Search

People often engage in various forms of self-disclosure and social sharing with others when trying to regulate the impact of emotional distress. Here we introduce a novel long-term mediated intervention aimed at supporting informal caregivers to cope with emotional distress via self-disclosing their emotions and needs to a social robot. Research has shown that informal caregivers often struggle in managing the emotional and practical demands of the caregiving situation, and also highlights the lack of social support and paucity of social interaction some experience. Accordingly, we were interested in the extent of informal caregivers' self-disclosure behaviour towards a social robot (Pepper, SoftBank Robotics) over time, and how (social and usability-related) perceptions of the robot develop over time. Moreover, we wished to examine how this intervention made informal caregivers feel (in terms of reported mood, perceptions of the robot as comforting, feelings of loneliness, and stress), and the extent to which interacting with the robot affected these individuals' emotion regulation. Informal caregivers conversed with the social robot Pepper 10 times across 5 weeks about general everyday topics. Our results show that informal caregivers self-disclosed increasingly more to the robot across time and perceived it as increasingly social and competent over time. Furthermore, participants' moods positively changed after interacting with the robot, which they perceived more comforting over time. Participants also reported feeling increasingly less lonely and stressed. Finally, our results showed that after self-disclosing to the robot for 5 weeks, informal caregivers reported being more accepting of their caregiving situation, reappraising it more positively, and experiencing fewer feelings of blame towards others. These results set the stage for situating social robots as conversational partners in social settings, as well as highlight how communicating with social robots holds potential for providing emotional support for people coping with emotional distress.

Predictors of the endorsement of gendered romance norms among single women

Previous research suggests that humans are not reliable judges of the amount of synchrony between individuals who are moving their bodies, but that the presence of more synchronicity enhanced how much the observers enjoyed watching the movements. This study aims to ascertain whether humans really are poor judges of synchrony, or successful quantification of synchrony is dependent on kinematic features of the movement, aesthetic judgements made by the perceiver, or perhaps even perceiver traits.

Alcohol abuse, reported by 1/8th critically ill patients, is an independent risk factor for death in sepsis. Sepsis kills over 270,000 patients/year in the US. We reported that the ethanol-exposure suppresses innate-immune response, pathogen clearance, and decreases survival in sepsis-mice via sirtuin 2 (SIRT2). SIRT2 is an NAD+-dependent histone-deacetylase with anti-inflammatory properties. We hypothesized that in ethanol-exposed macrophages, SIRT2 suppresses phagocytosis and pathogen clearance by regulating glycolysis. Immune cells use glycolysis to fuel increased metabolic and energy demand of phagocytosis. Using ethanol-exposed mouse bone marrow- and human blood monocyte-derived macrophages, we found that SIRT2 mutes glycolysis via deacetylating key glycolysis regulating enzyme phosphofructokinase-platelet isoform (PFKP), at mouse lysine 394 (mK394, human: hK395). Acetylation of PFKP at mK394 (hK395) is crucial for PFKP function as a glycolysis regulating enzyme. The PFKP also facilitates phosphorylation and activation of autophagy related protein 4B (Atg4B). Atg4B activates microtubule associated protein 1 light chain-3B (LC3). LC3 is a driver of a subset of phagocytosis, the LC3-associated phagocytosis (LAP), which is crucial for segregation and enhanced clearance of pathogens, in sepsis. We found that in ethanol-exposed cells, the SIRT2-PFKP interaction leads to decreased Atg4B-phosphorylation, decreased LC3 activation, repressed phagocytosis and LAP. Genetic deficiency or pharmacological inhibition of SIRT2 reverse PFKP-deacetylation, suppressed LC3-activation and phagocytosis including LAP, in ethanol-exposed macrophages to improve bacterial clearance and survival in ethanol with sepsis mice.

Openly available natural language generation (NLG) algorithms can generate human-like texts across domains. Given their potential, ethical challenges arise such as being used as a tool for misinformation. It is necessary to understand both how these texts are generated from an algorithmic point of view, and how they are evaluated by a general audience. In this study, our aim was to investigate how people react to texts generated algorithmically, whether they are indistinguishable from original/human-generated texts, and the value people assign these texts. Using original text-based archives, and text-based archives generated by artificial intelligence (AI), findings from our pre-registered study (N = 228) revealed that people were more likely to preserve original archives compared to AI-generated archives. Although participants were unable to accurately distinguish between AI generated and original archives, participants assigned lower value to archives they categorized as AI-generated compared to those they categorized as original. People's judgements of value were also influenced by their attitudes toward AI. These findings provide a richer understanding of how the emergent practice of automated text creation alters the practices of readers and writers, and have implications for how readers' attitudes toward AI affect the use and value of AI-based applications and creations.

Using subjective behavioural and objective cortical (fNIRS) measures, we aim to explore the extent to which embodied experience, sociocultural background, and performance ability impact aesthetic perceptions of dance choreography.

Preregistration plan for a study investigating the role of benevolent sexism, within the mediated association between division of labour and sexual desire, through perceived fairness of division of labour.

This study will assess the impact of previous arts experience and sociocultural background on aesthetic perceptions of dance movements.

The potential for robots to support education is being increasingly studied and rapidly realised. However, most research evaluating education robots has neglected to examine the fundamental features that make them more or less effective, given the needs and expectations of learners. This study explored how children’s perceptions, expectations and experiences are shaped by aesthetic and functional features during interactions with different robot ‘reading buddies’. We collected a range of quantitative and qualitative measures of subjective experience before and after children read a book with one of three different robots. An inductive thematic analysis revealed that robots have the potential offer children an engaging and non-judgemental social context to promote reading engagement. This was supported by children’s perceptions of robots as being intelligent enough to read, listen and comprehend the story, particularly when they had the capacity to talk. A key challenge in the use of robots for this purpose was the unpredictable nature of robot behaviour, which remains difficult to perfectly control and time using either human operators or autonomous algorithms. Consequently, some children found the robots’ responses distracting. We provide recommendations for future research seeking to position seemingly sentient and intelligent robots as an assistive tool within and beyond education settings., Scientific Reports, 13 (1), ISSN:2045-2322

Intention prediction often plays a crucial role in successful social interaction. Previous studies have attempted to understand this skill by focusing on the role of movement kinematics in isolation. However, this approach is limited as the same kinematics typically map to multiple action possibilities (affordances) and as a result, individual's also employ contextual information to predict others' intentions. In this study we present preliminary findings from a qualitative study aimed at investigating intention prediction in naturalistic contexts. Participants viewed an individual reaching for a cup with one of two object-directed intentions: to drink OR to clear the table. A third non-object-directed intention was also included where the observed individual placed their hand on the table next to the cup. For each intention the contextual information was varied by changing the environmental scene between (1) cups full of juice, (2) almost empty cups, and (3) half-empty cups. The findings reveal that participants perceived the cup's functional (most salient) affordance - drink - for the intention so far as the movement kinematics specified an object-directed intention (drink or clear) with a context that clearly afforded it (full and half-full cups). However, participants were also sensitive to the kinematic differences between the object-directed intentions when the context made the functional affordance seem improbable (almost empty cups).

