Search

Health-related social control refers to intentional attempts to influence people's health behaviors, often seen in personal relationships. Social robots hold promise in influencing people's health by exerting health-related social control, but it is unclear which social control strategies used by robots are appropriate and potentially effective. This study investigates the effects of positive versus negative, and relationship-oriented versus target-oriented social control strategies from a social robot on people's psychological reactions. In an online video prototype study, participants viewed scenarios of a social robot attempting to change their sedentary behaviors by using different strategies. We found that positive (versus negative) strategies elicited stronger positive affect, enjoyment, and perceived social appropriateness, reduced perceived threats to freedom, and strengthened behavioral intention. Meanwhile, the relationship-oriented (versus target-oriented) strategies elevated people's negative affect, reduced enjoyment and perceived appropriateness, elevated perceived threats to freedom, and weakened behavioral intentions. Given these findings, we give recommendations for designing health influence strategies in social robots.

During the pre-design phase of supportive intelligent systems, interviews with future users, such as people with dementia, aim to explore the problems for which technology may present a solution. However, interviewing someone with dementia in this unstructured design phase is reported to be challenging resulting in systems that are not truly addressing the needs of the people with dementia. Existing guidelines on creating engaging interviews are often based on post-hoc reflections of researchers and participants. In this study we performed line-to-line sequential in-depth analyses on transcripts of 10 interviews conducted to co-design an interactive intelligent gaming system. Both interviewers and interviewees were subsequently asked to reflect on their experiences. Results show that information exchange with people with dementia can be supported by carefully constructed questions, well-designed probes, and an appropriate length of the interview. To create a pleasant atmosphere during the interviews, researchers should design the interview as a fulfilling experience, be prepared to respond respectfully to negative responses and be able to adapt to different signs of consent. Improving the engagement with people with dementia during interviews can enhance the design and development of intelligent systems that cater to their unique challenges and requirements.

Users with large domain knowledge can be reluctant to use prediction models. This also applies to the sports domain, where running coaches rarely rely on marathon prediction tools for race-plan advice for their runners’ next marathon. This paper studies the effect of adding interactivity to such prediction models, to incorporate and acknowledge users’ domain knowledge. In think-aloud sessions and an online study, we tested an interactive machine learning tool that allowed coaches to indicate the importance of earlier races feeding into the model. Our results show that coaches deploy rich knowledge when working with the model on runners familiar to them, and their adaptations improved model accuracy. Those coaches who could interact with the model displayed more trust and acceptance in the resulting predictions.

The application of social robots in persuading people to change health behaviors is an increasing research topic. However, little is known in what ways, and under what conditions, effective health persuasion can be achieved in human-robot interaction (HRI). This position paper presents a conceptual model that integrates interpersonal relationship theories to postulate a mechanism through which social robots can change people’s health behaviors. In this paper, we first briefly describe the two interpersonal relationship theories we selectively focus on, namely social control and interdependence theory, and we discuss the possibility of people forming relationships with social robots. Then, we propose the conceptual model depicting the potential positive and negative influence of social robots’ health persuasion on people’s psychological and behavioral reactions and the modulating role of human-robot relationships. Finally, we discuss the implications of this model for future research.

Artificial intelligence (AI) has been applied to health behavior change research for over a decade. Current research programs include machine learning for delivering just-in-time adaptive interventions, computational modeling of behavior change processes, and the use of social AI for communication and persuasion. With new advances in AI, we propose an international workshop to bring together experts from all related disciplines to discuss and explore the potentials of AI for behavior change research. We discuss in this proposal the aims, planned activities, expected outcomes, and a promotion strategy for the workshop.

The need for touch exists below the horizon of consciousness. Interpersonal touch, or gentle physical contact between individuals, is an important aspect of human social interaction and has been shown to have numerous benefits for senior individuals. One of the main ways that interpersonal touch can benefit senior individuals is by reducing feelings of loneliness and social isolation. The outbreak of COVID-19 necessitated social distancing measures to mitigate the negative health consequences of the pandemic, which were particularly pronounced amongst vulnerable populations, especially those who live with dementia. At the same time, seniors’ psychosocial wellbeing was compromised as opportunities for interpersonal touch became severely restricted. In this paper, we share what we have learned during the pandemic both on the critical importance of social touch for senior individuals, as well as on innovative ways to ameliorate the forced absence of social touch. Through a targeted literature review and an online survey study, our findings highlights the fact that social relationships and physical contact are key to personal wellbeing and underline the importance of social touch at an advanced age.

Advancements in artificial intelligence and (social) robotics raise pertinent questions as to how these technologies may help shape the society of the future. The main aim of the chapter is to consider the social and conceptual disruptions that might be associated with social robots, and humanoid social robots in particular. This chapter starts by comparing the concepts of robots and artificial intelligence, and briefly explores the origins of these expressions. It then explains what a social robot is, as well as what humanoid robots are. A key notion in this context is the idea of anthropomorphism: the human tendency to attribute human qualities, not only to our fellow human beings, but also to parts of nature and to technologies. This tendency to anthropomorphize technologies by responding to and interacting with them as if they have human qualities is one of the reasons why social robots – in particular social robots designed to look and behave like human beings – can be socially disruptive. As is explained in the chapter, while some ethics researchers think that anthropomorphization is a mistake that can lead to various forms of deception, others – including both ethics researchers and social roboticists – believe it can be useful or fitting to treat robots in anthropomorphizing ways. The chapter explores that disagreement by, among other things, considering recent philosophical debates about whether social robots can be moral patients; that is, whether it can make sense to treat them with moral consideration. Where one stands on this issue will depend on one’s views about whether social robots can have, imitate, or represent morally relevant properties, or on how people relate to social robots in their interactions with them. Lastly, the chapter urges that the ethics of social robots should explore intercultural perspectives, and highlights some recent research on ubuntu ethics and social robots.

