT produced results in line with those expected from previous studies, with faster responses under conditions of low load than high load. The result of faster responses to high-salience stimuli compared to low-salience stimuli was also predicted. The primary finding from mean DRT RT was that participants who were presented an auditory signal responded faster than those who were presented a visual signal. This could be a result of the auditory pathway generally producing faster RTs than the

Conclusion

Cognitive models were applied to both a continuous primary tracking task and a software-based DRT to assess the differential impact of an auditory and visual DRT stimulus on user workload. The analysis of primary tracking task performance utilised SFT in a novel application, subjecting continuous tracking data and transformed pseudo-RT distributions to capacity analysis. Both measures agreed in their finding of limited capacity. A shifted-Wald model was applied to DRT data, which identified a

References (47)

  • CohenJ.

    Statistical power analysis for the behavioral sciences

    (1988)
  • FificM. et al.

    Stretching mental processes: an overview of and guide for SFT applications

  • GibneyK.D. et al.

    Visual distractors disrupt audiovisual integration regardless of stimulus complexity

    Frontiers in Integrative Neuroscience

    (2017)
  • HaapalainenE. et al.

    Psycho-physiological measures for assessing cognitive load

    Proceedings of the 12th ACM international conference on ubiquitous computing

    (2010)
  • HeathR.

    Nonlinear dynamics: Techniques and applications in psychology

    (2000)
  • HeathcoteA.

    Fitting wald and ex-wald distributions to response time data: An example using functions for the S-PLUS package

    Behavior Research Methods, Instruments, & Computers

    (2004)
  • HeathcoteA. et al.

    Working memory’s workload capacity

    Memory and Cognition

    (2015)
  • HouptJ.W. et al.

    Systems factorial technology with r

    Behavior Research Methods

    (2014)
  • HouptJ.W. et al.

    Statistical analyses of the resilience function

    Behavior Research Methods

    (2017)
  • HsiehL. et al.

    A surrogate test for cognitive demand: Tactile detection response task (TDRT)

  • InnesR. et al.

    Cognitive workload measure and analysis

    (2018)
  • Road vehicles. transport information and control systems. Detection-response task (DRT) for assessing attentional effects of cognitive load in driving

    (2016)
  • JeffreysH.

    Theory of probability

    (1961)
  • Cited by (0)

    This research was supported by an Australian Research Council grant DP160102360 to AE and JTT, and by an Australian Government Research Training Program (RTP) Scholarship awarded to Alexander Thorpe. We thank Mario Fific and an anonymous reviewer for their helpful feedback on earlier versions of this manuscript. We thank Bryan Paton for his invaluable contribution.

    View full text