The use of social robots, defined here as physically embodied artificial agents that humans can interact with in-situ, has become increasingly common over the past decade (Johal, 2020). Social robots have been deployed across a variety of learning tasks (affective and / or cognitive), in a variety of roles (48% as tutor, 38% as teacher, 9% as a peer or novice co-learner), and in various educational and research settings (Belpaeme et al., 2018). Compared to virtual agents or video/voice-based tutors, social robots offer promise for educational settings where it is well documented that social and physical interaction supports human learning (Belpaeme et al., 2018). In a comparative review, more positive outcomes were associated with tasks where the artificial agent was both physically embodied and co-located, compared to tele-present robots or other virtual agents (Li, 2015). Overall, social robot research in education provides evidence of positive affective and/or cognitive learning outcomes compared to other artificial agents (Belpaeme et al., 2018). For instance, social robots have been shown to improve cognitive gains in a puzzle task, compared to virtual or voice-only tutors (Leyzberg, et al., 2012), and increase task compliance, enjoyment, and performance in other tasks (Belpaeme et al., 2018). However, what it is about the physical presence of the robot that drives these performance gains remains not yet well understood. By convenience, much of the research with social robots in educational settings uses existing and readily available robots. Nao, a relatively inexpensive humanoid robot is the most popular option, used in 48% of studies reviewed by Belpaeme and colleagues (2018). Beyond this, however, the range of social robots used in educational research settings is highly varied across the robot’s form, features and functional capabilities, and few studies have directly compared the effectiveness of different types of social robots across equivalent tasks (Papakostas et al., 2021; Hortensius & Cross, 2018; Cross & Ramsey, 2021). The form and function of these off-the-shelf robots may influence performance outcomes in ways that are not yet well understood (Henschel et al., 2021). Social neuroscience research provides powerful evidence that human--robot interactions (HRI) are influenced by the beliefs, perceptions and expectations users hold about the artificial agents they engage with (e.g., Hortensius & Cross, 2018). Many different factors may influence how we perceive a robot, and what we feel or how we act in their presence (Cross & Ramsey, 2021), including what the robot looks like, and its perceived capabilities. For example, Woods (2006), in exploring the design space for robot development, provided evidence that it is the overall appearance, rather than just the facial features, that influence a child’s perception of the robot. Children were consistent in relating different forms in design, including shape, anthropomorphic features, and modes of movement, to differing personality characteristics. For example, aggressive robots were seen to be more machine- like, and happy / friendly robots more animal-like (Woods, 2006). In support of this, Hwang et al (2013) also found that the shape of the robot influences the emotions and personality attributed to the robot. Despite this, there is limited research into how the appearance and functions of a robot influence perception and expectations in educational tasks. It will be important to build a clear understanding of these issues if robots are to be used effectively to support educational outcomes in the future. To begin to address this, Caruana, Moffat, Blanco & Cross (in prep -https://osf.io/jdv2y/?view_only=f0e43b6b292747d587468865c2a02dd2), sought to compare the first impressions, expectations, and subsequent experiences amongst early learners during a reading task with three different commercially available social robots (Nao, MiRo, Cozmo) that varied across both form and function. A key finding from this study supports the view that the aesthetic and functional capabilities of the robot influences perceptions around intelligence and friendliness (Caruana, Moffat et al., in prep). Nonetheless, a limitation of this recent study remains the pragmatic necessity to use available robots, which constrains our ability to explore what children want and expect from learning-assistant robots before they set foot in the lab (or classroom). Why reading? Different educational tasks (and learner needs) influence what is required and expected from a social robot from the perspective of the user. Exploring form and function within a specific task keeps the task variable constant. Reading is a universal learning task – all children need to learn to read – and require frequent practice and feedback on their progress in order to do so. It is also one that can be difficult for some children, who require more intensive teaching, involvement, or other interventions. Approximately 16% of children struggle to learn to read for a variety of reasons (Shaywitz et al., 1992), and difficulty with learning to read is associated with various internalising conditions, including anxiety, depression (Francis et al.,2021) and reduced self-concept (Mcarthur et al, 2020). Whilst this can be addressed with combined and targeted reading instruction and psychological interventions (e.g., cognitive behavioural therapy) (Francis et al., 2021), tools which support children’s self-engagement in reading outside one-on-one intervention settings are also needed. Emerging evidence from reading-assistant dog programs suggest that the presence of a non-human and non-judgemental social agent can support reading outcomes in children (see Hall, Gee & Mills, 2018 for a systematic review). Along these lines, social robots may offer a more practical, accessible, and flexible alternative to dogs, as they have the capacity to offer a non-judgmental social presence, but do not require the same sort of care and expert handling as a live animal does and can potentially offer more forms of relevant feedback (Caruana, Moffat et al., in prep). A design-based approach Ample research exists exploring the necessary computational elements that create a “social” robot – such as responding appropriately to cues, understanding social rules, maintaining eye contact as appropriate, and animacy in voice intonations and gestures (Breazeal, 2003). However, setting aside the complex task of programming social robots to be appropriately responsive to the cues and adaptive to individual needs of the moment, a simple question has yet to be asked of the children who are expected to engage with these robots, without biasing their perceptions: What is it young children want or expect in a reading buddy robot? Current Study. A child’s perspective on what is desirable in a robot-assistant can and does differ from that of an adult (Lin et al. 2014). To this end, our current project sets out to explore which features matter to children in a robot designed to accompany them while they learn to read, free from the bias that may come from being presented with a physical robot in the experiment. Some work has been done to understand the design space for children’s robots previously. Woods (2006) found evidence that children prefer cartoon or toy-like characters with human characteristics, over more realistic robots, for example, with robots that are too human-like creating a sense of discomfort (the so called ‘Zombie’ effect, (Woods, 2006)). Lin and colleagues (2013) further explored expectations and design features for a service robot for children’s libraries. They also noted a preference for a more cartoon-oriented design, function-orientated movement, and visual feedback amongst children (Lin et al., 2013). However, this is the first time, as far as we can tell, young learner readers will be consulted as to their expectations and desires for a reading buddy robot. Unlike previous studies which have used pre-determined robot stimuli to probe the design space of robots (Wood, 2006), our aim is to use qualitative research via a “blank-sheet” discussion with young children of learning to read age (5-9 years) to build a better understanding of what they would like in a reading buddy robot. In a recent paper, Cross & Ramsey (2021) proposed a feature-based, or dimensional, framework to explore how robot variability influences human-robot interaction. Whilst not prescriptive in the dimensions proposed, they included size, human-like form or motion, functional repertoire, automated features, socialness, intelligence, and pre-existing expectations. Our open discussion approach will aid us in identifying and exploring which dimensions are spontaneously relevant from the child’s perspective, in a reading-buddy robot. Beyond this, we intend to probe for specific dimensions identified within Cross & Ramsey’s (2021) proposed framework, including physical form (shape, anthropomorphism, size, colour, material), functional capabilities (mobility, communication, feedback, actions), and character attributes (such as intelligence, friendliness, socialness).

