The Effects of Experience and Strategy on Visual Attention Allocation in an Automated Multiple-Task Environment

Operators and users interacting with computer environments often have to deal with multiple tasks at once, responding to each in series. Diagnostic automation, that is, automation that alerts users when and where to look, has been suggested to support the unique challenges of multiple task environments: activating tasks, switching between tasks, and tasks interfering with each other. Automation is not always reliable, however. Because of the common interaction with novel systems and the importance of training, the Simultaneous Task Environment Platform program—a multiple-task environment—was developed to understand the effects of experience on interaction with these automation-supported systems, as well as what strategies were developed. It was found that participants became more efficient with experience only when they interacted with higher reliability automation. Furthermore, the strategies participants developed focused on the differences between tasks and patterns across those tasks. Automated systems training should be sure to employ these findings.     

Age-related performance in a multiple-task environment.

Younger and older adult participants performed a dynamic multiple-task requiring concurrent processing of 4 independent tasks. Component-task performance emphasis (i.e., task priorities) was biased by differential point allocations across task components. After training, the point structure was modified. Older adults exhibited larger multiple-task performance deficits compared with younger adults; however, the age-related gap in multiple-task performance decreased with practice. The age-related performance difference increased again when task emphasis was changed, but not when demands were changed. Consistent with the training data, the age-related differences diminished again with additional experience on this new task-component emphasis. The data suggest that higher-order, strategic processing may be an important source of age-related differences in complex multiple-task performance. Actual or potential applications of this research include the facilitation of techniques for age-related comprehensive usability testing for products of even moderate complexity.     

Human performance in a multiple-task environment: Effects of automation reliability on visual attention allocation

Diagnostic automation has been posited to alleviate the high demands of multiple-task environments; however, mixed effects have been found pertaining to performance aid success. To better understand these effects, attention allocation must be studied directly. We developed a multiple-task environment to study the effects of automation on visual attention. Participants interacted with a system providing varying levels of automation and automation reliability and then were transferred to a system with no support. Attention allocation was measured by tracking the number of times each task was viewed. We found that participants receiving automation allocated their time according to the task frequency and that tasks that benefited most from automation were most harmed when it was removed. The results suggest that the degree to which automation affects multiple-task performance is dependent on the relative attributes of the tasks involved. Moreover, there is an inverse relationship between support and cost when automation fails.     

Detecting transient changes in dynamic displays: the more you look, the less you see

OBJECTIVE: Two experiments examined the detectability of transient changes in cluttered and dynamic displays and optimal scan strategies for performance. BACKGROUND: Research has demonstrated that onset changes are prioritized by the attention system and onsets are often used to signal important display changes. However, research has mostly used uncluttered, static displays and has largely ignored the role of scan strategy. METHOD: We had participants monitor a cluttered and dynamic display and respond when an object changed color or onset abruptly. Scan strategies were also evaluated. In another experiment participants were instructed to use particular strategies to detect changes. RESULTS: Consistent with previous results, onset events were detected better than were color change events. Scan strategy accounted for a large proportion of variance in detection performance. Participants who made few eye movements performed best. Participants who actively scanned the display performed worst. When poor performers were instructed to make few eye movements, their performance matched that of the best performers. CONCLUSION: Onset alerts can be an effective means of signaling important events in complex displays. Additionally, scan strategy plays an important role in the detection of transient events. APPLICATION: These results have important implications for training operators to search for transient changes within dynamic and cluttered displays.     

Height effects in real and virtual environments

The study compared human perceptions of height, danger, and anxiety, as well as skin conductance and heart rate responses and postural instability effects, in real and virtual height environments. The 24 participants (12 men, 12 women), whose average age was 23.6 years, performed "lean-over-the-railing" and standing tasks on real and comparable virtual balconies, using a surround-screen virtual reality (SSVR) system. The results indicate that the virtual display of elevation provided realistic perceptual experience and induced some physiological responses and postural instability effects comparable to those found in a real environment. It appears that a simulation of elevated work environment in a SSVR system, although with reduced visual fidelity, is a valid tool for safety research. Potential applications of this study include the design of virtual environments that will help in safe evaluation of human performance at elevation, identification of risk factors leading to fall incidents, and assessment of new fall prevention strategies.     

Favor-based decision: A novel approach to modeling the strategy diffusion in causal multiagent societies

In previous work on collective motion, agents always tend to imitate the behavior strategies of higher ranks; this model is called rank-based strategy diffusion. Unfortunately, this model is, by itself, insufficient in causal multiagent societies where agents may have causal links with each other. In causal environments, agents will develop positive (or negative) attitudes (favor) about those who can increase (or decrease) their own utilities. Naturally, for collective motion, agents will be inclined to imitate those who are well-favored and avoid those who are disfavored. This paper presents the concept of favor in causal environments, and presents a model for favor-based strategy diffusion. In the proposed model, agents in causal environments are inclined to associate with and imitate the strategies of those who are well-favored. Obviously, such diffusion effects well reflect the impact of causal relations in the real world.     

