Viewers tend to underestimate correlation in positively correlated scatterplots. However, systematically changing the size and opacity of scatterplot points can bias estimates upwards, correcting for this underestimation. Here, we examine whether the application of these visualisation techniques goes beyond a simple perceptual effect and could actually influence beliefs about information from trusted news sources. We present a fully-reproducible study in which we demonstrate that scatterplot manipulations that are able to correct for the correlation underestimation bias can also induce stronger levels of belief change compared to conventional scatterplots presenting identical data. Consequently, we show that novel visualisation techniques can be used to drive belief change, and suggest future directions for extending this work with regards to altering attitudes and behaviours.

When visualising data, chart designers have the freedom to choose the upper and lower limits of numerical axes. Axis limits can determine the physical characteristics of plotted values, such as the physical position of data points in dot plots. In two experiments (total N=300), we demonstrate that axis limits affect viewers' interpretations of the magnitudes of plotted values. Participants did not simply associate values presented at higher vertical positions with greater magnitudes. Instead, participants considered the relative positions of data points within the axis limits. Data points were considered to represent larger values when they were closer to the end of the axis associated with greater values, even when they were presented at the bottom of a chart. This provides further evidence of framing effects in the display of data, and offers insight into the cognitive mechanisms involved in assessing magnitude in data visualisations.

Systematically changing the size and opacity of points on scatterplots can be used to induce more accurate perceptions of correlation by viewers. Evidence points to the mechanisms behind these effects being similar, so one may expect their combination to be additive regarding their effects on correlation estimation. We present a fully-reproducible study in which we combine techniques for influencing correlation perception to show that in reality, effects of changing point size and opacity interact in a non-additive fashion. We show that there is a great deal of scope for using visual features to change viewers’ perceptions of data visualizations. Additionally, we use our results to further interrogate the perceptual mechanisms at play when changing point size and opacity in scatterplots.

Data visualisation software provides the ability to create highly customisable choropleth maps.This presents an abundance of design choices.The colour legend, one particular aspect of choropleth map design, has the potential to effectively convey data points' absolute magnitudes (how large or small they are).Colour legends present the mapping between a specific range of colours and a specific range of numerical values.In this experiment, we demonstrate that manipulating this range affects interpretations of the plotted values' absolute magnitudes.Participants (N = 100) judged the urgency of addressing pollution levels as greater when the colour legend's upper bound was equal to the maximum plotted value, compared to when it was significantly larger than the maximum plotted value.This provides insight into the cognitive processing of plotted data in choropleth maps that are designed to promote inferences about overall magnitude.

Websites are frequently used by programmers to support the development process. This paper investigates programmer-Web interactions when coding, and combines observations of behaviour with assessments of the resulting source code. We report on an online observational study with ten undergraduate student programmers as they engaged in programming tasks of varying complexity. Screens were recorded of participants’ activities, and each participated in an interview. Videos and interviews were thematically analysed. Novice programmers employed various strategies for seeking and utilising online knowledge. The resulting source code was examined to determine the extent to which it met requirements and whether it contained errors. The source code analysis revealed that coding with the websites involved more coding time and effort, but increased the possibility of producing correct code. However, coding with websites also introduced instances of either incorrect or non-executable source code.

Resilience engineering (RE) is most commonly applied at the organisational level, and has historically been associated with safety-critical industries such as nuclear, medical or aviation. This paper explores the application of RE frameworks within software engineering, and investigates resilient performance of the socio-technical system that supports the creation of software. We present a preliminary study based on a secondary analysis of data from previous ethnographic studies of commercial software practice. This analysis uses an RE framework devised for small team practice in safety critical settings. We present and discuss three salient episodes of software practice that illustrate the application of RE principles to software engineering, and suggest how this kind of analysis may benefit software engineering. We present challenges and opportunities based on our experience and propose future research directions.