Is there something unique about observing dynamic social interactions between humans, robots, and objects?

Access to the data, figures, and experiment builder used in the HeartBots experiment (Riddoch & Cross, 2020).

This study compares the influence of a synchronised audience on inhibition to that of a simple non-synchronised audience. We use the mirror game as a motor synchrony intervention and a Go/NoGo task to measure inhibition.

Background: To fulfil the need for a convenient and effective long-acting formulation to de-liver drugs to HIV/AIDS patients over a sustained period, we have developed an injectable in situ forming peptide hydrogel implant for the delivery of HIV/AIDS drugs for ≥28 days. Our formulation is an enzyme-responsive self-assembling low molecular weight D or L-peptide hydrogelator, namely phosphorylated (naphthalene-2-ly)-acetyl-diphenylalanine-lysine-tyrosine-OH (NapFFKY[p]-OH), to which zidovudine is conjugated covalently via an ester linkage. This forms a powder that can be readily dissolved in aqueous buffer to create an in-jectable solution. Aims/Objectives: To understand the macroscopic properties of gels and the relationship between their underyling structure and the diffusion kinetics of the HIV/AIDS drug zidovudine using primarily small angle neutron scattering (SANS), rheology and drug release assays. Methods: SANS [D11 ILL,], oscillatory rheology (loss/storage moduli, critical strain, viscosity, time sweeps), scanning electron microscopy, circular dichroism, in vitro (phosphate buffered saline) and in vivo (Sprague Dawley rats, subcutaneous injection) drug release for ≥28 days. Findings: Rheology showed peptides demonstrated enzyme-instructed self-assembly, form-ing hydrogels (~1KPa – 10KPa) within minutes. Drug attached peptide gels were less stiff than gels formed from parent peptide with no drug. SANS showed gels form fibres of narrow ra-dius (~2nm) and large length, a common property of entangled gel fibers]. Data closely fitted the flexible cylinder elliptical model and the composition of the gel fibres were similar at low Q. Differences in gel stiffness when zidovudine is attached to the peptide was attribut-ed to entanglement of fibres, rather than the composition of fibres themselves or their sec-ondary structures (primarily beta-sheets). The D-peptide variant Napffk(zidovudine)Y[p]G-OH demonstrated zero-order drug release kinetics over 28 days. Clinically relevant plasma levels of zidovudine were detected in rat models for up to 35 days. Conclusions: A large component of gel strength can be controlled by external conditions and this may allow a change in formulation parameters to optimise material specifications, most notably gel strength and drug release kinetics for long-acting drug delivery.

We can watch formal dance performances in different event spaces. For example, we might watch a dance performance live on a stage, or live at home on TV, a laptop, or even a mobile phone. During the Covid-19 pandemic, opera houses streamed more dance events than ever before. People from all over the world joined these streamed performances, either in cinemas with large screens, or from the comfort of their own homes. Sometimes, these live performances were recorded, which provided people with the ability to access the performances at a later stage for on-demand watching – again, at home or in cinemas or theatres, all over the world. An intriguing empirical question worthy of study concerns how these different types of Event Space (in theatre, home) and Mode (live, recorded) impact our aesthetic experience of a dance performance? Does it matter whether we watch a dancer live in front of us on a stage, or whether we see the same dancer while sitting at home, watching a screen? And, does it matter whether the dance is recorded, or live? Empirical concert studies certainly underscore the importance of the ‘frame’ in which music is received as a component of the aesthetic experience of music (Wald-Fuhrmann et al., 2021). Some previous research with music suggests that aesthetic enjoyment of music is higher when the presentation mode is live (Swarbrick et al., 2018, 2021), while other studies suggest that the enjoyment is comparable for live and recorded concerts, but dependent on interindividual differences in musical expectations (Belfi et al., 2021). With regards to dance, to our knowledge, principled assessments of this research question remain limited. With the rise of AI-assisted software, and avatar-powered dance media during the Covid-19 pandemic, an additional research question about the Type of Dancer has moved centre-stage: In the realm of dance performance, how does the physical form a dancer takes (i.e., human vs. artificial agent/avatar) shape our perceptions and enjoyment of a dance performance? Is watching a performance of an avatar as exciting and aesthetically valued as watching a human dancer? Some previous work suggests that movement smoothness is a key factor that predicts human observers’ aesthetic experience of a humanoid robot’s movements (Miura et al., 2010; Cross et al., 2016). Finally, and most importantly, in Western dance performance work, ‘Expressivity’ is a very important factor. During professional dance education, dance students are taught to dance with emotional expressivity, and not ‘just’ to perform movements technically correctly. For experts, expressivity is what differentiates dance from other impressive bodily practices, like, for instance, gymnastics or martial arts. Dancers can perform a dance piece either simply expressing neutrality (no expressivity), or expressing some internal state or emotion (with expressivity). With two dance pieces otherwise being equal, but performed expressively by the dancer, while the other is ‘merely’ technically accurate in its performance, which piece will audiences prefer? Previous work suggests that, when shown two stimuli with the same dance movement sequence, of which one has been danced with an expressive intention by the dancer, then dance-naïve observers rate the expressive version of the sequence as more expressive, and find it more beautiful and more likable (Christensen et al., 2018, 2021). By obtaining objective movement kinematic features like acceleration and speed it has been reported that the expressive intention is translated into a more varied velocity profile and this predicts human observers’ aesthetic experience of the dance movements (Orlandi, Orgs, & Cross, 2021). To provide first answers to the above questions, we will collect audience responses to purpose-choregraphed 1-minute dance pieces during a hybrid public engagement event to take place in the ARTlab of the Max Planck Institute for Empirical Aesthetics in Frankfurt/M, Germany. The audience will watch these dance pieces via different Modes (live, recorded), and the audience will be divided into different Event Spaces (at the institute Artlab that is like a theatre, or, at home). In addition, one of the pieces will be danced by a different Type of Dancer, a humanoid avatar, allowing us to compare how different types of dancers are received (human, avatar). In addition, the dancer will dance one 1-minute dance piece six times. Half of the times, he will dance the piece with expressivity, and the other half without. Hence, the audience will rate a total of six 1-minute dance pieces (3 with expressivity, 3 without). The question is: Does dance ‘Expressivity’ ‘reach’ the audience, no matter the type of Event Space (at theatre, at home), the Format (live, recorded), and the Type of Dancer (human, avatar)? We predict that the ‘Expressivity’ factor may cross through the different types of Event Space, Format, and Type of Dancer explained above; i.e., that the expressive versions of the dance pieces will be liked more, regardless of Event Space and presentation Format. The ratings that we will ask the audience members to provide will be on visual analogical scales including “how expressive was this movement?”, “how much did you like the movement?”, “how interesting did you find the movement?”, “how much would you like to see the movement again?” and “how boring did you find the movement?” all scales will be from 0 (not at all to 100 (very much).