Evolutionary Dynamics of Strategies without Complete Information on Complex Networks

As for the behaviors of multi‐agent system, recent years have witnessed the growing interest in the study of cooperative behaviors by the aid of evolutionary game dynamics on complex networks. Thereinto, the updating rules deciding the evolution of strategies will significantly influence the steady state distribution of the system. The strategy updating rooted in the pursuit of larger benefits, will drive the system to evolve into the coexistence of different states or the domination by some strategies. To relax the often‐used rules required explicit knowledge of the exact payoffs, this paper describes a new approach of updating strategy based on switching probabilities, which is independent on players' payoffs and degrees. And then the equilibrium state of the strategy evolution in the networks is studied. Our work here provides a computationally feasible way of estimating the steady characteristics of the strategy adoption of agents situating on complex networks. The stability analysis elucidates two important features: (i) the takeover of cooperation can be enhanced by the appropriate settings of the switching probabilities between strategies and (ii) larger average degree and power exponent in the employed scale‐free network can make it easier for the coexistence of strategies. The results can help the design of initial strategy distribution of agents located on social networks to promote cooperation.     

Predicting vigilance: A fresh look at an old problem

Individual differences in vigilance are ubiquitous and relevant to a variety of work environments in industrial, transportation, medical and security settings. Despite much previous work, mostly on personality traits, it remains difficult to identify vigilant operators. This paper reviews recent research that may point towards practically useful predictor variables for vigilance. Theoretical approaches to identifying predictors that accommodate the heterogeneous nature of vigilance tasks are compared. The article surveys recent empirical studies using personality measures, ability tests and scales for stress and coping as predictors of vigilance. Promising new constructs include trait scales linked to fatigue, abnormal personality and the stress state of task engagement. Implications of the data reviewed for occupational selection are discussed. Selection should be based on a multivariate assessment strategy, cognitive task analysis of the operational vigilance task and use of work sample measures to capture typical stress responses to the task. This review paper surveys recent research that may point towards practically useful predictor variables for vigilance. The article surveys recent empirical studies using personality measures, ability tests and scales for stress and coping as predictors of vigilance. Selection should be based on a multivariate assessment strategy.     

Effect of varying levels of mental workload on startle eyeblink modulation

Previous research using punctuate reaction time and counting tasks has found that the startle eyeblink reflex is sensitive to attentional demands. The present experiment explored whether startle eyeblink is also modulated during a complex continuous task and is sensitive to different levels of mental workload. Participants (N = 14) performed a visual horizontal tracking task either alone (single-task condition) or in combination with a visual gauge monitoring task (multiple-task condition) for three minutes. On some task trials, the startle eyeblink reflex was elicited by a noise burst. Results showed that startle eyeblink was attenuated during both tasks and that the attenuation was greater during the multiple-task condition than during the single-task condition. Subjective ratings, endogenous eyeblink rate, heart period, and heart period variability provided convergent validity of the workload manipulations. The findings suggest that the startle eyeblink is sensitive to the workload demands associated with a continuous visual task. The application of startle eyeblink modulation as a workload metric and the possibility that it may be diagnostic of workload demands in different stimulus modalities is discussed.     

Partitioning of Independent Tasks for Minimizing Completion Time and Total Waiting Time

Parallel processors provide fast computing environments for various users.But the real efficiencies of parallel processors intensively depend on the partitioning strategies of tasks over the processors.In this paper,the partitioning problems of independent tasks for homogeneous system of parallel processors are quantitatively studied.We adopt two criteria,minimizing the completion time and the total waiting time, to determine the optimal partitioning strategy.     

Effect of varying levels of mental workload on startle eyeblink modulation

Previous research using punctuate reaction time and counting tasks has found that the startle eyeblink reflex is sensitive to attentional demands. The present experiment explored whether startle eyeblink is also modulated during a complex continuous task and is sensitive to different levels of mental workload. Participants (N = 14) performed a visual horizontal tracking task either alone (single-task condition) or in combination with a visual gauge monitoring task (multiple-task condition) for three minutes. On some task trials, the startle eyeblink reflex was elicited by a noise burst. Results showed that startle eyeblink was attenuated during both tasks and that the attenuation was greater during the multiple-task condition than during the single-task condition. Subjective ratings, endogenous eyeblink rate, heart period, and heart period variability provided convergent validity of the workload manipulations. The findings suggest that the startle eyeblink is sensitive to the workload demands associated with a continuous visual task. The application of startle eyeblink modulation as a workload metric and the possibility that it may be diagnostic of workload demands in different stimulus modalities is discussed.     