Scatterplots are common data visualizations that can be used to communicate a range of ideas, the most intensively studied being the correlation between two variables. Despite their ubiquity, people typically do not perceive correlations between variables accurately from scatterplots, tending to underestimate the strength of the relationship displayed. Here we describe a two-experiment study in which we adjust the visual contrast of scatterplot points, and demonstrate a systematic approach to altering the bias. We find evidence that lowering the total visual contrast in a plot leads to increased bias in correlation estimates and show that decreasing the salience of points as a function of their distance from the regression line, by lowering their contrast, can facilitate more accurate correlation perception. We discuss the implications of these findings for visualization design, and provide a framework for online, reproducible, and large-sample-size (N = 150 per experiment) testing of the design parameters of data visualizations.

Websites are frequently used to support the development process. This paper investigates how websites are used when writing code and programmers’ perceptions of the potential impact of this on their behaviour and the quality of the resulting software. We interviewed 18 programmers (13 students enrolled in undergraduate computer science courses, and 5 experienced professionals), and analysed the data thematically. The findings were used to develop a survey, which was distributed to 276 programmers (251 students, 25 experienced professionals). The results indicate that use of websites, especially Stack Overflow, is viewed as an essential part of programming by both students completing coursework and professionals developing code in industry. We also found that developers have experience of encountering a diverse set of problematic code snippets online, that copying code from websites without checking its quality or understanding how it worked is common, and that using online resources in this way had a potentially counter-productive effect on learning. Based on these findings, we make a number of recommendations, including better consideration of online code reuse in taught programmes, co-development and code-reuse practices in professional settings, and software licensing training for professional developers. Editor’s note: Open Science material was validated by the Journal of Systems and Software Open Science Board.

There is a growing interest in the use of robotics and AI in the nuclear industry, however it is important to ensure these systems are ethically grounded, trustworthy and safe. An emerging technique to address these concerns is the use of explainability. In this paper we present the results of an interview study with nuclear industry experts to explore the use of explainable intelligent systems within the field. We interviewed 16 participants with varying backgrounds of expertise, and presented two potential use cases for evaluation; a navigation scenario and a task scheduling scenario. Through an inductive thematic analysis we identified the aspects of a deployment that experts want to know from explainable systems and we outline how these associate with the folk conceptual theory of explanation, a framework in which people explain behaviours. We established that an intelligent system should explain its reasons for an action, its expectations of itself, changes in the environment that impact decision making, probabilities and the elements within them, safety implications and mitigation strategies, robot health and component failures during decision making in nuclear deployments. We determine that these factors could be explained with cause, reason, and enabling factor explanations.

Physical activity reconstruction is a process whereby self-trackers reflect on their physical activities and goals in an episodic fashion by recalling a series of past experiences and events. Tracking tools often include spatio-temporal cues (i.e. maps and timelines) to provide further context to these physical activity patterns, yet little is known about whether these features actually aid reflection. Here, we report a four-week user study (N=26) run with FitReconstruct, an app that fosters reconstruction of physical activity data using spatio-temporal and contextual cues to support recall of episodic information. We found that compared to a baseline version, presenting spatio-temporal and contextual cues increased the quality of reflection and user satisfaction. These findings inform design recommendations to support explicit reflective prompts in physical activity tracking apps.

Trust is a critical issue in human–robot interactions as it is at the base of the establishment of solid relationships. Theory of Mind (ToM) is the cognitive skill that allows us to understand what others think and believe. Several studies in HRI and psychology suggest that trust and ToM are interdependent concepts since we trust another agent based on our representation of its actions, beliefs, and intentions. However, very few works take ToM of the robot into consideration while studying trust in HRI. In this paper, we aim to examine whether the perception of ToM abilities on a robotic agent influences human-robot trust over time in an iterative game scenario. To this end, participants played an Investment Game with a humanoid robot (Pepper) that was presented as having either low-level ToM or high-level ToM. During the game, the participants were asked to pick a sum of money to invest in the robot. The amount invested was used as the main measurement of human-robot trust. Our experimental results show that robots possessing a high-level of ToM abilities were trusted more than the robots presented with low-level ToM skills.