Describing interactions between objects often requires the speaker to choose a perspective, and mention one object first and the other second. Speakers have a tendency to make the most animate object in a scene the subject of a sentence, and to mention this object first (Prat-Sala & Branigan, 2000). For example, it is more common to say the man ran away from the dog rather than the dog chased the man, therefore taking the perspective of the man. This bias is considered to occur because of the way speakers perceive animacy: on a hierarchy, with humans at the top and concepts or inanimate objects near the bottom (Harris, 1978; Aissen, 2003). An animacy hierarchy is thought to be the product of conceptual accessibility; the more conceptually accessible something is, the easier it is to retrieve and the more likely it is to be uttered earlier on in a sentence. Animate entities are considered to be more conceptually accessible than less animate entities (McDonald, Bock & Kelly, 1993; Bock & Warren 1985), and this results in an animacy bias whereby speakers utter them earlier on in their sentences. Participants will view scenes of two figures interacting, who will differ in animacy (one human and one robot). Participants will describe these scenes using so-called symmetrical predicates (Gleitman, Gleitman, Miller, & Ostrin, 1996). These are verbs that imply a symmetry of action between the two figures (e.g. arguing, kissing, playing), but when used in a sentence they require the speaker to make a decision as to which figure should be mentioned first, which may be potentially conditioned on a factor such as animacy. Participants will form transitive sentences (e.g. “Kate is arguing with Botz”) to describe the events, which they will type into a text box, and we will measure if they mention the human or the robot first. The events depicted in this experiment constitute three “valences”; intimate (e.g. kissing), negative (e.g. arguing) and casual (e.g. playing). We will investigate if the animacy bias is stronger on trials depicting emotional (intimate/negative) events. In terms of animacy, robots constitute a special control category for human stimuli. Contemporary humanoid robots are capable of self-propelled motion, speech (in some cases), and can exhibit what appears to be intentional behaviour (for example through eye movements or gestures), hence, they provide some cues to animacy to human observers (Bartneck, Kulić, Croft & Zoghbi, 2009). However, as previous studies on the perception of mind have shown, humans would be hesitant to call any robot “alive” (Gray, Gray & Wegner, 2007). When rated on their capacity of mind, robots rank below human adults and infants, and receive somewhat comparable ratings to pets, with little means to sense or feel the world, but with some capacity of agency. A study that asked participants to rate robots’ ability to plan and act (their agency) and their ability to sense and feel (their experience of the world) confirmed that participants assigned high rankings to pictures of human faces for both agency and experience, whereas robots where ascribed some agency but no experience, and objects ranking lowest on both criteria (Henschel, Bargel, & Cross, 2020). We therefore propose that robots should rank lower than humans on the animacy hierarchy, while controlling for humanlike features (i.e. having a comparable body shape and similar set of facial features). Our previous research suggests that closeness to the speaker influences the conceptual accessibility of an entity. When describing an interaction between two people, speakers are more likely to take the perspective of the figure most similar to themselves in social group (e.g. the same race or gender as themselves). These studies are in progress, but the pre-registrations can be found here (https://osf.io/bfhrn) and here (https://osf.io/t928m). The results of these studies suggest that speakers often demonstrate a ‘Like Me’ bias to name the figure from their own social group first and the figure from the out-group second. Thus, in addition to our main analysis of the animacy bias, we will further consider the ‘Like Me’ bias in exploratory analyses by varying the social group of the human in these pairs (a black woman, a black man, a white woman and a white man) and conducting this experiment with participants from these four demographic groups. In doing so, we will measure if participants demonstrate an animacy bias by routinely taking the perspective of the human figure, and if this bias is stronger on trials where the human figure is in the participant’s ingroup than their outgroup. Prior to this main experiment, we conducted a norming study with a separate group of 80 participants (of the same four demographic groups as in this experiment) to validate our novel set of robot stimuli. We investigated how much mind is ascribed to the four robotic agents to be used in the main study, and found that two of our robots (‘Pepper’ and ‘NAO’) were rated as having higher (and similar levels of) perceived mind than the other two (‘Gresh’ and ‘Princess Firecracker’ – see Files for portraits of all four robots used in this experiment). According to this norming study, people perceive the humanoid robots (Pepper and NAO) as having greater agency and experience (Henschel et al., 2020) than the Lego robots (Gresh and Princess Firecracker). We expect the animacy bias to favour the human figure even more on trials where the human is paired with an object with low perceived mind. We will test this hypothesis by entering the continuous variables of perceived robot agency and perceived robot experience (ratings provided in the norming study) into our model as predictors of mentioning the human figure first. As gender and race are of importance in the current experiment, we also included rating items about the robots’ perceived race, sex and gender. In order to establish gender-neutral names for the robots that are not too distinct from the set of human names, participants further rated a set of six names on perceived gender and sex. The four robot names used in this study were those that were rated as being most gender-neutral on 9-point Likert scales (very female – very male, very feminine – very masculine). Descriptions of how we intend to use these items of perceived robot race and combined gender/sex from the norming study are explained in the Exploratory Analyses section.