As a consequence of the outbreak of the COVID-19 global pandemic in the spring of 2020, large-scale social distancing measures were implemented, resulting in the forced adoption of online or digital forms of psychological treatment. This sudden transition to digital care offered a unique opportunity to investigate if and how this experience impacted mental healthcare professionals’ perceptions and use of Digital Mental Health tools. The current paper presents findings of a repeated cross-sectional study consisting of three iterations of a national online survey in the Netherlands. This survey contained open and closed questions on professionals’ adoption readiness, frequency of use, perceived competency, and perceived value of Digital Mental Health collected in 2019 (before the pandemic), in 2020 (after the first wave), and in 2021 (after the second wave). The inclusion of data gathered prior to the COVID-19 pandemic offers a unique window to assess how professionals’ adoption has developed through this transition from voluntary to mandatory use of Digital Mental Health tools. Our study also re-assesses the drivers, barriers, and needs of mental healthcare professionals after having gained experience with Digital Mental Health. In total, 1,039 practitioners completed the surveys (Survey 1: n = 432, Survey 2: n = 363, and Survey 3: n = 244). Results indicate that compared to the period before the pandemic, there was a particularly large increase in use, competency, and perceived value regarding videoconferencing. Small differences were also found for some other basic tools that were crucial to ensure the continuation of care, such as e-mail, text messaging, and online screening, but not for more innovative technologies, such as virtual reality and biofeedback. Many practitioners reported to have gained skills regarding Digital Mental Health and experienced several benefits of it. They expressed the intention to continue with a blended approach, using Digital Mental H

Social touch is essential to human development and communication. Mediated social touch is suggested as a solution for circumstances where distance prevents skin-to-skin contact. However, past research aimed at demonstrating efficacy of mediated touch in reducing stress and promoting helping have produced mixed findings. These inconsistent findings could possibly be due to insufficient control of contextual factors combined with unnatural interaction scenarios. For example, touch occurs less frequently among strangers and is often accompanied with nonverbal visual cues. We investigated how visual presentation of touch, and interpersonal relationship to the sender influence perception, affective experiences, and autonomic responses the touch evoke. Fifty couples of mixed gender were recruited. A mediated touch was repeatedly applied by either the male partner or male confederate to female participants. The latter witnessed through a webcam as the sender caressed a rubber hand or touchpad to send the touch. Following our hypotheses, touch sent by one's partner was perceived softer and more comforting than stranger touch. The partner's touch also resulted in weaker skin conductance responses, particularly when sent by touching a touchpad. In sum, how a mediated touch is experienced depends both on who is touching, and on how the touch is visually represented.

As a consequence of the outbreak of the COVID-19 global pandemic in the spring of 2020, large-scale social distancing measures were implemented, resulting in the forced adoption of online or digital forms of psychological treatment. This sudden transition to digital care offered a unique opportunity to investigate if and how this experience impacted mental healthcare professionals’ perceptions and use of Digital Mental Health tools. The current paper presents findings of a repeated cross-sectional study consisting of three iterations of a national online survey in the Netherlands. This survey contained open and closed questions on professionals’ adoption readiness, frequency of use, perceived competency, and perceived value of Digital Mental Health collected in 2019 (before the pandemic), in 2020 (after the first wave), and in 2021 (after the second wave). The inclusion of data gathered prior to the COVID-19 pandemic offers a unique window to assess how professionals’ adoption has developed through this transition from voluntary to mandatory use of Digital Mental Health tools. Our study also re-assesses the drivers, barriers, and needs of mental healthcare professionals after having gained experience with Digital Mental Health. In total, 1,039 practitioners completed the surveys (Survey 1: n = 432, Survey 2: n = 363, and Survey 3: n = 244). Results indicate that compared to the period before the pandemic, there was a particularly large increase in use, competency, and perceived value regarding videoconferencing. Small differences were also found for some other basic tools that were crucial to ensure the continuation of care, such as e-mail, text messaging, and online screening, but not for more innovative technologies, such as virtual reality and biofeedback. Many practitioners reported to have gained skills regarding Digital Mental Health and experienced several benefits of it. They expressed the intention to continue with a blended approach, using Digital Menta

Biosensing technologies have advanced to a level where measurements of our physiology have become easily accessible, and new areas of application are being explored, including interpersonal communication via biosignals. To date, it is largely unknown what the communication characteristics of biosignal sharing are and how they influence experiences of biosignal-mediated interactions. By conceptualizing biosignal sharing as a new communication medium, we try to benefit from earlier theories and models that help describe and understand the communication properties that are inherently part of biosignal sharing, but have not previously been examined in this way. In this paper, we review the published literature on biosignal sharing and analyze the current state of the art using two complementary methods of analysis. First, we extend an existing framework of relevant communication dimensions and apply this to examine the characteristics of biosignal sharing systems. Second, we perform a thematic analysis on user experiences related to biosignal sharing available from literature. Then, we combine these findings and link the communication characteristics to the experiences of biosignal sharing. We discuss our findings, elucidating the way in which biosignal sharing affects the interpersonal communication process. Based on these insights, we provide several considerations for future research and design.

Autonomous mobile robots that operate in environments with people are expected to be able to deal with human proxemics and social distances. Previous research investigated how robots can approach persons or how to implement human-aware navigation algorithms. However, experimental research on how robots can avoid a person in a comfortable way is largely missing. The aim of the current work is to experimentally determine the shape and size of personal space of a human passed by a robot. In two studies, both a humanoid as well as a non-humanoid robot were used to pass a person at different sides and distances, after which they were asked to rate their perceived comfort. As expected, perceived comfort increases with distance. However, the shape was not circular: passing at the back of a person is more uncomfortable compared to passing at the front, especially in the case of the humanoid robot. These results give us more insight into the shape and size of personal space in human–robot interaction. Furthermore, they can serve as necessary input to human-aware navigation algorithms for autonomous mobile robots in which human comfort is traded off with efficiency goals.