Media is evolving from traditional linear narratives to personalised experiences, where control over information (or how it is presented) is given to individual audience members. Measuring and understanding audience engagement with this media is important in at least two ways: (1) a post-hoc understanding of how engaged audiences are with the content will help production teams learn from experience and improve future productions; (2), this type of media has potential for real-time measures of engagement to be used to enhance the user experience by adapting content on-the-fly. Engagement is typically measured by asking samples of users to self-report, which is time consuming and expensive. In some domains, however, interaction data have been used to infer engagement. Fortuitously, the nature of interactive media facilitates a much richer set of interaction data than traditional media; our research aims to understand if these data can be used to infer audience engagement. In this paper, we report a study using data captured from audience interactions with an interactive TV show to model and predict engagement. We find that temporal metrics, including overall time spent on the experience and the interval between events, are predictive of engagement. The results demonstrate that interaction data can be used to infer users' engagement during and after an experience, and the proposed techniques are relevant to better understand audience preference and responses.

Self-trackers reflect on their personal data to understand their behaviour and plan accordingly. Often, this reflection involves uncertainty, which can affect decision-making. To better understand the role of uncertainty, we conducted an interview study to comprehend how uncertainty influences reflection and the resulting actions. Our findings suggest that, in addition to the conventional role of uncertainty as a barrier, uncertainty also manifests as a trigger and facilitator to reflection. We discuss functionalities to alleviate the negative effects of uncertainty (e.g. incorporating users' expectations in self-tracking), and leverage its positive effects through activity reconstruction mechanisms.

The electrocardiogram (ECG), a graphical representation of the heart's electrical activity, is used for detecting cardiac pathologies. Certain medications can produce a complication known as 'long QT syndrome', shown on the ECG as an increased gap between two parts of the waveform. Self-monitoring for this could be lifesaving, as the syndrome can result in sudden death, but detecting it on the ECG is difficult. Here we evaluate whether using pseudo-colour to highlight wave length and changing the coordinate system can support lay people in identifying increases in the QT interval. The results show that introducing colour significantly improves accuracy, and that whilst it is easier to detect a difference without colour with Cartesian coordinates, the greatest accuracy is achieved when Polar coordinates are combined with colour. The results show that applying simple visualisation techniques has the potential to improve ECG interpretation accuracy, and support people in monitoring their own ECG.

The use of mobile devices during television viewing is now commonplace, and broadcasters are increasingly supplying programme-related ‘companion content’. To produce an optimal user experience, it is important to determine how the delivery of companion content affects and is perceived by the viewer — without this we risk distracting the viewer, leading to frustration and disengagement. We present a controlled study investigating how attention, cognitive load (as measured by the NASA TLX), and users’ preferences are affected by the provision of two different content delivery modes: pushed and pulled . We find that delivery mode affected the temporal distribution of gaze to the tablet, with a consistent viewing pattern for pushed updates, which attracted attention within a few seconds, and a more diverse set of viewing patterns when updates were pulled. Cognitive load was similar in both conditions, and there was no consensus as to which mode was preferred, but users showed strong, polarised individual preferences. The advantages of each delivery mode are presented as a set of recommendations for the delivery of companion content.

OBJECTIVE: The study sought to quantify a layperson's ability to detect drug-induced QT interval prolongation on an electrocardiogram (ECG) and determine whether the presentation of the trace affects such detection. MATERIALS AND METHODS: Thirty layperson participants took part in a psychophysical and eye-tracking experiment. Following training, participants completed 21 experimental trials, in which each trial consisted of 2 ECGs (a baseline and a comparison stimulus, both with a heart rate of 60 beats/min). The experiment used a 1 alternative forced-choice paradigm, in which participants indicated whether or not they perceived a difference in the QT interval length between the 2 ECGs. The ECG trace was presented in 3 ways: a single complex with the signals aligned by the R wave, a single complex without alignment, and a 10-second rhythm strip. Performance was analyzed using the psychometric function to estimate the just noticeable difference threshold, along with eye-tracking metrics. RESULTS: The just noticeable difference 50% and 75% thresholds were 30 and 88 ms, respectively, showing that the majority of laypeople were able to detect a clinically significant QT-prolongation at a low normal heart rate. Eye movement data indicated that people were more likely to appraise the rhythm strip stimulus systematically and accurately. CONCLUSIONS: People can quickly be trained to self-monitor, which may help with more rapid identification of drug-induced long QT syndrome and prevent the development of life-threatening complications. The rhythm strip is a better form of presentation than a single complex, as it is less likely to be misinterpreted due to artifacts in the signal.