This project aims to distinguishing between two theoretically important pathways that provides insight into why men’s endorsement of hostile sexism is associated with aggressive behavior towards female partners. We expect men's hostile sexism to predict both (a) lower perceived relationship power, and (b) greater desire for power. Both of these power variables should independently predict greater reports of aggression towards female partners.

A strong motivator for engaging in prosocial behaviour is empathy (Decety et al., 2016). Empathy is a multidimensional construct build from interrelated cognitive and affective components that lets individuals perceive and react to the emotional state of others (Davis, 1983, Preston & de Waal, 2002). In this study, we define empathy as an affective response caused by, and congruent with, the inferred emotional state of others (Eisenberg et al, 1991). An important characteristic of empathy is intergroup empathy bias. This bias describes the fact that humans feel more empathy for members of their ingroup as opposed to members of an outgroup (Cikara et al., 2011). On some occasions, people even experience pleasure when an outgroup member encounters misfortune, a response called schadenfreude (Vanman, 2016). Feelings of schadenfreude are known to be modulated by interpersonal factors, such as an individual’s responsibility for or the perceived deservingness of their misfortune (van Dijk et al., 2005), malicious envy (van der Ven et al., 2015) and ingroup-identification when the outgroup is a rival team (Hoogland et al., 2015). It is important to understand when and why these biases occur, as they are associated with motivating people to take part in collective violence (Cikara, 2015). Categorization of individuals in different groups or teams based on arbitrary criteria has already been shown to be successful in inducing intergroup empathy bias (Montalan et al., 2012), and intergroup schadenfreude bias (Cikara et al., 2014). However, participants in these studies would only rate their feelings of empathy and schadenfreude in relation to scenarios which were not directly relevant to the task or the source of the competition. In the current study, we will directly assess whether people display an intergroup empathy and schadenfreude bias when rating their feelings during a competitive game when they are arbitrarily assigned to competing teams.

Is there a relationship between anthropomorphism and Theory-of-Mind? So far, data provide mixed results. Partial support comes from neuroimaging studies looking at an association between situational (e.g., task-based) and dispositional measures of anthropomorphism and Theory-of-Mind network activation and anatomy (Chaminade et al., 2007; Cullen et al., 2014; Waytz, Morewedge, et al., 2010). Overall, activity patterns overlap when people engage in anthropomorphism and Theory-of-Mind (Hortensius & Cross, 2018). However, we found no consistent relationship between dispositional anthropomorphism and activity within and across the Theory-of-Mind network during the observation of a social animation film (Experiment 1). While a quadratic relationship described the data better than a linear relationship, the data were in favour of the null hypothesis. This is in line with a recent behavioural study that found no relationship between direct and indirect measures of situational and dispositional anthropomorphism and inter-individual differences in theory-of-mind in adults and children (Tahiroglu & Taylor, 2019). Research question: In this follow-up experiment, we explore the relationship between Theory-of-Mind and situational and dispositional anthropomorphism in more detail. Specifically, we ask if there is a linear or quadratic relationship between situational and dispositional anthropomorphism and behavioural measures of Theory-of-Mind.

This project is aimed at investigating the extent to which social robots’ cognitive architectures can elicit emotionally rich expressions and information from people to identify their needs and emotional states over time. By conducting an online experiment this research will examine people’s expressions to social robots, as well as how people disclose their needs and emotions to social robots via video chat over time. By measuring peoples’ disclosures and emotional states through different channels of communication and behaviour, emotional patterns of peoples’ experiences regarding different topics (i.e., coping with COVID-19 and general everyday experiences) will be further evaluated. Finally, the study results will provide evidence for peoples’ expressions, perceptions, and reactions to a social robot in reoccurring interactions over time, and will reduce the current knowledge gap regarding novel effects in human-robot interaction (HRI) research. The research findings will be employed for developing speech-based health and well-being interventions using social robots. Furthermore, these will contribute to our broader understanding of how individuals share information and disclose their emotions to social robots, and how social robots can be used as tools to support people's emotional health. This study investigates the extent to which social robots can elicit and support socially meaningful conversation with people over time. A between-subjects repeated measures online experiment is planned with participants completing 10 online video chats with a social robot across five weeks. By measuring the quantity and quality of peoples’ disclosures, expressions, and perceptions, this study aims to provide a nuanced overview of the impact of affective interactions with social robots over time.

New natural language generation (NLG) algorithms, like GPT-2, are able to generate human-like text across diverse domains. For instance, artificial intelligence (AI) algorithms can generate news articles based on standardised data, or creative texts such as poems (van Dalen, 2012; Koebis & Mossink, 2021). In the current study, we wanted to investigate 1) whether archives generated by GPT-2 can be accurately identified as AI-generated i.e. can people distinguish between "fake" and "real" archives; 2) do people value real archives more than AI-generated archives both when they are aware of the distinction, and when they are not; 3) are people more biased to "destroy" fake archives compared to true archives, and 4) how do people's attitudes toward AI modulate their responses to how much they value AI-generated archives?

In the present study, we aim to examine the link between aesthetic appreciation and stimulus and knowledge cues to human animacy. Through a series of experiments, this study explores the relationship between stimulus (agent) and knowledge (source of movement) cues to human animacy, and their link to aesthetic appreciation. Previous evidence suggests higher ratings of smoothness and liking when movement was believed to originate from human movement compared to computer animation, independent of weather the agent's form was a human or robot (Cross et al., 2016). Moreover, human-generated choreographies are preferred to choreographies that are human-generated (Darda & Cross, 2021). In the current experiment, we aim to delve more deeply into the relationship between stimulus and knowledge cues to human animacy on people’s aesthetic preferences for dance movements, and how this might interact with choreographies that are intentionally choreographed by a human (human-generated), and those that are choreographed by a computer (computer-generated).