Model-agnostic explainable AI tools explain their predictions by means of ’local’ feature contributions. We empirically investigate two potential improvements over current approaches. The first one is to always present feature contributions in terms of the contribution to the outcome that is perceived as positive by the user (“positive framing”). The second one is to add “semantic labeling”, that explains the directionality of each feature contribution (“this feature leads to +5% eligibility”), reducing additional cognitive processing steps. In a user study, participants evaluated the understandability of explanations for different framing and labeling conditions for loan applications and music recommendations. We found that positive framing improves understandability even when the prediction is negative. Additionally, adding semantic labels eliminates any framing effects on understandability, with positive labels outperforming negative labels. We implemented our suggestions in a package ArgueView[11].

Persuasion aims at changing peoples’ motivations and/or behaviors. This study explores how and when physiology reflects persuasion processes and specifically whether individual differences in motivations and behaviors affect psychophysiologic reactions to persuasive information. Participants (N = 70) with medium or high meat consumption patterns watched a persuasive video advocating limited meat consumption, while their electrodermal and cardiovascular physiology was measured. Results indicated that the video increased participants’ moral beliefs, perceived behavioral control, and reduction intentions. This study also found an increase in physiologic arousal during the persuasive video and that people with motivations less aligned to the persuasion objective had more physiologic arousal. The findings encourage further psychophysiologic persuasion research, especially as these insights can potentially be used to personalize persuasive messages of behavior change applications.

Measurement of physiological reactions to persuasive messages can improve our understanding of psychological processes of persuasion, and potentially further enhance and personalize current persuasion interventions. However, little is known about the relationship between psychophysiology and persuasive processes. This study focused on four persuasion principles: scarcity, commitment, consensus, and authority, and people's susceptibility to them. Physiological measures included the cardiovascular, respiratory, and electrodermal system, as well as facial motor systems. Psychological measures consisted of self-reported attitude towards oral care and susceptibility to persuasion (STPS). We performed a randomized within-subject experiment in which fifty-six participants viewed persuasive messages deploying the aforementioned persuasion principles to improve their oral care. Results indicated different physiological patterns during persuasion versus rest. We found no different physiological patterns in exposure to distinct persuasion principles, nor a clear correlation with susceptibility to individual persuasion principles. However, mixed model analysis illustrated that overall STPS scores help explain variance in reactivity of skin conductance level and skin conductance response, and reactivity in the zygomaticus major: lower susceptibility relates to higher reactivity. Summarizing, we have found no conclusive support for distinct psychophysiological patterns associated with different persuasion principles, although overall susceptibility seems to be reflected in physiology to some extent.

Psychological theories of habit posit that when a strong habit is formed through behavioral repetition, it can trigger behavior automatically in the same environment. Given the reciprocal relationship between habit and behavior, changing lifestyle behaviors is largely a task of breaking old habits and creating new and healthy ones. Thus, representing users’ habit strengths can be very useful for behavior change support systems, for example, to predict behavior or to decide when an intervention reaches its intended effect. However, habit strength is not directly observable and existing self-report measures are taxing for users. In this paper, building on recent computational models of habit formation, we propose a method to enable intelligent systems to compute habit strength based on observable behavior. The hypothesized advantage of using computed habit strength for behavior prediction was tested using data from two intervention studies on dental behavior change (N= 36 and N= 75), where we instructed participants to brush their teeth twice a day for three weeks and monitored their behaviors using accelerometers. The results showed that for the task of predicting future brushing behavior, the theory-based model that computed habit strength achieved an accuracy of 68.6% (Study 1) and 76.1% (Study 2), which outperformed the model that relied on self-reported behavioral determinants but showed no advantage over models that relied on past behavior. We discuss the implications of our results for research on behavior change support systems and habit formation.

Tracking health data, for example, through wearable devices or health apps, is increasingly commonplace. Consequently, health coaches (e.g., personal trainers, dieticians) are facing growing numbers of clients who bring their data to the clinic. These data potentially add value to the coaching process, for example, by showing more objective and specific information on clients' behaviors. However, in practice, it turns out to be hard to effectively utilize health data in a coaching setting, and it is not yet fully understood how data affect the coaching process and the coach-client communication. We organized a workshop (12 coaches, 3 clients) and a field study (5 coaches, 6 clients), where we observed coach-client interactions enriched with data. By including both familiar and unfamiliar coach-client pairs, as well as alternating the timing of the data presented (i.e., at the beginning, or halfway through the session), we acquired a variety of data-driven coaching interactions and analyzed this using a mixture of qualitative and quantitative methods. Our results show that data are not “plug-and-play.” There is an extensive process of interpreting and contextualizing data, in which the client has a key role, which is essential to gain relevant and actionable insights from the data useful to the coaching process. We also observed that data affect the coach-client communication on both content and relationship levels. We will reflect on these insights in terms of design recommendations for wearable tracking devices and e-health technology to effectively support health coaches and their interactions with their clients.

Autonomous mobile robots that operate in environments with people are expected to be able to deal with human proxemics and social distances. Previous research investigated how robots can approach persons or how to implement human-aware navigation algorithms. However, experimental research on how robots can avoid a person in a comfortable way is largely missing. The aim of the current work is to experimentally determine the shape and size of personal space of a human passed by a robot. In two studies, both a humanoid as well as a non-humanoid robot were used to pass a person at different sides and distances, after which they were asked to rate their perceived comfort. As expected, perceived comfort increases with distance. However, the shape was not circular: passing at the back of a person is more uncomfortable compared to passing at the front, especially in the case of the humanoid robot. These results give us more insight into the shape and size of personal space in human–robot interaction. Furthermore, they can serve as necessary input to human-aware navigation algorithms for autonomous mobile robots in which human comfort is traded off with efficiency goals.

