AI is reshaping academic research, yet its role in peer review remains polarising and contentious. Advocates see its potential to reduce reviewer burden and improve quality, while critics warn of risks to fairness, accountability, and trust. At ICLR 2025, an official AI feedback tool was deployed to provide reviewers with post-review suggestions. We studied this deployment through surveys and interviews, investigating how reviewers engaged with the tool and perceived its usability and impact. Our findings surface both opportunities and tensions when AI augments in peer review. This work contributes the first empirical evidence of such an AI tool in a live review process, documenting how reviewers respond to AI-generated feedback in a high-stakes review context. We further offer design implications for AI-assisted reviewing that aim to enhance quality while safeguarding human expertise, agency, and responsibility.

Digital climate actions are an important aspect in centring voices in climate activism and therefore, it is important to study how digital channels support or hinder the actions of climate activists. Specifically, due to identity-related factors, people of colour (POC) could feel marginalised and othered, making it difficult to fit in in climate communities but online anonymity might help relieve these concerns. However, little research informs us what digital platforms POC actually use for building and engaging in climate communities including social media platforms (e.g., Facebook and Instagram), instant messaging channels (e.g., WhatsApp and Telegram), microblogging (e.g., Twitter), online newsletters and petitions and how these platforms and their anonymity features support or do not support their participation. Therefore, the current mixed-methods survey study was conducted to investigate what POC's choices of digital platforms are and how anonymity features support their participation in digital climate actions including their sense of safety and belonging. The results from 264 participants of five racial backgrounds i.e., White, Black, South Asian, Central, East and South East (CESE) Asians and mixed race, showed that there were different patterns in digital platform usage where most participants of colour had more distributed platform choices than White participants who were more likely to engage on identified sites such as Facebook and petitions. We also found that POC are more likely to engage on encrypted communication channels such as Telegram and WhatsApp and some anonymous actions than White participants. Results understanding the effects of anonymity reinforced that anonymity is important for POC's engagement because of safety reasons, however, being able to identify other POC is also important to their engagement as they might feel safer, giving them more courage to engage unanonymously. POC's need for anonymity, therefore, presented a contradictory but interesting paradox. Our study contributes to highlighting the importance of pluralistic and inclusive frameworks including trauma-informed computing and intersectionality in technological design which emphasises aspects of individual safety, community and digital platform trust and community support, collaboration and intersectional needs.

Spreadsheet collaboration provides valuable opportunities for learning and expertise sharing between colleagues. Sharing expertise is essential for the retention of important technical skillsets within organisations, but previous studies suggest that spreadsheet experts often fail to disseminate their knowledge to others. We suggest that social norms and beliefs surrounding the value of spreadsheet use significantly influence user engagement in sharing behaviours. To explore this, we conducted 31 semi-structured interviews with professional spreadsheet users from two separate samples. We found that spreadsheet providers face challenges in adapting highly personalised strategies to often subjective standards and in evaluating the appropriate social timing of sharing. In addition, conflicted self-evaluations of one's spreadsheet expertise, dismissive normative beliefs about the value of this knowledge, and concerns about the potential disruptions associated with collaboration can further deter sharing. We suggest that these observations reflect the challenges of long-term learning in feature-rich software designed primarily with initial learnability in mind. We therefore provide implications for design to navigate this tension. Overall, our findings demonstrate how the complex interaction between technology design and social dynamics can shape collaborative learning behaviours in the context of feature-rich software.

Voice Agents (VAs) are touted as being able to help users in complex tasks such as cooking and interacting as a conversational partner to provide information and advice while the task is ongoing. Through conversation analysis of 7 cooking sessions with a commercial VA, we identify challenges caused by a lack of contextual awareness leading to irrelevant responses, misinterpretation of requests, and information overload. Informed by this, we evaluated 16 cooking sessions with a wizard-led context-aware VA. We observed more fluent interaction between humans and agents, including more complex requests, explicit grounding within utterances, and complex social responses. We discuss reasons for this, the potential for personalisation, and the division of labour in VA communication and proactivity. Then, we discuss the recent advances in generative models and the VAs interaction challenges. We propose limited context awareness in VAs as a step toward explainable, explorable conversational interfaces.