Our aim is to understand whether dog behaviours important for the human-pet bond, translate well (are understood and perceived similarly) when demonstrated in a robotic platform.

The Economic Games with Robots 2 (ECONOBOT2) project explores the impact of robots’ emotional displays on people’s tendency to cooperate with a robot opponent in prisoner’s dilemma games. Participants played iterated prisoner’s dilemma games with a non-expressive robot (as a measure of cooperative baseline), followed by an angry and a sad robot, in turn. Based on the Emotion as Social Information model, we expected participants with higher cooperative to cooperate less when a robot displayed anger, and cooperate more when the robot displayed sadness. Contrarily, according to this model, participants with lower cooperative predispositions should cooperate more with an angry robot and less with a sad robot. The results of 60 participants failed to support the predictions. Only participants’ cooperative predispositions significantly predicted their cooperative tendencies during gameplay. Participants who cooperated more in the baseline measure also cooperated more with the robots displaying sadness and anger. In exploratory analyses, we revealed the influence of individual differences in emotion perception on people’s cooperative decisions. Participants who accurately recognised the robots’ sad and angry displays tended to cooperate less with them overall. The study highlights the impact of personal factors in human–robot cooperation. and how these factors might surpass the influence of bottom-up emotional displays by robots.

Describing interactions between objects often requires the speaker to choose a perspective and mention one object first and the other second. Speakers have a tendency to make the most animate object in a scene the subject of a sentence, and to mention this object first (Prat-Sala & Branigan, 2000). For example, it is more common to say the man ran away from the dog rather than the dog chased the man, therefore taking the perspective of the man. This animacy bias is considered to occur because of the way speakers perceive animacy: on a hierarchy, with humans at the top and concepts or inanimate objects below (Harris, 1978; Aissen, 2003). The current experiment follows on from the work detailed in a previous pre-registration, Study 1, which can be found here: https://osf.io/a5fby. Speakers viewed scenes of humans and robot interacting, and described these interactions using transitive sentences (e.g. “Fred is dancing with Botz”). We found speakers would reliably mention the human first and the robot second, with statistical modelling showing a strong main effect of relative animacy. The following experiment attempts to explore the ways in which animacy is affected by perceived mind. In doing so, we question the extent to which mind attribution determines animacy and designates an object’s place on the “animacy hierarchy”. Researchers in psycholinguistics have offered a few explanations for the animacy bias phenomenon. According to the theory of conceptual accessibility, the more significant an object is in human thought, the more conceptually accessible it is considered to be, and therefore the greater the likelihood of a speaker mentioning that object earlier on in speech (McDonald, Bock & Kelly, 1993; Bock & Warren 1985). Animate entities are also considered to be more easily or rapidly lexically retrieved, meaning the speed with which we retrieve labels for animate objects from memory is faster than that for inanimate objects and concepts, thus increasing the likelihood of these labels appearing earlier in speech (Branigan, Pickering & Tanaka, 2008). Regarding the specific animacy differences in our own study (human vs robot), we might look to neuroimaging studies to understand how human and robot figures are processed differently neurologically. When Wang and Quadflieg (2015) compared participants’ responses to interactions between either two humans or between a human and a robot, they identified that different regions of a neural network involved in formulating impressions relating to beliefs, desires, motives, feelings or intentions were activated depending on the interaction type (human/human or human/robot). The results from this study suggest that social cognition relating to impression formation may differ when we are witnessing interactions between two figures of the same relative animacy, and two figures who fall on different positions on the animacy hierarchy. Crucially, humans and robots also differ in agency and experience. While “agency” designates the thematic role of an object doing something to another object in linguistics (Dowty (1979, 1991), in the study of mind perception, agency has been defined as an object’s ability to plan and act (Gray, Gray, & Wegner, 2007). Experience, a second important feature of perceived mind, refers to an object’s ability to sense and feel (Gray et al., 2007). Evidence for humans and robots differing in perceived mind comes from Henschel, Bargel and Cross (2020), who asked participants to rate humans, robots and objects for capacities of mind. They found that participants rated humans as high for both features and ascribed robots with some agency but no experience (while objects had the lowest ratings for both agency and experience). The current study therefore considers two definitions of “agency” - firstly who, linguistically, is designated the sentence agent, and secondly who appears to the speaker as being the most agentive figure. An important cognitive feature of perceiving social interactions is intention, which relates to both the representation of a figure’s covert mental state and that which arises from perceiving an action they are carrying out (Gobbini et al., 2007). In addition to Gray, Gray and Wegner’s definition of agency, Dowty (1991) also proposes features of intentionality as key facets of agenthood, such as causation (causing another object to experience an event or change in state) and volition (choosing to be involved the event or state change). In general, we understand that humans have the capacity to plan and act – an ability that is certainly more questionable in robots, and we are also aware that humans are able to execute actions and bring about change. Again, this is an ability many people are likely to be uncertain about for robots. Thus, we could argue that people might perceive intention differently between human and robot figures executing actions. Social cognition is dependent on a person’s ability or willingness to consider the mind of another person. Dehumanised perception results when someone fails to engage typical social cognition and automatically consider the mind of another person (Harris & Fiske, 2011). On the other side of the spectrum is empathy – the capacity to understand and share another’s feelings. Empathy is an important feature of mind attribution while processing social interactions, and is something we are capable of demonstrating for robots as well as humans, given the right circumstances (Küster & Swiderska, 2020, Rosenthal-von der Pütten et al., 2013). The current study explores the role that mind attribution has on perceptions of animacy. We are investigating if it is possible to use narrative to alter how people perceive the mind of a non-human object, such as a robot, to the extent that they reverse the normal rules of the animacy bias (human mentioned first, robot mentioned second, as seen in Study 1). We will also use this narrative to reduce the perceived mind of human figures. In doing so, we aim to encourage speakers to view events from the perspective of the usually less-animate robot, therefore making this figure the sentence agent over the depicted human and mentioning it first in their descriptions of events. We will use the same paradigm as in Study 1, but we will manipulate features of mind, including agency, experience, empathy and dehumanisation, by prefacing the critical trials with a short story. In a robot-oriented condition, this narrative will frame the robot figures as highly animate and the human figures as dehumanised clones grown by the robot figures. A second and control condition, human-oriented, will depict the human figures as animate agents and the robots as clones the humans have grown. We will also use this control condition to measure the extent to which linguistic accessibility influenced the human bias detected in Study 1. There, we used common Anglophone names for our human figures (e.g. Fred, Kate) and fictional names for our robot figures (e.g. Cybe, Mech). In the current study, both human and robot figures will have fictional names, which would allow us to confirm that the results of Study 1 could not be explained simply in terms of name familiarity. Participants will view scenes of two figures interacting, who will differ in animacy (one robot and one human) and will describe these scenes using so-called symmetrical predicates (Gleitman, Gleitman, Miller, & Ostrin, 1996). These are verbs that imply a symmetry of action between the two figures (e.g. arguing, kissing, playing), but when used in a sentence they require the speaker to make a decision as to which figure should be mentioned first. Participants will form transitive sentences (e.g. “botz is arguing with kote”) to describe the events, which they will type into a text box, and we will measure if they mention the robot or the human first.