Persuasive technology (PT) can assist in behavior change. PT systems often rely on user models, based on behavior and self-report data, to personalize their functionalities and thereby increase efficiency. This review paper shows how physiological measurements could be used to further improve user models for personalization of PT by means of bio-cybernetic loops and data-driven approaches. Furthermore, we outline the advantages of using physiological measures for personalization compared to self-report and behavior measurement. Additionally, we show how two types of physiological information—physiological states and physiological reactivity—can be relevant for PT adaptations. To illustrate this, we present a model with two types of physiology-based PT adaptations as part of a bio-cybernetic loop; state-based and reactivity-based. Next, we discuss the implications of physiology-aware PT for persuasive design and theory. And lastly, because of the potential impact of such systems, we also consider important ethical implications of physiology-aware PT.

Physiological responses to persuasion can help to increase our understanding of persuasive processes, and thereby the effectiveness of persuasive interventions. However, a clear relationship between psychophysiology and persuasion is not yet established. This study investigates if peripheral physiology predicts persuasion effectiveness, and whether peripheral physiology yields information that is not represented in other predictors of persuasion. We studied physiological reactions in the cardiovascular, electrodermal, facial muscle and respiratory systems of 75 participants while they read gain- or loss-framed persuasive messages advocating increased oral health care behavior. Persuasion effectiveness was measured as pre to post intervention changes in self-reported attitudes and intentions, as well as through changes in behavioral compliance over three weeks. Overall, participants showed stronger attitudes and intentions directly after the intervention (short-term persuasion), but did not show changes in behavior or attitudes two weeks later (no long-term persuasion). On an individual level, physiological reactivity parameters yielded additional information next to self-report measures to predict persuasion effectiveness on attitude, intention and behavioral compliance. To conclude, our findings suggested a positive relationship between physiological reactivity to persuasive messages and subsequent attitudes, intentions and behavior, and quantified the extra personalization that psychophysiological measures can bring to persuasive messaging

Background: A major factor hampering the adoption of technology in mental health care is a lack of knowledge and skills. Serious gaming offers a potentially effective strategy to enhance the skills needed through experiencing and learning-by-doing in a playful way. However, serious gaming solutions are not widely available for mental health care. Therefore, the development of a game-based training environment in mental health care was pursued in a design project. The first step in such a design project is to identify user requirements that should be met. Objective: This study aims to deliver user requirements that inform the design of a game-based training environment for mental health care professionals. This environment aims to support mental health care professionals’ knowledge and skill enhancement regarding the use of e–mental health (eMH); for example, video calling, mobile apps, web-based treatment modules, and techniques such as virtual or augmented reality. Methods: We used an exploratory multiple methods design consisting of a web-based questionnaire, co-design sessions, and interviews. To ensure a good representation of the target user group, professionals from various disciplines within mental health care were included in the research. The multiple methods design facilitates a broad view of user needs and in-depth knowledge of specific design requirements. We describe the protocol for this research project in a protocol paper published in the JMIR Research Protocols in February 2021. Results: The user requirements analysis revealed three types of users for the envisioned game-based training environment: mental health care professionals who want to learn about the basic possibilities of eMH, mental health care professionals who want to develop their eMH skills to the next level, and mental health care professionals who want to experiment with new technologies. This reflects the diversity of needs that were identi

Affective haptic devices (AHDs) have been developed with the aim of communicating touch acts, symbolic messages, emotions, and/or providing a sense of social awareness. Within AHDs, three categories can be distinguished: mediated social touch (MST), symbolic communication systems, and awareness systems. For each of these categories, prototypes have been developed and discussed in the literature. Each such prototype, however, describes but a small part of the design space of AHDs. What is lacking is a description of the design space itself—of all choices that can be considered during the design process. Such a description will allow for a more systematic exploration of AHD designs and provides a means of combining insights gained from individual point solutions (i.e., existing prototypes). Therefore, in this article, we provide a systematic description of the design space of AHDs and its underlying dimensions based on general (e.g., revisability or synchronicity) and AHD-specific (e.g., actuation type) communication system characteristics. This resulted in 17 design dimensions, each consisting of two or more categories (the design characteristics). Based on a systematic literature search from devices up to 2019, 89 AHD prototypes were identified, and each was classified on the design dimensions. The empirical analysis of where these AHDs are located in the design space revealed, first, that potentially interesting characteristics from mediated communication, such as revisability and reviewability, are underexplored in AHDs. Second, MST devices were found to often lack those system characteristics, such as real-time modalities, that seem crucial for providing the affordances needed to simulate social touch. In particular, when comparing symbolic and awareness devices to MST devices, we found the latter to more frequently lack some of the key characteristics of face-to-face communication (i.e., bi-directional and symmetric communication). Limitations and implication

Affective haptic devices (AHDs) are communication technologies utilizing the sense of touch, and include mediated social touch (MST), symbolic haptic messaging, and awareness systems that, for example, let one feel another person's heartbeat. The COVID-19 pandemic and consequent social distancing measures have led to a reemphasis of the importance of social touch, and many people have experienced firsthand what it is like to miss touching loved ones. This offers an excellent opportunity to study people's intention to use AHDs. For this purpose, a survey study (n = 277) was conducted combining qualitative and quantitative data analysis methods. Touch deprivation, resulting from not being able to touch a loved one, was associated with intention to use AHDs: the more deprived an individual, the higher his or her intention to use AHDs. Technology readiness and touch aversion did not affect intention to use AHDs. AHDs for symbolic messaging gained higher interest than MST and awareness devices, and long-distance relationships were seen as the most likely scenario for using AHDs. Bi-directionality, synchronicity, and symmetry were regarded as important features for providing shared meaning and a sense of connectedness. Reviewability, multimodality, and actuation type were also deemed important. Limitations of the study and implications for the design of AHDs are discussed.