Informal Knowledge Sharing (KS) is vital for enduser programmers to gain expertise. To better understand how personal (self-efficacy), social (reputational gains, trust between colleagues), and software-related (codification effort) variables influence spreadsheet KS intention, we conducted a multiple regressions analysis based on survey data from spreadsheet users ($n=100$) in administrative and finance roles. We found that high levels of spreadsheet self-efficacy and a perception that sharing would result in reputational gains predicted higher KS intention, but individuals who found knowledge codification effortful showed lower KS intention. We also observed that regardless of occupation, users tended to report a lower sense of self-efficacy in their general spreadsheet proficiency, despite also reporting high self-efficacy in spreadsheet use for job-related contexts. Our findings suggest that acknowledging and designing for these social and personal variables can help avoid situations where experienced individuals refrain unnecessarily from sharing, with implications for spreadsheet design.

As visualisations reach a broad range of audiences, designing visualisations that attract and engage becomes more critical. Prior work suggests that semantic icons entice and immerse the reader; however, little is known about their impact with informational tasks and when the viewer’s attention is divided because of a distracting element. To address this gap, we first explored a variety of semantic icons with various visualisation attributes. The findings of this exploration shaped the design of our primary comparative online user studies, where participants saw a target visualisation with a distracting visualisation on a web page and were asked to extract insights. Their engagement was measured through three dependent variables: (1) visual attention, (2) effort to write insights, and (3) self-reported engagement. In Study 1, we discovered that visualisations with semantic icons were consistently perceived to be more engaging than the plain version. However, we found no differences in visual attention and effort between the two versions. Thus, we ran Study 2 using visualisations with more salient semantic icons to achieve maximum contrast. The results were consistent with our first Study. Furthermore, we found that semantic icons elevated engagement with visualisations depicting less interesting and engaging topics from the participant’s perspective. We extended prior work by demonstrating the semantic value after performing an informational task (extracting insights) and reflecting on the visualisation, besides its value to the first impression. Our findings may be helpful to visualisation designers and storytellers keen on designing engaging visualisations with limited resources. We also contribute reflections on engagement measurements with visualisations and provide future directions.

Participation in climate activism is often facilitated by joining groups on digital platforms, for online and in-person participation in the movement. However, despite the easy access to communities digitally, people of colour (POC) are underrepresented in the UK in climate activism. Little research provides insight into why the voices of POC are underrepresented and how to level the playing field for participation in climate activism, which is just as important to POC. To investigate this, we conducted a qualitative semi-structured interview study with POC who are active in the climate movement (N=12). We identified five main themes: (i) gaining membership to the climate activism community, (ii) experiencing challenges in belonging with respect to identity characteristics, (iii) participating in protest through relevant issues and values, (iv) feeling fear due to differences and marginalisation, and (v) developing strategies for control and utilising anonymity. Our findings show how people's racial and other minoritised identity is crucial in influencing their sense of belonging and community and in turn their participation in climate activism and perhaps other opinion-based movements. For movements to be more inclusive, we urge the HCI and CSCW community to rethink designs influenced by structures of power and centre diverse values and issues instead of universal ones to amplify the voices of minoritised identities. Suggestions on how technologies could achieve this are discussed.

Research suggests that the temporal flexibility advertised to crowdworkers by crowdsourcing platforms is limited by both client-imposed constraints (e.g., strict completion times) and crowdworkers’ tooling practices (e.g., multitasking). In this article, we explore an additional contributor to workers’ limited temporal flexibility: the design of crowdsourcing platforms, namely requiring crowdworkers to be ‘on call’ for work. We conducted two studies to investigate the impact of having to be ‘on call’ on workers’ schedule control and job control. We find that being ‘on call’ impacted (1) participants’ ability to schedule their time and stick to planned work hours, and (2) the pace at which participants worked and took breaks. The results of the two studies suggest that the ‘on-demand’ nature of crowdsourcing platforms can limit workers’ temporal flexibility by reducing schedule control and job control. We conclude the article by discussing the implications of the results for (a) crowdworkers, (b) crowdsourcing platforms, and (c) the wider platform economy.