This study investigated human-robot cooperation in the context of prisoner’s dilemma (PD) games and the extent to which people’s willingness to cooperate with a robot varies according to incentives provided by the game context. We manipulated the payoff matrices of human-robot PD games and predicted that people would cooperate more often in the situation where cooperating with the robot was a relatively more rewarding option. Our results showed that, in the early rounds of the game, participants made significantly more cooperative decisions when the game structure providing more incentives for cooperation. However, subsequent game decisions were dominantly driven by other two factors. The first was the robot’s previous game choices, where participants played a tit-for-tat strategy against the robot even though its decisions were random. Second, presentation of real-time game scores significantly impacted people’s cooperative tendencies, which were further shaped by participant performance relative to the robot. Participants only showed prosocial tendencies toward the robot when they had achieved high scores. In conclusion, people appear to display a strong reciprocal tendency toward social robots in economic games, and other relevant factors shaping people’s behavior in PD games played against robots include incentive structure, presentation of game scores, and people’s predisposition to anthropomorphism.

The present study uses a story-stem method to gain qualitative insights into the narratives which people associate with human-robot encounters. Further, we propose to investigate whether and how these narratives differ between participants from the UK and Japan. We are particularly interested in how far participants ascribe agency and human-like characteristics (anthropomorphism) to a fictional robot, and how the human-robot relationship is portrayed in the narratives, including the proxemic behaviour of the characters. The study supplements an ongoing quantitative study (registered on OSF, embargoed until 28th April, 2022, Rooksby et al., 2020), which investigates how UK and Japanese residents compare on their proxemic preferences for an approaching robot.

This in-situ experiment employs a repeated measures design with a within-participant condition to investigate behaviours of human participants while approaching a humanoid robot (NAO, Softbank Robotics) in the context of simple verbal exchanges. Our specific focus is on the range of distances that human participants choose when moving to and from the robot, as well as how they do so, in terms of the angle and posture during the approach (i.e., manners in which to approach a robot). The data will be compared across a group of Japanese and British nationals to investigate the role of cultural group membership to identify and explore any differences between the groups in terms of the distance and manner. Further analysis will explore any effect of demographic background and other personality/trait information of the participants. The UK and Japan are generally considered nations associated with ‘non-contact’ cultures in terms of proximity preferences during social interaction; however, other differences are generally accepted between them, such as sitting on the floor being more common in Japan or aversion of direct gaze known as an aspect in the Japanese culture. Our main goal in the study concerns how the two groups may be distinguishable when a robot sits in a posture familiar in Japan or around a robotic behaviour that is aimed to instigate eye contact. A secondary goal is to explore how participants’ existing experience, such as exposure to robots, and other demographic information, may account for any differences observed between the groups.

An exploration into the attachment-promoting and engagement-sustaining features of children's companions, nonexclusive to any specific pet, animal, or toy, within the context of family units, culture, and communities. Furthermore, this study aims to identify commonalities in the preferences of children across sociodemographic groups.

Robots are becoming more prominent in daily life. Aside from being present in industrial environments, they are finding their ways in our homes and workplaces. One aim of roboticists is for these machines to not only serve functional roles, but also interact socially with humans. A more communicative robot is seen as more favourable than an efficient robot in some situations (Hamacher et al., 2016). Ideal human-robot interaction requires not only the robot’s ability to interact with humans but also the human’s willingness to interact socially with this robot. A strong motivator for engaging in such kind of prosocial behaviour is empathy (Decety et al., 2016). Empathy is a multidimensional construct build from interrelated cognitive and affective components that lets us perceive and react to the emotional state of others (Davis, 1983; Preston & de Waal, 2002). In this study, we further define empathy as an affective response caused by, and congruent with, the inferred emotional state of others (Eisenberg et al., 1991). An important characteristic of empathy is that we are more inclined to feel more empathy towards individuals with whom we share commonalities. This propensity is referred to as the intergroup empathy bias. People can feel more empathy for members of their ingroup as opposed to members of an outgroup, and on some occasions, people can even experience pleasure when an outgroup member encounters misfortune, a response called schadenfreude (Vanman, 2016). Feelings of schadenfreude are known to be modulated by interpersonal factors, such as an individual’s responsibility for, or the perceived deservingness of, their misfortune (Dijk et al., 2005), malicious envy (van de Ven et al., 2015), and ingroup-identification when the outgroup is a rival team (Hoogland et al., 2015). It is important to understand when and why these biases occur, as they are associated with motivating people to take part in collective violence (Cikara et al., 2011). Robots regarded and treated as social beings might also be subjected to this kind of social categorisation (Vanman & Kappas, 2019). However, it is not yet known whether and when people might classify robots as members of their ingroup, and thus feel more empathy toward them in certain situations, and also whether it might be the case that people can feel schadenfreude when outgroup robots experience misfortune. Previous studies had tried to make robots more human-like in order to facilitate participants’ perceptions that these robots belong to their ingroup (Eyssel & Kuchenbrandt, 2011). Another way to promote robots as being seen to belong to a person’s ingroup could be by assigning humans and robots to the same team (Fraune et al., 2017). Categorization of individuals into different groups or teams based on arbitrary criteria has already been shown to be successful in inducing intergroup empathy bias (Montalan et al., 2012), and intergroup schadenfreude bias (Cikara et al., 2011). So far, this has not yet been investigated in mixed human-robot teams. Across two studies, we will investigate whether humans show an intergroup empathy bias towards robots when humans and robots form a team (I). We will also examine whether we can generalise our findings across different robots (II), an often-unaddressed issue in the field of social robots (Cross et al., 2019; Hortensius et al., 2018). In the first study, we will use the Cozmo robot (Figure 1), a small entertainment robot with a mechanical design, while in the second we will use the NAO robot (Figure 2), a humanoid robot often used in HRI research.