Are posture data from simulated tasks representative of field conditions? Case study for overhead electric utility workers

Many ergonomics studies are conducted in laboratory-simulated work environments to assess risks for the development of musculoskeletal disorders under more controlled conditions. However, the simulated conditions could be of questionable validity with respect to reproduction of field conditions involving risk factors. The objective of this study was to verify whether the postures recorded for neck extension/flexion and upper arm elevation of overhead electric utility workers in a simulated environment were similar to those recorded in a field environment. Of the three frequently performed tasks analysed, two presented similar exposure in both conditions. However, differences were identified for a more complex task (relay replacement). These results suggest that simulated tasks may be more representative for more standardised tasks. This may indicate that researchers investigating risks should avoid simplifying complex tasks when reproducing field posture exposure in laboratories, since omitting extra subtasks may lead to an inaccurate reproduction of field exposure.

Practitioner Summary: Studies comparing results from field and simulated environments are necessary to evaluate to what degree postural exposure reproduced in laboratory is representative of the exposure occurring in real work situations. This is particularly relevant for tasks involving training in simulated environment due to safety constraints.     

Generalized exploration in policy search

To learn control policies in unknown environments, learning agents need to explore by trying actions deemed suboptimal. In prior work, such exploration is performed by either perturbing the actions at each time-step independently, or by perturbing policy parameters over an entire episode. Since both of these strategies have certain advantages, a more balanced trade-off could be beneficial. We introduce a unifying view on step-based and episode-based exploration that allows for such balanced trade-offs. This trade-off strategy can be used with various reinforcement learning algorithms. In this paper, we study this generalized exploration strategy in a policy gradient method and in relative entropy policy search. We evaluate the exploration strategy on four dynamical systems and compare the results to the established step-based and episode-based exploration strategies. Our results show that a more balanced trade-off can yield faster learning and better final policies, and illustrate some of the effects that cause these performance differences.     

Characterisation of trunk muscle activation amplitude patterns during a simulated checkstand operation with continuously changing flexor and lateral moment demands

While the typical physical exposure to modern-day workers has changed from heavy to low level repetitive demands, there is limited research that examines light occupations. This study examined trunk muscle recruitment strategies in response to a simulated checkout operation. Surface electromyography and kinematic variables were recorded from 29 healthy subjects. Four principal patterns accounted for 95.3% of the variation. Significant differences in scores captured different strategies in response to reach conditions and external moment directions. Synergistic co-activation of ipsilateral back sites and differential activation among external oblique and erector spinae sites suggests that the central nervous system may control different regions of the trunk musculature to optimally account for asymmetrical demands. The strategy between the internal oblique and back extensor sites suggests that a specific co-activation strategy may be needed during lighter work. During low-load occupational tasks, several recruitment strategies were required to maintain spinal stability and account for changing external moments.

Statement of Relevance: Different recruitment strategies found in response to changing external moments offer new insights into neuromuscular control for lighter work. Specifically, multiple trunk muscle sites interact in a complex manner, taking into account task specificity and individual variation that are valuable in workstation design, evaluating injury risk and estimating spinal loads.     

Deployment strategies for crowdsourcing text creation

Automatically generating text of high quality in tasks such as translation, summarization, and narrative writing is difficult as these tasks require creativity, which only humans currently exhibit. However, crowdsourcing such tasks is still a challenge as they are tedious for humans and can require expert knowledge. We thus explore deployment strategies for crowdsourcing text creation tasks to improve the effectiveness of the crowdsourcing process. We consider effectiveness through the quality of the output text, the cost of deploying the task, and the latency in obtaining the output. We formalize a deployment strategy in crowdsourcing along three dimensions: work structure, workforce organization, and work style. Work structure can either be simultaneous or sequential, workforce organization independent or collaborative, and work style either by humans only or by using a combination of machine and human intelligence. We implement these strategies for translation, summarization, and narrative writing tasks by designing a semi-automatic tool that uses the Amazon Mechanical Turk API and experiment with them in different input settings such as text length, number of sources, and topic popularity. We report our findings regarding the effectiveness of each strategy and provide recommendations to guide requesters in selecting the best strategy when deploying text creation tasks.     