Embellishments are features commonly used in everyday visualisations which are demonstrated to enhance assimilation and memorability. Despite their popularity, little is known about their impact on enticing readers to explore visualisations. To address this gap, we conducted 18 interviews with a diverse group of participants who were consumers of news media but non-experts in visualisation and design. Participants were shown ten embellished and plain visualisations collected from the news and asked to rank them based on enticement and ease of understanding. Extending prior work, our interview results suggest that visualisations with multiple embellishment types might make a visualisation perceived as more enticing. An important finding from our study is that the widespread of certain embellishments in the media might have made them part of visualisation conventions, making a visualisation appear more objective but less enticing. Based on these findings, we ran a follow-up online user study showing participants variations of the visualisations with multiple embellishments to isolate each embellishment type and investigate its effect. We found that variations with salient embellishments were perceived as more enticing. We argue that to unpack the concept of embellishments; we must consider two factors: embellishment saliency and editorial styles. Our study contributes concept and design considerations to the literature concerned with visualisation design for non-experts in visualisation and design.

Message deletion in mobile messaging apps allows people to “unsay” things they have said. This paper explores how and why people use (or do not use) this feature within remediation strategies after a communication error is identified. We present findings from a multi-stage survey designed to explore people’s general experiences of the message deletion feature (N = 401), peoples’ experiences of using this feature during the remediation of an error (N = 70), and receivers’ perceptions around recent message deletions (N = 68). While people are typically aware of the deletion feature, it is infrequently used. When used, it is primarily done so to improve conversations by reducing confusion between conversation partners. We found people being aware of message deletions creating information-gaps which can provoke curiosity in recipients, causing them to develop narratives to help address the uncertainty. We found concerns amongst senders that these narratives would be of a negative nature, having an undesirable impact on how others perceive them. We use our findings to suggest ways in which mobile messaging apps could improve conversational experiences around erroneous and regrettable messages.

Although e-mail interfaces have not changed much in the past decade, how we manage our inbox has changed a lot, as a result of the growing number of messages we receive (The Radicati Group, 2015) a...

Many human activities, such as manufacturing and assembly, are sequence-constrained procedural tasks (SPTs): they consist of a series of steps that must be executed in a specific spatial/temporal order. However, these tasks can be error prone - steps can be missed out, executed out-of-order, and repeated. The ability to automatically predict if a person is about to commit an error could greatly help in these cases. The prediction could be used, for example, to provide feedback to prevent mistakes or mitigate their effects. In this paper, we present a novel approach for real-time error prediction for multi-step sequence tasks which uses a minimum viable set of behavioural signals. We have three main contributions. The first we present an architecture for real-time error prediction based on task tracking and intent prediction. The second is to explore the effectiveness of using hand position and eye-gaze tracking for task tracking. We confirm that eye-gaze is more effective for intent prediction, hand tracking is more accurate for task tracking and that combining the two provides the best overall response. We show that using Hands and Gaze tracking data we can predict selection/placement errors with an F1 score of 97%, approximately 300ms before the error would occur. Finally, we discuss the application of this hand-gaze error detection architecture used in conjunction with head-mounted AR displays, to support industrial manual assembly.

Reliable and accurate planning is essential for modern knowledge workers. However, there is limited insight about when, how and why planning is inaccurate, and the circumstances in which those inaccuracies are troublesome. To investigate this, we asked 20 academics to keep a diary for a single work day. They estimated the duration of the tasks they wanted to achieve at the start of the day and noted down in detail the tasks they actually achieved during the day. Semi-structured interviews were conducted to complement this diary data. The diaries showed that some tasks, such as email and coding, were more susceptible to time underestimation bias while other tasks, such as writing and planning, were more susceptible to time overestimation bias in planning. Based on interviews, a typology of common reasons for delays in planned daily work is presented. It suggests that vague and optimistic planning leads to the observed discrepancy between planned and actual work. Finally, interviews suggested that participants adopted four planning strategies that vary in the frequency of planning, from minimal planning to daily, weekly and multi-level planning. We close by discussing ways support systems for accurate planning can be better designed for different use cases.