This is a follow up experiment for the Something in the Way You Move Project. We aimed to replicate the results from Experiment 1 with a higher level of experimental control. RQ3: How do different types of dancers (avatar, human dancer) impact audiences’ aesthetic experience of a dance performance? RQ4: Two dance pieces being equal, but one is being danced with expressivity by the dancer, while the other is ‘merely’ technically perfect (i.e., without any intended expressivity), which piece will audiences prefer (i.e., rate higher in aesthetic enjoyment)? RQ5 (exploratory): What are the kinematic profiles of dance sequences that are intended to be expressive as compared to those that are not, in terms of speed, acceleration and velocity? RQ6 (exploratory): Does dance expertise modulate aesthetic judgment as a function of the above parameters? This research question is exploratory because we will not seek to control the variable dance expertise when recruiting participants, but include questions about dance experience (yes/no, ‘years of experience’, and whether they are ‘professional dancers’; yes/no). RQ7 (exploratory): Do interindividual differences like dance expertise, personality, aesthetic and emotional sensitivity have an effect on aesthetic enjoyment of dance movements as a function of our factors type of dancer and intended expressivity? Explore different rating scales To replicate the results from Experiment 1 with a higher level of experimental control, the rating scales of Experiment 2 again included liking, interest, want to see again, and boredom. Furthermore, a rating scale for ‘beauty’ was added because: 1) previous research that has shown that beauty and liking are separate dimensions of aesthetic enjoyment (J.F. Christensen et al., 2018), and we aimed to confirm or refute this finding. And 2) we aimed to explore which of the two prototypical affective responses (liking or beauty) would be more positively related to the epistemic and utilitarian emotions that may propel audiences to wish to come back to watch a dance performance (interest and want to see again), and negatively related to the emotion of aesthetic displeasure (boredom) that may cause the opposite effect. This latter aim was of course motivated by general problem of the dance performance world post-Covid-19, that we outlined in the introduction. Interindividual differences Demographics, dance experience and additional interindividual difference measures were obtained, including personality (Big Five Extra Short version; BFI-XS; (Rammstedt, Danner, Soto, & John, 2018), and a measure of aesthetic sensitivity (the Aesthetic Responsiveness Assessment; AReA; (Schlotz et al., 2020)). In keeping with the evidence reviewed earlier, we aimed to explore the interactions between the factors type of dancer and intended expressivity and personality, and aesthetic sensitivity. Data for Experiment 2 was obtained via Prolific©, a recognized platform for behavioural research (Peer, Rothschild, Gordon, Evernden, & Damer, 2021; Sauter, Draschkow, & Mack, 2020).

Preregistration plan for a study investigating couples where men experience financial strain, examining the associations between endorsing benevolent sexism with men's personal well-being, as well as with women's relationship well-being

This registration contains the materials and relevant data and code for Sample C in the paper entitled, "Feeling Appreciated Buffers against the Negative Effects of Unequal Division of Household Labor on Relationship Satisfaction."

This project will investigate the potential for emotional exchange in chatbot-human-interaction (CHI), informing their appropriate role in mental health care. More specifically, we are interested in the extent to which the perception of a disembodied conversational agent’s origin will affect one’s emotional disclosure and their perceptions of the agent, the interaction, and their emotional well-being.

While a growing body of literature has begun to examine proxemics in light of human–robot interactions, it is unclear how insights gained from human–human or human–robot interaction (HRI) apply during human–drone interactions (HDI). Understanding why and how people locate themselves around drones is thus critical to ensure drones are socially acceptable. In this paper, we present a proxemic user study (N=45) in virtual reality focusing on 1) the impact of the drone’s height and 2) the type of cover story used to introduce the drone (framing) on participants’ proxemic preferences. We found that the flying height has a statistically significant effect on the preferred interpersonal distance, whereas no evidence was found related to how the drone was framed. While results also highlight the value of using Virtual Reality for HDI experiments, further research must be carried out to investigate how these findings translate from the virtual to the real world.

One limitation of social robots has been the ability of the models they operate on to infer meaningful social information about people's subjective perceptions, specifically from non-invasive behavioral cues. Accordingly, our paper aims to demonstrate how different deep learning architectures trained on data from human-robot, human-human, and human-agent interactions can help artificial agents to extract meaning, in terms of people's subjective perceptions, in speech-based interactions. Here we focus on identifying people's perceptions of their subjective self-disclosure (i.e., to what extent one perceives to be sharing personal information with an agent). We approached this problem in a data-first manner, prioritizing high quality data over complex model architectures. In this context, we aimed to examine the extent to which relatively simple deep neural networks could extract non-lexical features related to this kind of subjective self perception. We show that five standard neural network architectures and one novel architecture, which we call a Hopfield Convolutional Neural Network, are all able to extract meaningful features from speech data relating to subjective self-disclosure.

Openly available natural language generation (NLG) algorithms can generate human-like texts across multiple domains. Given the increasing potential of NLG algorithms, many ethical challenges arise such as being used as a tool for misinformation. It is necessary to understand not just how these texts are generated from an algorithmic point of view, but also how they are evaluated by a general audience. In the current study, our aim was to shed light on how people react to texts generated algorithmically, whether they are indistinguishable from original (human-generated) texts, and the value people assign these texts in a rapidly automated world. Using original text-based archives, and fake text-based archives generated by artificial intelligence (AI), findings from our pre-registered, statistically powered study (N=228) revealed that people assigned lower values to archives that were AI-generated compared to original archives. Although participants were unable to accurately distinguish between AI generated and original archives, original archives were more likely to be preserved than AI-generated archives. This bias against AI archives persisted when people were aware of the archive’s source and when they were not, as well as when they categorised the archive as AI-generated (even though it may not be). People’s judgements of value were also influenced by their attitudes toward AI. These findings provide a richer understanding of how the emergent practice of automated text content creation alters the practices of readers and writers alike and have implications for how readers’ attitudes toward AI affect the use and value of AI-based applications and creations.