Digital support during self-regulated learning can improve metacognitive knowledge and skills in learners. Previous research has predominantly focused on embedding metacognitive support in domain-specific content. In this study, we examine a detached approach where digital metacognitive support is offered in parallel to ongoing domain-specific training via a digital tool. The primary support mechanism was self-explication, where learners are prompted to make, otherwise implicit, metacognition concrete. In a controlled pre-test/post-test quasi-experiment, we compared domain-specific and domain-general support and assessed the effects, use, and learners' perceptions of the tool. The results showed that self-explication is an effective mechanism to support and improve metacognition during self-regulated learning. Furthermore, the results confirm the effectiveness of offering detached metacognitive support. While only domain-specific metacognitive support was found to be effective, quantitative and qualitative analysis warrant further research into domain-general and detached metacognitive support. The results also indicated that, while students with higher metacognition found a lack of relevance of using the tool, students with lower metacognition are less likely to make (structural) use of the available support. A key challenge for future research is thus to adapt metacognitive support to learner needs, and to provide metacognitive support to those who would benefit from it the most. The paper concludes by formulating implications for future research as well as design of digital metacognitive support.

Robots navigate ever more often in close proximity to people. In the current work, we focused on two distinctive navigational scenarios: passing and overtaking a person who is walking. In the first experiment, we compared nine different passing distances for a humanoid robot and found that human comfort increased with passing distance and that their relationship could be described by an inverted Gaussian. In the second experiment, we validated this relationship for an industrial autonomous robot and extended the study to also include overtaking distances and different robot moving speeds. The results showed that overtaking was considered to be less comfortable than passing but that the overtaking distance had a similar relationship with human comfort. Human comfort decreases with a higher robot movement speed. Results obtained through location trackers furthermore showed that people actively take a larger distance from the robot when it starts its trajectory closer to them. The current results can be used to quantify human comfort in environments where humans and robots co-exist and they can be used as input for human-aware navigational models for autonomous robots.

In what ways could the future of emotional bonds between humans and conversational AI change us? To explore this question in a multi-faceted manner, designers, engineers, philosophers as separate focus groups were given a design fiction probe— a story of a chatbot’s disappearance from a person’s life. Though articulated in discipline-specific ways, participants expressed similar concerns and hopes: 1) caring for a machine could teach people to emotionally care for themselves and others, 2) the boundary between human and non-human emotions may become blurred when people project their own emotions onto AI, e.g., a bot’s ”breakdown” as one’s own, and 3) people may then intertwine their identities with AI through emotions. We consider ethical ramifications of socially constructed emotions between humans and conversational agents.

Psychological theories of habit posit that when a strong habit is formed through behavioral repetition, it can trigger behavior automatically in the same environment. Given the reciprocal relationship between habit and behavior, changing lifestyle behaviors is largely a task of breaking old habits and creating new and healthy ones. Thus, representing users’ habit strengths can be very useful for behavior change support systems, for example, to predict behavior or to decide when an intervention reaches its intended effect. However, habit strength is not directly observable and existing self-report measures are taxing for users. In this paper, building on recent computational models of habit formation, we propose a method to enable intelligent systems to compute habit strength based on observable behavior. The hypothesized advantage of using computed habit strength for behavior prediction was tested using data from two intervention studies on dental behavior change (N= 36 and N= 75), where we instructed participants to brush their teeth twice a day for three weeks and monitored their behaviors using accelerometers. The results showed that for the task of predicting future brushing behavior, the theory-based model that computed habit strength achieved an accuracy of 68.6% (Study 1) and 76.1% (Study 2), which outperformed the model that relied on self-reported behavioral determinants but showed no advantage over models that relied on past behavior. We discuss the implications of our results for research on behavior change support systems and habit formation.

Model-agnostic explainable AI tools explain their predictions by means of ’local’ feature contributions. We empirically investigate two potential improvements over current approaches. The first one is to always present feature contributions in terms of the contribution to the outcome that is perceived as positive by the user (“positive framing”). The second one is to add “semantic labeling”, that explains the directionality of each feature contribution (“this feature leads to +5% eligibility”), reducing additional cognitive processing steps. In a user study, participants evaluated the understandability of explanations for different framing and labeling conditions for loan applications and music recommendations. We found that positive framing improves understandability even when the prediction is negative. Additionally, adding semantic labels eliminates any framing effects on understandability, with positive labels outperforming negative labels. We implemented our suggestions in a package ArgueView[11].

Persuasion aims at changing peoples’ motivations and/or behaviors. This study explores how and when physiology reflects persuasion processes and specifically whether individual differences in motivations and behaviors affect psychophysiologic reactions to persuasive information. Participants (N = 70) with medium or high meat consumption patterns watched a persuasive video advocating limited meat consumption, while their electrodermal and cardiovascular physiology was measured. Results indicated that the video increased participants’ moral beliefs, perceived behavioral control, and reduction intentions. This study also found an increase in physiologic arousal during the persuasive video and that people with motivations less aligned to the persuasion objective had more physiologic arousal. The findings encourage further psychophysiologic persuasion research, especially as these insights can potentially be used to personalize persuasive messages of behavior change applications.