Many computer tasks involve looking up information from different sources, and these self-interruptions can be disruptive. In this article, we investigate whether giving people feedback on how long they are away from their task influences their self-interruption behaviour. We conducted a contextual inquiry on self-interruption behaviour in an office workplace. Participants were observed to postpone physical interruptions until a convenient moment in the task if they were expected to take time. In contrast, observations revealed that digital interruptions were addressed immediately; participants reported these were presumed to be quick to deal with. To increase awareness of time spent on digital interruptions, we developed TimeToFocus, a notification tool showing people the duration of their interruptions while working on a task. A field study deployment of TimeToFocus in an office workplace found that feedback on the duration of interruptions made participants reflect on what they were doing during interruptions. They reported that they used this insight to avoid task-irrelevant activities. To confirm whether participants’ perceptions of the benefit of the tool could be measured, we conducted an online experiment, where participants had to retrieve information from an email sent to their personal email addresses and enter it into a spreadsheet. Participants who used our tool made shorter interruptions, completed the spreadsheet task faster and made fewer data entry errors. We conclude that feedback on the length of interruptions can assist users in focusing on their primary task and thus improve productivity.

Automated driving will mean that people can engage in other activities and an important concern will be how to alert the driver to critical events that require their intervention. This study evaluates how various levels of assertiveness of voice command in a semi-AV and different degrees of immersion of a non-driving task may affect people's attention on the road. In a simulated set-up, 20 participants were required to execute actions on the steering wheel when a voice command was given while playing a mobile game. Regardless of how immersed the driver was in the game, a more assertive voice resulted in faster reaction time to the instructions and was perceived as more urgent than a less assertive voice. Automotive systems should use an assertive voice to effectively grab people's attention. This is effective even when they are engaged in an immersive secondary task.

Crowdworkers receive no formal training for managing their tasks, time or working environment. To develop tools that support such workers, an understanding of their preferences and the constraints they are under is essential. We asked 317 experienced Amazon Mechanical Turk workers about factors that influence their task and time management. We found that a large number of the crowdworkers score highly on a measure of polychronicity; this means that they prefer to frequently switch tasks and happily accommodate regular work and non-work interruptions. While a preference for polychronicity might equip people well to deal with the structural demands of crowdworking platforms, we also know that multitasking negatively affects workers' productivity. This puts crowdworkers' working preferences into conflict with the desire of requesters to maximize workers' productivity. Combining the findings of prior research with the new knowledge obtained from our participants, we enumerate practical design options that could enable workers, requesters and platform developers to make adjustments that would improve crowdworkers' experiences.

We review the history of human-automation interaction research, assess its current status and identify future directions. We start by reviewing articles that were published on this topic in the International Journal of Human-Computer Studies during the last 50 years. We find that over the years, automated systems have been used more frequently (1) in time-sensitive or safety-critical settings, (2) in embodied and situated systems, and (3) by non-professional users. Looking to the future, there is a need for human-automation interaction research to focus on (1) issues of function and task allocation between humans and machines, (2) issues of trust, incorrect use, and confusion, (3) the balance between focus, divided attention and attention management, (4) the need for interdisciplinary approaches to cover breadth and depth, (5) regulation and explainability, (6) ethical and social dilemmas, (7) allowing a human and humane experience, and (8) radically different human-automation interaction.

Mobile devices are increasingly used while watching video, both as a secondary device and for dedicated viewing. However, devices frequently issue notifications that can interrupt viewing. This study investigated the effect of interruptions from notifications on viewer immersion. Participants watched 10 minutes of a movie without notifications, and 10 minutes while receiving message notifications. There were two participant groups: one watched video on a 30-inch monitor with messages sent to a separate smartphone; while another watched on a smartphone while also receiving messages on it. Viewer immersion was assessed after each condition via questionnaire. We also considered message response times. Results showed that immersion scores were lower when the video was interrupted with notifications, regardless of viewing device. Message response times were fastest when using the phone as a secondary device. Our results suggest that device-driven interruptions should be minimised for an immersive film experience.