Diversity of tasks and information technologies used by office workers at and away from work

Background. Computer use is associated with musculoskeletal complaints among office workers. Insufficient exposure diversity between tasks is a proposed etiological factor, but little information exists on diversity of tasks and information and communication technologies (ICT) among office workers. Method. Direct observation and self-report data were collected on tasks performed and ICT used among 24 office workers, over 12 h in work and non-work environments. Self-reports were repeated on four additional days. Results. Observations were for a mean [SD] 642[40] min. Productive tasks comprised 63% of observations, instrumental 17%, self-care 12% and leisure 8%. Non-ICT tasks comprised 44% of observations; New electronic-based ICT 36%; Old paper-based ICT 15%, and Combined ICT tasks 4%. Proportions of tasks and ICT use differed between environments and days. Conclusion. Information about diversity in tasks and ICT provides the basis for future investigations into exposure variation in ICT-intensive environments and possible musculoskeletal health risks.

Statement of relevance: Information and communication technologies (ICT) provide office workers access to perform work-related tasks after work hours and in away-from-work locations. Musculoskeletal disorder risk assessment for office workers should account for actual tasks performed over a work day, including away from work exposures. This study provides rich, detailed data on occurrence of tasks performed and ICT used by office workers throughout the day.     

Selection of strategies in judgment-based effort estimation

We currently know little about the factors that motivate the selection and change of strategy in judgment-based effort estimation. A better understanding of these issues may lead to more accurate judgment-based effort estimates and motivates the four experiments reported in this paper. The experiments’ two main results are the identification of the importance of “estimation surprises” (large estimation errors) to motivate estimation strategy change and the large individual variation in the initial choice of estimation strategy. The individual variation seems not only to be a result of differences in previous experiences, but also a result of differences in the mental “accessibility” of the strategies. We found, for example, that the use of a type of strategy was increased when we instructed a developer to use the same type of strategy on unrelated tasks immediately before.     

A formative analysis of how preservice teachers learn to use technology

Abstract   A comprehensive, formal comparison of strategies used by preservice teachers to learn how to use new technology has yet to be researched. Understanding the relative strengths and weakness of learning strategies would provide useful guidance to educators and students. The purpose of the current study was to explore the effectiveness of four learning strategies: collaboration, using authentic tasks, formal instruction and exploratory learning. Seventy-four preservice teachers (25 male, 49 female) were surveyed at the beginning and end of an 8-month, consecutive, Bachelor of Education programme, with respect to their learning strategies, change in computer knowledge and use of computers in the classroom. Collaborative learning and use of authentic tasks were the most preferred strategies – formal instruction was the least preferred. A collaborative approach to learning was the best predictor of gains in computer knowledge. Authentic tasks and collaborative strategies were significant predictors of teacher use of computers in the classroom. Preference for authentic tasks was the only predictor of student use of computers. Regardless of strategy preference, selecting more than one primary learning tool was significantly correlated with amount learned and use of the computers in the classroom. Ability was not related to strategy preference. Finally, females preferred collaborative approach to learning, although they were significantly more open to using multiple strategies than males.     

自适应迁移预测的动态多目标差分演化算法

针对动态多目标优化环境下寻找并跟踪变化的Pareto最优前沿和Pareto最优解集的难题,提出两个策略：自适应迁移策略和预测策略。自适应迁移策略是根据环境的变化自适应地插入迁移个体来提高算法种群的多样性,从而提高算法对动态环境的适应能力。预测策略是通过时间序列并加上一定的扰动来产生预测种群,来预测环境变化之后的Pareto最优解集,以达到对其快速跟踪的目的。通过两个策略在多目标差分演化算法上的应用来解决动态多目标优化问题。实验过程中,通过平均最优解集分布均匀度和平均决策空间世代距离等指标表明,基于自适应迁移策略和预测策略的多目标差分演化算法能够很好适应变化的环境,并能够快速找到Pareto最优解集。     

Investigation of task performance variations according to task requirements and alertness across the 24-h day in shift workers

The study investigated time-of-day effects on task performance in shift workers in different tasks (reaction time, discrimination, probe recognition, free recall), by varying task-specific features. On each of six recordings, each programmed on a different day and in a randomised order, operators rated alertness and performed different tasks. Self-rated alertness varied according to a typical diurnal trend. Time of day also affected reaction time (slower responses at 03:00 hours), discrimination performance (lower accuracy at 03:00 hours in the most difficult condition) and recall (superior recall at 07:00 and 11:00 hours following deeper processing at encoding). The data demonstrated time-of-day effects on cognitive processes also involved in many real-job activities, despite the lack of control for a number of exogenous factors known to interfere with performance in work settings. Since in the cognitively more loaded tasks, time-of-day effects depended on task conditions, the findings are of operational concern in shift-work situations involving differential task requirements. In a real-job setting, performance variations were observed according to time of day and task requirements in a set of cognitively more or less demanding tasks. Task-specific research across the 24-h day enables a better understanding of operators' tasks and the development of supporting technology.