Tracking health data, for example, through wearable devices or health apps, is increasingly commonplace. Consequently, health coaches (e.g., personal trainers, dieticians) are facing growing numbers of clients who bring their data to the clinic. These data potentially add value to the coaching process, for example, by showing more objective and specific information on clients' behaviors. However, in practice, it turns out to be hard to effectively utilize health data in a coaching setting, and it is not yet fully understood how data affect the coaching process and the coach-client communication. We organized a workshop (12 coaches, 3 clients) and a field study (5 coaches, 6 clients), where we observed coach-client interactions enriched with data. By including both familiar and unfamiliar coach-client pairs, as well as alternating the timing of the data presented (i.e., at the beginning, or halfway through the session), we acquired a variety of data-driven coaching interactions and analyzed this using a mixture of qualitative and quantitative methods. Our results show that data are not “plug-and-play.” There is an extensive process of interpreting and contextualizing data, in which the client has a key role, which is essential to gain relevant and actionable insights from the data useful to the coaching process. We also observed that data affect the coach-client communication on both content and relationship levels. We will reflect on these insights in terms of design recommendations for wearable tracking devices and e-health technology to effectively support health coaches and their interactions with their clients.

Measurement of physiological reactions to persuasive messages can improve our understanding of psychological processes of persuasion, and potentially further enhance and personalize current persuasion interventions. However, little is known about the relationship between psychophysiology and persuasive processes. This study focused on four persuasion principles: scarcity, commitment, consensus, and authority, and people's susceptibility to them. Physiological measures included the cardiovascular, respiratory, and electrodermal system, as well as facial motor systems. Psychological measures consisted of self-reported attitude towards oral care and susceptibility to persuasion (STPS). We performed a randomized within-subject experiment in which fifty-six participants viewed persuasive messages deploying the aforementioned persuasion principles to improve their oral care. Results indicated different physiological patterns during persuasion versus rest. We found no different physiological patterns in exposure to distinct persuasion principles, nor a clear correlation with susceptibility to individual persuasion principles. However, mixed model analysis illustrated that overall STPS scores help explain variance in reactivity of skin conductance level and skin conductance response, and reactivity in the zygomaticus major: lower susceptibility relates to higher reactivity. Summarizing, we have found no conclusive support for distinct psychophysiological patterns associated with different persuasion principles, although overall susceptibility seems to be reflected in physiology to some extent.

Affective haptic devices (AHDs) are communication technologies utilizing the sense of touch, and include mediated social touch (MST), symbolic haptic messaging, and awareness systems that, for example, let one feel another person's heartbeat. The COVID-19 pandemic and consequent social distancing measures have led to a reemphasis of the importance of social touch, and many people have experienced firsthand what it is like to miss touching loved ones. This offers an excellent opportunity to study people's intention to use AHDs. For this purpose, a survey study (n = 277) was conducted combining qualitative and quantitative data analysis methods. Touch deprivation, resulting from not being able to touch a loved one, was associated with intention to use AHDs: the more deprived an individual, the higher his or her intention to use AHDs. Technology readiness and touch aversion did not affect intention to use AHDs. AHDs for symbolic messaging gained higher interest than MST and awareness devices, and long-distance relationships were seen as the most likely scenario for using AHDs. Bi-directionality, synchronicity, and symmetry were regarded as important features for providing shared meaning and a sense of connectedness. Reviewability, multimodality, and actuation type were also deemed important. Limitations of the study and implications for the design of AHDs are discussed., Human Information Communication Design

Despite the growing popularity of digital health interventions, limitations of traditional behavior change theories and a lack of theory integration hinder theory-driven behavior change applications. In this paper, we aim to review theories relevant to lifestyle behavior change from the broader psychology literature and then integrate these theories into a new theoretical framework called adaptive decision-making to address two specific problems. First, our framework represents lifestyle behaviors at two levels-one of individual daily decisions (action level) and one of larger behavioral episodes (reflection level)-to more closely match the temporal characteristics of lifestyle behaviors and their associated digital data. Second, the framework connects decision-making theories and learning theories to explain how behaviors and cognitive constructs dynamically influence each other, making it a suitable scaffold for building computational models. We map common digital intervention techniques onto the behavioral and cognitive processes in the framework and discuss possible contributions of the framework to both theory development and digital intervention design.

Psychological theories of habit posit that when a strong habit is formed through behavioral repetition, it can trigger behavior automatically in the same environment. Given the reciprocal relationship between habit and behavior, changing lifestyle behaviors (e.g., toothbrushing) is largely a task of breaking old habits and creating new and healthy ones. Thus, representing users' habit strengths can be very useful for behavior change support systems (BCSS), for example, to predict behavior or to decide when an intervention reaches its intended effect. However, habit strength is not directly observable and existing self-report measures are taxing for users. In this paper, built on recent computational models of habit formation, we propose a method to enable intelligent systems to compute habit strength based on observable behavior. The hypothesized advantage of using computed habit strength for behavior prediction was tested using data from two intervention studies, where we trained participants to brush their teeth twice a day for three weeks and monitored their behaviors using accelerometers. Through hierarchical cross-validation, we found that for the task of predicting future brushing behavior, computed habit strength clearly outperformed self-reported habit strength (in both studies) and was also superior to models based on past behavior frequency (in the larger second study). Our findings provide initial support for our theory-based approach of modeling user habits and encourages the use of habit computation to deliver personalized and adaptive interventions.

In “Theory Integration for Lifestyle Behavior Change in the Digital Age: An Adaptive Decision-Making Framework” (J Med Internet Res 2021;23(4):e17127) the authors noted one error. In the originally published manuscript, the heading “Mapping Digital Intervention Techniques to the Framework” was incorrectly set as a level 3 subheading rather than a level 2 heading. This has been changed to a level 2 heading in the corrected version of the manuscript. The correction will appear in the online version of the paper on the JMIR Publications website on April 15, 2021, together with the publication of this correction notice. Because this was made after submission to PubMed, PubMed Central, and other full-text repositories, the corrected article has also been resubmitted to those repositories.