Data-entry is a common activity that is usually performed accurately. When errors do occur though, people are poor at spotting them even if they are told to check their input. We considered whether making people pause for a brief moment before confirming their input would make them more likely to check it. We ran a lab experiment to test this idea. We found that task lockouts encouraged checking. Longer lockout durations made checking more likely. We ran a second experiment on a crowdsourcing platform to find out whether lockouts would still be effective in a less controlled setting. We discovered that longer lockouts induced workers to switch to other activities. This made the lockouts less effective. To be useful in practice, the duration of lockouts needs to be carefully calibrated. If lockouts are too brief they will not encourage checking. If they are too long they will induce switching.

Link-shortening services save space and make the manual entry of URLs less onerous. Short links are often included on printed materials so that people using mobile devices can quickly enter URLs. Although mobile transcription is a common use-case, link-shortening services generate output that is poorly suited to entry on mobile devices: links often contain numbers and capital letters that require time consuming mode switches on touch screen keyboards. With the aid of computational modeling, we identified problems with the output of a link-shortening service, bit.ly. Based on the results of this modeling, we hypothesized that longer links that are optimized for input on mobile keyboards would improve link entry speeds compared to shorter links that required keyboard mode switches. We conducted a human performance study that confirmed this hypothesis. Finally, we applied our method to a selection of different non-word mobile data-entry tasks. This work illustrates the need for service design to fit the constraints of the devices people use to consume services.

Obtaining high-quality data from crowds can be difficult if contributors do not give tasks sufficient attention. Attention checks are often used to mitigate this problem, but, because the roots of inattention are poorly understood, checks often compel attentive contributors to complete unnecessary work. We investigated a potential source of inattentiveness during crowdwork: multitasking. We found that workers switched to other tasks every 5 minutes, on average. There were indications that increasing switch frequency negatively affected performance. To address this, we tested an intervention that encouraged workers to stay focused on our task after multitasking was detected. We found that our intervention reduced the frequency of task switching. It also improves on existing attention checks because it does not place additional demands on workers who are already focused. Our approach shows that crowds can help to overcome some of the limitations of laboratory studies by affording access to naturalistic multitasking behavior.

Menu interfaces often arrange options into semantic groups. This semantic structure is then usually conveyed to the user by supplementary visual grouping cues. We investigate whether these visual grouping cues actually help users locate items in menus faster, and whether there is potential for these powerful grouping cues to impede search when used inappropriately. Thirty-six participants performed known-item searches of word menus. These menus differed along three dimensions: (1) whether visual grouping cues were used, (2) whether items were semantically organized, and (3) the number of items belonging to each semantic group. Results show that the usefulness of visual grouping entirely depends on the underlying semantic structure of the menu. When menus were semantically organized, having visual grouping cues delineate the boundaries between large semantic groups resulted in the fastest search times. But when semantically unrelated items were visually grouped together, participants took far longer to locate targets. Menu designers should therefore take great care to avoid visually grouping semantically unrelated items as this has the potential to hinder menu interactions.

One reason that human interaction with technology is difficult to understand is because the way in which people perform interactive tasks is highly adaptive. One such interactive task is menu search. In the current article we test the hypothesis that menu search is rationally adapted to (1) the ecological structure of interaction, (2) cognitive and perceptual limits, and (3) the goal to maximise the trade-off between speed and accuracy. Unlike in previous models, no assumptions are made about the strategies available to or adopted by users, rather the menu search problem is specified as a reinforcement learning problem and behaviour emerges by finding the optimal Markov Decision Process (MDP). The model is tested against existing empirical findings concerning the effect of menu organisation and menu length. The model predicts the effect of these variables on task completion time and eye movements. The discussion considers the pros and cons of the modelling approach relative to other well-known mod- elling approaches.