The development of ontology engineering tools has traditionally lacked a user-centred perspective, instead being guided by the need to address particular gaps indicated by anecdotal evidence. This has typically resulted in prototypes that do not obtain traction beyond a narrow scope. Understanding the authoring patterns of ontology engineers is crucial to informing the development of ontology engineering tools that cater for the activity workflows of the users and, consequently, boosting the adoption of these tools. We report evidence about how Protégé is used across three different authoring settings, addressing the threats to validity of relying on a single user study. These settings address the continuum of expertise (from intermediate to expert users), the type of tasks (whether they are free-form or prescriptive) and the effect of the location (laboratory, tutorial or on their own) and how the studies are administered (whether or not there is a close supervision). While there are activity workflows that are particular to settings, the results indicate a number of core workflows that are common to all of them. We discuss actionable recommendations for ontology engineering tools in light of these results.

We report the process used to create artefacts for self-reporting Parkinson's Disease symptoms. Our premise was that a technology-based approach would provide participants with an effective, flexible, and resilient technique. After testing four prototypes using Bluetooth, NFC, and a microcontroller we accomplished almost full compliance and high acceptance using a paper diary to track day-to-day fluctuations over 49 days. This diary is tailored to each patient's condition, does not require any handwriting, allows for implicit reminders, provides recording flexibility, and its answers can be encoded automatically. We share five design implications for future Parkinson's self-reporting artefacts: reduce participant completion demand, design to offset the effect of tremor on input, enable implicit reminders, design for positive and negative consequences of increased awareness of symptoms, and consider the effects of handwritten notes in compliance, encoding burden, and data quality.

Scanpath analysis is a controversial and important topic in eye tracking research. Previous work has shown the value of scanpath analysis in perceptual tasks; little research has examined its utility for understanding human reasoning in complex tasks. Here, we analyze n-grams, which are continuous ordered subsequences of participants' scanpaths. In particular we studied the length of n-grams that are most appropriate for this form of analysis. We reuse datasets from previous studies of human cognition, medical diagnosis and art, systematically analyzing the frequency of n-grams of increasing length, and compare this approach with a string alignment-based method. The results show that subsequences of four or more areas of interest may not be of value for finding patterns that distinguish between two groups. The study is the first to systematically define the parameters of the length of n-gram suitable for analysis, using an approach that holds across diverse domains.

Objective: Allergies are increasing, but the reasons for this are unclear. Although environmental factors are thought to be important, there is a lack of data on how they contribute to symptom development. To understand this relationship better, we need accurate data about both symptoms and environmental factors. Our objective here is to ascertain whether experience sampling is a reliable approach for collecting allergy symptom data in the general population, allowing us to map symptoms and understand etiology. Materials and Methods: We conducted a 32-week cross-sectional study where individuals reported their seasonal allergy symptoms and severity via a mobile application. Symptom geographical location and timestamp were also collected automatically. Results: The experience sampling method reliably infers the incidence of seasonal allergies as indicated by the strong correlation (r = 0.93, P < .003) between the reported lack of wellness and the number of antihistamines prescribed by General Practitioners. Discussion and Conclusion: The project has resulted in the first dataset to map allergy symptoms over time and place and reveals periods of peak hay fever symptoms in the UK.