Studies on practitioners' perceptions of eMental Health repeatedly report experienced difficulties in establishing empathic interactions with their clients online – a fact that plays an important role in the adoption, or lack thereof, of technology-mediated treatments. Especially now, in times when social distancing and lockdown measures are effectuated globally to combat the COVID-19 pandemic, the need for high-quality online formats of mental health care has become more urgent than ever. To increase our understanding of the process of achieving empathic interactions in technology-mediated communication, the current paper presents an online survey study using open questions on practitioners' experiences of establishing empathy in online therapeutic interactions (n = 363). Responses were analyzed using thematic analysis. From the derived themes, a conceptual model was built showing several properties of online communication that influence empathy and how these properties affect the therapeutic interaction on emotional, conversational and relational levels. Additionally, our findings highlight the behavioral strategies practitioners employ to find workarounds in order to manage these effects and attempt to (re)connect with their clients. Our findings thus provide experience-based insights into the process of building empathy in online treatments – insights that can be used to improve current and future eMental Health technologies, enabling practitioners to reach the desired level of empathic understanding during remote therapeutic interactions.

Background: Over the last decades, significant advances have been made in the development of digital tools and applications for mental healthcare. Yet, despite growing evidence for their effectiveness, their acceptance and use in clinical practice remain low. To gain further insights in the process of eMental Health adoption and to facilitate future research on this topic, a validated and easy-to-use instrument to assess professionals' readiness to adopt eMental Health is necessary. Objective: The aim of this study was to develop and validate an instrument for assessing mental healthcare professionals' adoption readiness for eMental Health. Methods: Item generation was guided by literature and input from mental healthcare professionals and experts in survey development. Exploratory factor analyses were conducted on an initial set of 29 items completed by a sample of mental healthcare professionals (N = 432), after which the scale was reduced to 15 items in an iterative process. The factor structure thus obtained was subsequently tested with a confirmatory factor analysis with a second sample of mental healthcare professionals (N = 363). Internal consistency, convergent validity and predictive validity of the eMHAR Scale were assessed. Results: Exploratory factor analyses resulted in a three-factor solution with 15 items. The factors were analyzed and labeled as ‘perceived benefits and applicability of EMH’, ‘EMH proactive innovation’, and ‘EMH self-efficacy’. These factors were confirmed through a confirmatory factor analysis. The total scale and subscales showed good internal consistency (Cronbach’s alpha = 0.73-0.88) and acceptable convergent and predictive relations to related constructs. Conclusions: The constructed eMHAR Scale showed a conceptually interpretable three-factor structure with satisfactory characteristics and relationships with relevant concepts. Its ease of use allows for a quick acquisition of data that can contribute to unders

Though educators often deal with stressful social conflicts, many face them ad hoc without much training. We studied if and how virtual agents can help University staff manage student-teacher conflicts. We explored educators' verbal, behavioral, and physiological reactions to a virtual agent that brought up a student-teacher conflict and held exit-interviews. Our qualitative analysis revealed that virtual agents for conflict training were positively received, but not for conflict mediation with cross-cultural differences. Those with non-Western backgrounds felt that an agent could help “save face,” whereas Westerners preferred to resolve conflicts in person. In line with this, participants with a Western background rated the virtual agent to be less competent compared to those with non-Western backgrounds. While physiological measures only allow for limited conclusions, we found that participants who believed that the agent was controlled by a human had higher normalized hear rate variability (for the entire conversation in total) than people who thought that the agent was autonomous. We discuss implications for implementing virtual agents for training purposes, the impact of physiological signals, and the need to consider cultural and individual differences.

Many of the choices that designers and engineers make during a design process impact not only the functionality, usability, or aesthetics of a technology, but also impact the values that might be supported or undermined via the technology design. Designers can actively design for values, and this awareness has led to the development of various ‘ethics by design' approaches. One such approach is capability sensitive design (CSD). Thus far, CSD is only developed from a theoretical-ethical point of view. This article aims to bridge the theory-practice gap by entering into dialogue with various design-experts on ethics by design in general and CSD in particular. An empirical study, consisting of thematic interviews with nine design-experts, was conducted in order to explore design-experts' experiences with designing for values, what they regard as the strengths and weaknesses of CSD, and if CSD could be of practical use to their design (research) practice.

As robots may take a greater part in our moral decision-making processes, whether people hold them accountable for moral harm becomes critical to explore. Blame and punishment signify moral accountability, often involving emotions. We quantitatively looked into people’s willingness to blame or punish an emotional vs. non-emotional robot that admits to its wrongdoing. Studies 1 and 2 (online video interaction) showed that people may punish a robot due to its lack of perceived emotional capacity than its perceived agency. Study 3 (in the lab) demonstrated that people were neither willing to blame nor punish the robot. Punishing non-emotional robots seems more likely than blaming them, yet punishment towards robots is more likely to arise online than offline. We reflect on if and why victimized humans (and those who care for them) may seek out retributive justice against robot scapegoats when there are no humans to hold accountable for moral harm.

Studies on practitioners' perceptions of eMental Health repeatedly report experienced difficulties in establishing empathic interactions when communicating with their clients online – a fact that plays an important role in the adoption, or lack thereof, of technology-mediated treatments. Especially now, in times when social distancing and lockdown measures are effectuated globally to combat the COVID-19 pandemic, the need for high-quality online formats of mental health care has become more urgent than ever. To increase our understanding of the process of achieving empathic interactions in technology-mediated communication, the current paper presents an online survey study using open questions on practitioners' experiences of establishing empathy in online therapeutic interactions (n = 363). Responses were analyzed using thematic analysis. From the derived themes, a conceptual model was built showing several properties of online communication that influence empathy and ways in which these properties affect the therapeutic interaction on emotional, conversational, and relational levels. Additionally, our findings highlight the behavioral strategies practitioners employ to find workarounds to manage these effects and attempt to (re)connect with their clients. Our findings thus provide experience-based insights into the process of building empathy in online treatments – insights that can be used to improve current and future eMental Health technologies, enabling practitioners to reach the desired level of empathic understanding during remote therapeutic interactions.