It is well established that driving while interacting with a secondary in-car device, such as a portable music player, is distracting and can lead to a decline in driver safety and performance. One aspect that has not received as much attention though is the extent to which drivers adapt their interactions with an in-car device to the changing demands of the road. Do drivers adopt compensatory strategies or tactics when driving in more demanding settings? We investigate this question by conducting a driving simulator study in which participants were required to drive either in a city or a highway environment. During these drives, participants were asked to interact with an MP3 music player and make a series of either easy or difficult selections. It was found that participants who drove in the city made shorter glances to the in-car iPod than those that drove on the highway. As a result of this tactical change, participants had better lane keeping performance in the city, which was important given the narrower traffic lanes. As expected, we also replicate the well-known effect that more complex secondary in-car tasks are more distracting than simpler in-car tasks. The contribution this paper makes to the automotive UI community is that it provides evidence that drivers adapt to the demands of the driving environment even when interacting with secondary devices.

This paper presents a novel mathematical model for visual search and selection time in linear menus. Assuming two visual search strategies, serial and directed, and a pointing sub-task, it captures the change of performance with five fac- tors: 1) menu length, 2) menu organization, 3) target position, 4) absence/presence of target, and 5) practice. The novel aspect is that the model is expressed as probability density distribution of gaze, which allows for deriving total selection time. We present novel data that replicates and extends the Nielsen menu selection paradigm and uses eye-tracking and mouse tracking to confirm model predictions. The same parametrization yielded a high fit to both menu selection time and gaze distributions. The model has the potential to improve menu designs by helping designers identify more effective solutions without conducting empirical studies.

Mobile phones have evolved significantly in recent years from single-purpose communication devices to multi-purpose computing devices. Despite this evolution, the interaction model for how incoming calls are handled has barely changed. Current-generation smartphones still use abrupt full-screen notifications to alert users to incoming calls, demanding a decision to either accept or decline the call. These full-screen notifications forcibly interrupt whatever activity the user was already engaged in. This might be undesirable when the user's primary task was more important than the incoming call. This paper explores the design space for how smartphones can alert users to incoming calls. We consider designs that allow users to postpone calls and also to multiplex by way of a smaller partial-screen notification. These design alternatives were evaluated in both a small-scale controlled lab study as well as a large-scale naturalistic in-the-wild study. Results show that a multiplex design solution works best because it allows people to continue working on their primary task while being made aware that there is a caller on the line. The contribution of this work is an enhanced interaction design for handling phone calls, and an understanding of how people use it for handling incoming calls.

When searching a display, users sometimes know what the target is but sometimes do not. It has generally been assumed that for this latter case people must engage in a deeper semantic evaluation of items during the search process. This idea is central to Information Foraging theory. But do people actually spend longer assessing items when engaged in a semantically demanding search task' We investigate this by having participants locate target items in 16-item menus. Participants were either told exactly what to look for (known-item search) or they were told the category that the target belonged to (semantic search). Participants were faster and more accurate at known-item searches. Eye-movement data show that this was because participants were more likely to skip over items when performing known-item searches. Contrary to expectation, we found limited empirical evidence to support the idea that deeper semantic evaluations of items lead to longer gaze durations (this occurred only when items were arranged very close together). This finding is important because it reveals how people adopt different eye gaze strategies depending on the kind of search activity they are engaged in.

Number entry is a common task in many domains. In safety-critical environments such as air traffic control or on hospital wards, incorrect number entry can have serious harmful consequences. Research has investigated how interface designs can help prevent users from making number entry errors. In this paper, we present an experimental evaluation of two possible interface designs aimed at helping users detect number entry errors using the idea of a checksum: an additional (redundant) number that is related to the to-be-entered numbers in such a way that it is sufficient to verify the correctness of the checksum, as opposed to checking each of the entered numbers. The first interface requires users to check their own work with the help of the checksum; the second requires the user to enter the checksum along with the other numbers so that the system can do the checking. In each case, two numbers needed to be entered, while the third number served as a checksum. With the first interface, users caught only 36% of their errors. The second interface resulted in all errors being caught, but the need to enter the checksum increased entry time by 46%. When participants were allowed to choose between the two interfaces, they chose the second interface in only 12% of the cases. Although these results cannot be generalized to other specific contexts, the results illustrate the strengths and weaknesses of each way of using checksums to catch number entry errors. Hence our study can serve as a starting point for efforts to improve each method.