Despite the popularity of photo editors used to improve image attractiveness and expressiveness on social media, many users have trouble making sense of color filter effects and locating a preferred filter among a set of designer-crafted candidates. The problem gets worse when more computer-generated filters are introduced. To enhance filter findability, we semantically name and organize color effects leveraging data curated by creative communities online. We first model semantic mappings between color themes and keywords in everyday language. Next, we index and organize each filter by the derived semantic information. We conduct three separate studies to investigate the benefit of the semantic features on filter exploration. Our results indicate that color theme semantics constructed through social curation enhances filter findability, providing important implications into how to use the wisdom of the crowd to improve user experience with image editors.

Ontologies are complex systems of axioms in which unanticipated consequences of changes are both frequent, and difficult for ontology authors to apprehend. The effects of modelling actions range from unintended inferences to outright defects such as incoherency or even inconsistency. One of the central ontology authoring activities is verifying that a particular modelling step has had the intended consequences, often with the help of reasoners. For users of Protégé, this involves, for example, exploring the inferred class hierarchy. This paper provides evidence that making entailment set changes explicit to authors significantly improves the understanding of authoring actions regarding both correctness and speed. This is tested by means of the Inference Inspector, a Protégé plugin we created that provides authors with specific details about the effects of an authoring action. We empirically validate the effectiveness of the Inference Inspector in two studies. In a first, exploratory study we determine the feasibility of the Inference Inspector for supporting verification and isolating authoring actions. In a second, controlled study we formally evaluate the Inference Inspector and determine that making changes to key entailment sets explicit significantly improves author verification compared to the standard static hierarchy/frame-based approach. We discuss the advantages of the Inference Inspector for different types of verification questions and find that our approach is best suited for verifying added restrictions where no new signature, such as class names, is introduced, with a 42% improvement in verification correctness.

Parkinson's Disease impairs the motor, cognitive and emotional functioning of people. Clinicians do not get an accurate image of the disease because patients visit them every six months and their symptoms can change within hours. Technology has been used to tackle this problem, but most approaches disrupt people's routines or are uncomfortable to use. We aim to monitor Parkinson's in an unobtrusive, longitudinal and naturalistic way based on digital biomarkers inferred from smartphone-collected heterogeneous data. We use Parkinson's clinical scales and self-reporting of symptoms as ground truth to evaluate our methodology. Here, we present three insights we gained after tracking four people 24/7 for four months: a) the monitoring smartphone needs to be people's primary device, b) participants prefer a paper diary for symptom self-reporting, and c) social and phone interaction will be explored as digital biomarkers. We expect this approach to improve patient's quality of life and the efficiency of health services.

Many citizen science projects ask people to create an account before they participate - some require it. What effect does the registration process have on the number and quality of contributions? We present a controlled study comparing the effects of mandatory registration with an interface that enables people to participate without registering, but allows them to sign up to 'claim' contributions. We demonstrate that removing the requirement to register increases the number of visitors to the site contributing to the project by 62%, without reducing data quality. We also discover that contribution rates are the same for people who choose to register, and those who remain anonymous, indicating that the interface should cater for differences in participant motivation. The study provides evidence that to maximize contribution rates, projects should offer the option to create an account, but the process should not be a barrier to immediate contribution, nor should it be required.

Ontologies are complex knowledge representation artefacts used widely across biomedical, media and industrial domains. They are used for defining terminologies and providing metadata, especially for linked open data, and as such their use is rapidly increasing, but so far development tools have not benefited from empirical research into the ontology authoring process. This paper presents the results of a study that identifies common activity patterns through analysis of eye-tracking data and the event logs of the popular authoring tool, Protégé. Informed by the activity patterns discovered, we propose design guidelines for bulk editing, efficient reasoning and increased situational awareness. Methodological implications go beyond the remit of knowledge artefacts: we establish a method for studying the usability of software designed for highly specialised complex domains.