Studies on practitioners' perceptions of eMental Health repeatedly report experienced difficulties in establishing empathic interactions with their clients online – a fact that plays an important role in the adoption, or lack thereof, of technology-mediated treatments. Especially now, in times when social distancing and lockdown measures are effectuated globally to combat the COVID-19 pandemic, the need for high-quality online formats of mental health care has become more urgent than ever. To increase our understanding of the process of achieving empathic interactions in technology-mediated communication, the current paper presents an online survey study using open questions on practitioners' experiences of establishing empathy in online therapeutic interactions (n = 363). Responses were analyzed using thematic analysis. From the derived themes, a conceptual model was built showing several properties of online communication that influence empathy and how these properties affect the therapeutic interaction on emotional, conversational and relational levels. Additionally, our findings highlight the behavioral strategies practitioners employ to find workarounds in order to manage these effects and attempt to (re)connect with their clients. Our findings thus provide experience-based insights into the process of building empathy in online treatments – insights that can be used to improve current and future eMental Health technologies, enabling practitioners to reach the desired level of empathic understanding during remote therapeutic interactions.

Studies on practitioners' perceptions of eMental Health repeatedly report experienced difficulties in establishing empathic interactions when communicating with their clients online – a fact that plays an important role in the adoption, or lack thereof, of technology-mediated treatments. Especially now, in times when social distancing and lockdown measures are effectuated globally to combat the COVID-19 pandemic, the need for high-quality online formats of mental health care has become more urgent than ever. To increase our understanding of the process of achieving empathic interactions in technology-mediated communication, the current paper presents an online survey study using open questions on practitioners' experiences of establishing empathy in online therapeutic interactions (n = 363). Responses were analyzed using thematic analysis. From the derived themes, a conceptual model was built showing several properties of online communication that influence empathy and ways in which these properties affect the therapeutic interaction on emotional, conversational, and relational levels. Additionally, our findings highlight the behavioral strategies practitioners employ to find workarounds to manage these effects and attempt to (re)connect with their clients. Our findings thus provide experience-based insights into the process of building empathy in online treatments – insights that can be used to improve current and future eMental Health technologies, enabling practitioners to reach the desired level of empathic understanding during remote therapeutic interactions.

In “Theory Integration for Lifestyle Behavior Change in the Digital Age: An Adaptive Decision-Making Framework” (J Med Internet Res 2021;23(4):e17127) the authors noted one error. In the originally published manuscript, the heading “Mapping Digital Intervention Techniques to the Framework” was incorrectly set as a level 3 subheading rather than a level 2 heading. This has been changed to a level 2 heading in the corrected version of the manuscript. The correction will appear in the online version of the paper on the JMIR Publications website on April 15, 2021, together with the publication of this correction notice. Because this was made after submission to PubMed, PubMed Central, and other full-text repositories, the corrected article has also been resubmitted to those repositories.

Psychological theories of habit posit that when a strong habit is formed through behavioral repetition, it can trigger behavior automatically in the same environment. Given the reciprocal relationship between habit and behavior, changing lifestyle behaviors (e.g., toothbrushing) is largely a task of breaking old habits and creating new and healthy ones. Thus, representing users' habit strengths can be very useful for behavior change support systems (BCSS), for example, to predict behavior or to decide when an intervention reaches its intended effect. However, habit strength is not directly observable and existing self-report measures are taxing for users. In this paper, built on recent computational models of habit formation, we propose a method to enable intelligent systems to compute habit strength based on observable behavior. The hypothesized advantage of using computed habit strength for behavior prediction was tested using data from two intervention studies, where we trained participants to brush their teeth twice a day for three weeks and monitored their behaviors using accelerometers. Through hierarchical cross-validation, we found that for the task of predicting future brushing behavior, computed habit strength clearly outperformed self-reported habit strength (in both studies) and was also superior to models based on past behavior frequency (in the larger second study). Our findings provide initial support for our theory-based approach of modeling user habits and encourages the use of habit computation to deliver personalized and adaptive interventions.

Despite the growing popularity of digital health interventions, limitations of traditional behavior change theories and a lack of theory integration hinder theory-driven behavior change applications. In this paper, we aim to review theories relevant to lifestyle behavior change from the broader psychology literature and then integrate these theories into a new theoretical framework called adaptive decision-making to address two specific problems. First, our framework represents lifestyle behaviors at two levels-one of individual daily decisions (action level) and one of larger behavioral episodes (reflection level)-to more closely match the temporal characteristics of lifestyle behaviors and their associated digital data. Second, the framework connects decision-making theories and learning theories to explain how behaviors and cognitive constructs dynamically influence each other, making it a suitable scaffold for building computational models. We map common digital intervention techniques onto the behavioral and cognitive processes in the framework and discuss possible contributions of the framework to both theory development and digital intervention design.

Though educators often deal with stressful social conflicts, many face them ad hoc without much training. We studied if and how virtual agents can help University staff manage student-teacher conflicts. We explored educators' verbal, behavioral, and physiological reactions to a virtual agent that brought up a student-teacher conflict and held exit-interviews. Our qualitative analysis revealed that virtual agents for conflict training were positively received, but not for conflict mediation with cross-cultural differences. Those with non-Western backgrounds felt that an agent could help “save face,” whereas Westerners preferred to resolve conflicts in person. In line with this, participants with a Western background rated the virtual agent to be less competent compared to those with non-Western backgrounds. While physiological measures only allow for limited conclusions, we found that participants who believed that the agent was controlled by a human had higher normalized hear rate variability (for the entire conversation in total) than people who thought that the agent was autonomous. We discuss implications for implementing virtual agents for training purposes, the impact of physiological signals, and the need to consider cultural and individual differences.