Journal Article Shared Input Multimodal Mobile Interfaces: Interaction Modality Effects on Menu Selection in Single-Task and Dual-Task Environments Get access Shengdong Zhao, Shengdong Zhao * 1Department of Computer Science, National University of Singapore, 13 Computing Drive, Computing 2, #01-04, Singapore 117417 *Corresponding author: [email protected] Search for other works by this author on: Oxford Academic Google Scholar Duncan P. Brumby, Duncan P. Brumby 2UCL Interaction Centre, University College London, Gower Street, London WC1E 6BT, UK Search for other works by this author on: Oxford Academic Google Scholar Mark Chignell, Mark Chignell 3Knowledge Media Design Institute (KMDI), University of Toronto, 27 King's College Circle, Toronto, Ont., Canada M5S 1A1 Search for other works by this author on: Oxford Academic Google Scholar Dario Salvucci, Dario Salvucci 4Drexel University, 3141 Chestnut Street, Philadelphia, PA 19104, USA Search for other works by this author on: Oxford Academic Google Scholar Sahil Goyal Sahil Goyal 5National University of Singapore, 13 Computing Drive, Computing 2, #01-04, Singapore 117417 Search for other works by this author on: Oxford Academic Google Scholar Interacting with Computers, Volume 25, Issue 5, September 2013, Pages 386–403, https://doi.org/10.1093/iwc/iws021 Published: 06 February 2013 Article history Received: 29 December 2011 Revision received: 11 October 2012 Accepted: 13 November 2012 Published: 06 February 2013

People are prone to making more errors when multitasking. Thus in safety-critical environments, it is often considered safer to perform tasks sequentially. Here we explore how the cost of accessing information affects the way people choose to interleave. An empirical study based on a medical scenario was conducted. Participants had to program infusion pump devices using information from a prescription form. The physical and mental effort involved in accessing information was manipulated. This was achieved by varying the physical distance between the prescription form and the devices. We demonstrate that by increasing information access cost, individuals are less likely to omit a required task step. This is because they adopt a more memory-intensive strategy, which encourages interleaving at natural boundaries, i.e., after completing the programming of one of the pumps. Interleaving during programming can result in task steps being forgotten.

Auto-locks are a necessary feature on many modern day mobile devices, but can they sometimes have detrimental consequences? In this paper we investigate how auto-locks can affect behavior in a demanding multitasking scenario. A study is conducted in which participants had to enter text using a touch-screen interface while driving a simulated vehicle in a lab setting. Different auto-lock mechanisms were implemented on the secondary device, manipulating both the duration of the lockout threshold (i.e., the period of inactivity before the auto-lock was initiated) and the complexity of the unlock procedure (i.e., how easy it was for the user to unlock the device once it had locked). Results showed that lane-keeping performance on the primary driving task was worse when there was a shorter lockout threshold. The reason for this was two-fold: (1) participants took fewer long pauses between typing actions, so as to avoid being locked out of the device, and (2) when the device did lock, unlocking it took time and further distracted the driver. In support of this latter finding, we also found that a more complex unlock procedure, which required a pin code to be entered, resulted in worse lane-keeping performance than when the device could be unlocked by making a simple button press. These findings suggest that auto-locks can dissuade users from regularly interleaving attention between other ongoing activities. Designers should keep this in mind when incorporating auto-locks in mobile devices.

In-car devices that use audio output have been shown to be less distracting than traditional graphical user interfaces, but can be cumbersome and slow to use. In this paper, we report an experiment that demonstrates how these performance characteristics impact whether people will elect to use an audio interface in a multitasking situation. While steering a simulated vehicle, participants had to locate a source of information in a short passage of text. The text was presented either on a visual interface, or using a text-to-speech audio interface. The relative importance of each task was varied. A no-choice/choice paradigm was used in which participants first gained experience with each of the two interfaces, before being given a choice on which interface to use on later trials. The characteristics of the interaction with the interfaces, as measured in the no-choice phase, and the relative importance of each task, had an impact on which output modality was chosen in the choice phase. Participants that prioritized the secondary task tended to select the (faster yet more distracting) visual interface over the audio interface, and as a result had poorer lane keeping performance. This work demonstrates how a user's task objective will influence modality choices with multimodal devices in multitask environments.