Ontologies have been employed across scientific and business domains for some time, and the proliferation of linked data means the number and range of potential authors is set to increase significantly. Ontologies using the Web Ontology Language (OWL) are complex artefacts, however: the authoring process requires not only knowledge of the application domain, but also skills in programming and logics. To date, there has been no systematic attempt to understand the effectiveness of existing tools, or explore what users really require to build successful ontologies. Here we address this shortfall, presenting insights from an interview study with 15 ontology authors. We identify the problems reported by authors, and the strategies they employ to solve them. We map the data to a set of design recommendations, which describe how tools of the future can support ontology authoring. A key challenge is dealing with information overload: improving the user's ability to navigate, populate and debug large ontologies will revolutionise the engineering process, and open ontology authoring up to a new generation of users.

The process of authoring ontologies appears to be fragmented across several tools and workarounds, and there exists no well accepted framework for common authoring tasks such as exploring ontologies, comparing versions, debugging, and testing. This lack of an adequate and seamless tool chain potentially hinders the broad uptake of ontologies, especially OWL, as a knowledge representation formalism. We start to address this situation by presenting insights from an interview-based study with 15 ontology experts. We uncover the tensions that may emerge between ontology authors including antagonistic ontology building styles (definition-driven vs. manually crafted hierarchies). We identify the problems reported by the ontology authors and the strategies they employ to solve them. These data are mapped to a set of key design recommendations, which should inform and guide future efforts for improving ontology authoring tool support, thus opening up ontology authoring to a new generation of users. We discuss future research avenues in light of these results.

The perception of the visual complexity of World Wide Web (Web) pages is a topic of significant interest. Previous work has examined the relationship between complexity and various aspects of presentation, including font styles, colours and images, but automatically quantifying this dimension of a web page at the level of the document remains a challenge. In this paper we demonstrate that areas of high complexity can be identified by detecting areas, or ‘chunks’, of a web page high in block-level elements. We report a computational algorithm that captures this metric and places web pages in a sequence that shows an 86% correlation with the sequences generated through user judgements of complexity. The work shows that structural aspects of a web page influence how complex a user perceives it to be, and presents a straightforward means of determining complexity through examining the DOM.

Dynamic micro-content—interactive or updating widgets and features—is now widely used on the Web, but there is little understanding of how people allocate attention to it. In this article we present the results of an eye-tracking investigation examining how the nature of dynamic micro-content influences whether or not the user views it. We propose and validate the Dynamic Update Viewing-likelihood (DUV) model, a CHi-squared Automatic Interaction Detector (CHAID) model that predicts with around 80% accuracy whether users view dynamic updates as a function of how they are initiated, their size, and their duration. The model is constructed with data from live Web sites and does not rely on knowledge of the user's task to make its predictions, giving it a high level of external validity. We discuss one example of its application: informing how dynamic content should be presented in audio via assistive technology for people with visual impairments.

This paper presents an eye-tracking study that examines how people use the visual elements of Web pages to complete certain tasks. Whilst these elements are available to play their role in these tasks for sighted users, it is not the case for visually disabled users. This lack of access to some visual elements of a page means that visually disabled users are hindered in accomplishing these tasks. Our previous work has introduced a framework that identifies these elements and then reengineers Web pages such that these elements can play their intended roles in an audio, as well as visual presentation. To further improve our understanding of how these elements are used and to validate our framework, we track the eye movements of sighted users performing a number of different tasks. The resulting gaze data show that there is a strong relationship between the aspects of a page that receive visual attention and the objects identified by our framework. The study also shows some limitations, as well as yielding information to address these short-comings. Perhaps the most important result is the support provided for a particular kind of object called a Way Edge - the visual construct used to group content into sections. There is a significant effect of Way Edges on the distribution of attention across tasks. This is a result that not only provides strong evidence for the utility of re-engineering, but also has consequences for our understanding of how people allocate attention to different parts of a page. We speculate that the phenomenon of 'Banner Blindness' owes as much to Way Edges, as it does to colour and font size.

