Predicting High Levels of Multitasking Reduces Between-Tasks Interactions
“Predicting High Levels of Multitasking Reduces Between-Tasks Interactions” by Rico Fischer and Gesine Dreisbach stated that the simultaneous handling of two tasks requires protecting the prioritized primary task from the interference of the secondary task processing. Therefore, the authors aimed to present an effective way of reducing multitasking by predicting its high levels to reduce the between-tasks interactions. The authors hypothesized that the performance of two tasks at the same time could be interrupted by secondary task processing, especially in cases when primary and secondary tasks are similar.
To test their hypothesis, researchers implemented an “item-specific proportion manipulation of temporal task overlap” (Fischer & Dreisbach, 2015, p. 1482). The experiment included 44 students (35 female and 9 male) from Technische Universitat Dresden. It is crucial to mention that all study participants had normal vision, and only one student reported left-handedness. Participants were given a specific task for completion, which included stimuli (S1 – digits 2, 3, 7, 8 and S2 – 1, 4, 6, 9), presented on a black and white monitor in 20-point Arial font.
The first group of stimuli was responded to with middle and right index fingers, while the second group of stimuli was responded to with left middle and index fingers. To respond to the given stimuli, participants were given to use a German QWERTZ keyboard. The entire experiment was presented and recorded by Presentation Software. The strength of the study was associated with the conducted experiment that gave reliable results, while, on the other hand, was limited to a relatively small sample size of only forty-four participants.
As to the results of the experiment, researchers removed the response time analyses, error trials in both tasks smaller than 150ms and larger than 3000ms. It was found that the critical interaction between item set and RC compatibility in Response Time 1 depended on the factor Block (F(1,43)=7.08, p=0.011, n2=0.14). A similar interaction was found in Response time 2 (F(1,43)=5.55, p=0.023, n2=0.11). Researchers implemented Greenhouse-Geisser correction to cases where it was appropriate.
The researcher concluded that the high temporal proximity between two similar tasks increased the interference of such tasks. The conducted experiment gave evidence of the possibility to significantly modulate the interference between two tasks using predicting the temporal correlation between Task 1 and Task 2. The study implied that the task with a small overlap in temporal indicators could potentially lead to task shielding, resulting in high interference between the given tasks. On the contrary, a larger overlap in temporal indicators can lead to the declined interference between tasks.
The findings of the research are important for providing knowledge on how to regulate cognitive control in all types of tasks, either dual and single. Previous research on single and dual tasks showed that “specific item or context features can become associated with attentional control settings that were repeatedly applied to process these items or items at these contexts” (Fischer & Dreisbach, 2015, p. 1485). However, this study showed that specific items cannot predict the conflict processing as well as control settings. Temporal task predictions specific to particular tasks’ overlaps can give knowledge not only about the risk of task interference but also on other requirements necessary for the completion of dual-tasks.
The Problem State: A Cognitive Bottleneck in Multitasking
“The Problem State: A Cognitive Bottleneck in Multitasking” by Borst, Taatgen, and van Rijn aimed to explore the challenges of multitasking and predict the method and time in which the tasks will overlap or interfere. Therefore, the researchers made a focus on ‘the problem state,’ which is a “directly accessible intermediate representation of the current state of a task” (Borst, Taatgen, & van Rijn, 2010, p. 363).
The authors put forward a hypothesis that the larger overlap in cognitive constructs among different tasks, the larger is the interference. Borst et al. (2010) gave an example of the two simultaneous tasks that use the same faculty, writing and talking, which result in great interference because of their similarity (p. 363). As to the newly proposed terminology, the problem state was defined as a resource necessary for keeping important information, which can facilitate the accomplishment of a task.
Researchers conducted three experiments: Subtraction and Text Entry, Subtraction and Text Entry – Two Responses Per Switch, and Triple Tasking. For Experiment 1, 15 students from the University of Groningen were chosen to participate. In Experiment 2, 15 students from the same University who did not participate in Experiment 1 were chosen. For Experiment 3, 22 students who did not participate in the first two experiments were chosen.
In Experiment 1, participants were asked to perform subtraction and entry tasks concurrently. In Experiment 2, the study group performed tasks similar to the first experiment; however, participants were assigned to giving answers to the task before moving to the next task. In Experiment 3, the same design of the tasks remained; although, columns with the subtractions that had already been solved were marked with the “#” sign, which prevented any strategies based on display from occurring. Furthermore, participants were asked to complete listening tasks that consisted of listening to short stories and answering multiple-choice questions at the end of each listening assignment (Borst et al., 2010, p. 374).
As to the results of the conducted experiments, Experiment 1 found an over additive interaction effect of task difficulty regarding the text entry assignments and response times. The participants of the experiment took the most time to complete hard-hard conditions. As to Experiment 2, the response time that was necessary with responses for subtractions and test entry tasks increased as the complexity of the assignments. Experiment 3 resulted in the over additive interaction effect between the difficulty of subtraction and text entry assignments. All of the three experiments were beneficial for testing the initial hypothesis of the problem state resources acting as a ‘bottleneck’ when it comes to performing different tasks at the same time. The initial hypothesis was supported by the three conducted experiments.
“The Problem State: A Cognitive Bottleneck in Multitasking” implied that the issue of the problem state bottleneck should be taken into account in real-life situations. For example, people steer the car wheel and talk on the phone at the same time. The research conducted by Borst et al. (2010) concluded that the problem state resource plays a role of the ‘bottleneck’ when it comes to multitasking (p. 379). Such a multitasking bottleneck can potentially create a negative impact on the task interference because intermediate representations of the problem state are to be maintained in the same condition for more than several seconds. Therefore, the problem state resource should be considered in the process of creating appropriate environments for multitasking.
References
Borst, J., Taatgen, N., & van Rijn, H. (2010). The problem state: A cognitive bottleneck in multitasking. Journal of Experimental Psychology, 36(2), 363-382.
Fischer, R., & Dreisbach, G. (2015). Predicting high levels of multitasking reduces between-tasks interactions. Journal of Experimental Psychology, 41(6), 1482-1487.