We investigate how people adapt their strategy for interleaving multiple concurrent tasks to varying objectives. A study was conducted in which participants drove a simulated vehicle and occasionally dialed a telephone number on a mobile phone. Experimental instructions and feedback encouraged participants to focus on either driving or dialing. Results show that participants adapted their task interleaving strategies to meet the required task objective, but in a manner that was nonetheless intricately shaped by internal psychological constraints. In particular, participants tended to steer in between dialing chunks of digits even when extreme vehicle drift implied that more reactive strategies would have generated better lane keeping. To better understand why drivers interleaved tasks at chunk boundaries, a modeling analysis was conducted to derive performance predictions for a range of dialing strategies. The analysis supported the idea that interleaving at chunk boundaries efficiently traded the time given up to dialing with the maintenance of a central lane position. We discuss the implications of this work in terms of contributions to understanding how cognitive constraints shape strategy adaptations in dynamic multitask environments.

When searching a novel Web page, people often estimate the likelihood that labeled links on the page will lead to their goal. A rational analysis of this activity suggests that people should adjust their estimate of the likelihood that any one item will lead to the goal in a manner that is sensitive to the context provided by the likelihoods that other items on the page will lead to the goal. Two experiments were designed to provide evidence to discriminate between this account and others found in the literature (e.g., satisficing and assess-all accounts). The experiments systematically manipulated the relevance of the distractor items and the location of the target item on the page. The results showed that (a) a high-value item was more likely to be selected when it was first encountered if the relevance of competing distractors was relatively low and (b) more items were assessed prior to selection when the distractors were of greater semantic relevance to the goal. The location manipulation showed that if more distractors were assessed prior to the target item, then the relevance of the distractors had a greater influence on the decision as to whether to select the target immediately. These results suggest that decisions as to when to select an item from the page are sensitive to the context provided by the likelihoods of all of the items so far assessed and not just to the most recent item. The findings are therefore inconsistent with both satisficing and assess-all accounts of interactive search.

The paper describes an approach to modeling the strategic variations in performing secondary tasks while driving. In contrast to previous efforts that are based on simulation of a cognitive architecture interacting with a task environment, we take an approach that develops a cognitive constraint model of the interaction between the driver and the task environment in order to make inferences about dual-task performance. Analyses of driving performance data reveal that a set of simple equations can be used to accurately model changes in the lateral position of the vehicle within the lane. The model quantifies how the vehicle's deviation from lane center increases during periods of inattention, and how the vehicle returns to lane center during periods of active steering. We demonstrate the benefits of the approach by modeling the dialing of a cellular phone while driving, where drivers balance the speed in performing the dial task with accuracy (or safety) in keeping the vehicle centered in the roadway. In particular, we show how understanding, rather than simulating, the constraints imposed by the task environment can help to explain the costs and benefits of a range of strategies for interleaving dialing and steering. We show how particular strategies are sensitive to a combination of internal constraints (including switch costs) and the trade-off between the amount of time allocated to secondary task and the risk of extreme lane deviation.

Portable music players such as Apple's iPod have become ubiquitous in many environments, but one environment in particular has elicited new safety concerns and challenges -- in-vehicle use while driving. We present the first study of portable music-player interaction while driving, examining the effects of iPod interaction by drivers navigating a typical roadway in a driving simulator. Results showed that selecting media on the iPod had a significant effect on driver performance as measured by lateral deviation from lane center; the effect was comparable to previously reported effects of dialing a cellular phone. In addition, selecting media and watching videos had a significant effect on car-following speed, resulting in speed reductions that presumably compensated for impaired lateral performance. Given that iPod interaction has become increasingly common while driving, these results serve as a first step toward understanding the potential effects of portable music-player interaction on driver behavior and performance.