Retrieving information presented visually is difficult for visually disabled users. Current accessibility technologies, such as screen readers, fail to convey presentational layout or structure. Information presented in graphs or images is almost impossible to convey through speech alone. In this paper, we present the results of an experimental study investigating the role of touch (haptic) and auditory cues in aiding structure recognition when visual presentation is missing. We hypothesize that by guiding users toward nodes in a graph structure using force fields, users will find it easier to recognize overall structure. Nine participants were asked to explore simple 3D structures containing nodes (spheres or cubes) laid out in various spatial configurations and asked to identify the nodes and draw their overall structure. Various combinations of haptic and auditory feedback were explored. Our results demonstrate that haptic cues significantly helped participants to quickly recognize nodes and structure. Surprisingly, auditory cues alone did not speed up node recognition; however, when they were combined with haptics both node identification and structure recognition significantly improved. This result demonstrates that haptic feedback plays an important role in enabling people to recall spatial layout.

Capturing detailed surface geometry currently requires specialized equipment such as laser range scanners, which despite their high accuracy, leave gaps in the surfaces that must be reconciled with photographic capture for relighting applications. Using only a standard digital camera and a single view, we present a method for recovering models of predominantly diffuse textured surfaces that can be plausibly relit and viewed from any angle under any illumination. Our multiscale shape-from-shading technique uses diffuse-lit/flash-lit image pairs to produce an albedo map and textured height field. Using two lighting conditions enables us to subtract one from the other to estimate albedo. In the absence of a flash-lit image of a surface for which we already have a similar exemplar pair, we approximate both albedo and diffuse shading images using histogram matching. Our depth estimation is based on local visibility. Unlike other depth-from-shading approaches, all operations are performed on the diffuse shading image in image space, and we impose no constant albedo restrictions. An experimental validation shows our method works for a broad range of textured surfaces, and viewers are frequently unable to identify our results as synthetic in a randomized presentation. Furthermore, in side-by-side comparisons, subjects found a rendering of our depth map equally plausible to one generated from a laser range scan. We see this method as a significant advance in acquiring surface detail for texturing using a standard digital camera, with applications in architecture, archaeological reconstruction, games and special effects.

We present a system that enables, for the first time, effective transatlantic cooperative haptic manipulation of objects whose motion is computed using a physically-based model. We propose a technique for maintaining synchrony between simulations in a peer-to-peer system, while providing responsive direct manipulation for all users. The effectiveness of this approach is determined through extensive user trials involving concurrent haptic manipulation of a shared object. A CAD assembly task, using physically-based motion simulation and haptic feedback, was carried out between the USA and the UK with network latencies in the order of 120ms. We compare the effects of latency on synchrony between peers over the Internet with a low latency (0.5ms) local area network. Both quantitatively and qualitatively, when using our technique, the performance achieved over the Internet is comparable to that on a LAN. As such, this technique constitutes a significant step forward for distributed haptic collaboration

Collaborative virtual environments (CVEs) enable two or more people, separated in the real world, to share the same virtual “space.” They can be used for many purposes, from teleconferencing to training people to perform assembly tasks. Unfortunately, the effectiveness of CVEs is compromised by one major problem: the delay that exists in the networks linking users together. Whilst we have a good understanding, especially in the visual modality, of how users are affected by delayed feedback from their own actions, little research has systematically examined how users are affected by delayed feedback from other people, particularly in environments that support haptic (force) feedback. The current study addresses this issue by quantifying how increasing levels of latency affect visual and haptic feedback in a collaborative target acquisition task. Our results demonstrate that haptic feedback in particular is very sensitive to low levels of delay. Whilst latency affects visual feedback from 50 ms, it impacts on haptic task performance 25 ms earlier, and causes the haptic measures of performance deterioration to rise far more steeply than visual. The “impact-perceive-adapt” model of user performance, which considers the interaction between performance measures, perception of latency, and the breakdown of perception of immediate causality, is proposed as an explanation for the observed pattern of performance.