Human—Robot companionship: A mixed-methods investigation

Riddoch, Katie Alexandra (2021) Human—Robot companionship: A mixed-methods investigation. PhD thesis, University of Glasgow.

Abstract

In recent years, the arts have brought robots to life in spectacular fashion. In popular fiction we have been presented with machines that can run, leap, fight, and (perhaps most impressively of all) robots which can ascend stairs with absolutely no trouble at all. Amidst these chaotic and often dystopian scenes, we are exposed to moments of humour and lightness – robots can be seen engaging in conversation, cracking jokes, and comforting someone in their time of need. In these relatively mundane moments (as we smile, laugh, and cry) the impression emerges that the robot is something special to the person depicted. Rather than simply being a household appliance, it appears to be something more: a sort of… friend.

Returning from the pages and screens of fiction to the real world, we find human society ever more fractured, and the loneliness epidemic at large. Unsurprisingly, given the engaging depictions in popular fiction, the idea of robots for companionship and social support is gaining traction and garnering increasing research attention. In care homes, robot animals can be found cooing and purring in the laps of individuals with dementia, while in schools, friendly humanoid robots may be seen teaching social skills to children with additional needs. What remains unknown, though, is the extent to which people will grow fond of such ‘social robots’ over time, and if so, whether their relationships with these machines might ever resemble (or indeed, replace) those with other humans. Is a ‘robot friend’ the stuff of science-fiction, or could it someday soon become sciencereality? In this thesis, this question is explored from a range of perspectives using a variety of methods spanning lab-based experiments, online surveys, and focus groups.

This thesis begins with an introduction to social robots, and an exploration of the background regarding the nature and importance of human social relationships. After introducing relevant theories, I highlight gaps in our understanding of human—robot companionship that I seek to explore through this thesis (Chapter 1). In the subsequent chapters, I present four empirical pieces of work, each offering a unique perspective on the subject. Specifically, in Chapter 2, I report results from a lab-based experiment in which a robot’s lights (located within its shoulders) were programmed to illuminate in a synchronous or asynchronous manner relative to a participant’s heart rate. I aimed to determine whether such a synchrony manipulation might increase prosocial behaviours and improve attitudes towards a social robot - based on prior work showing that experimentally-induced movement synchrony can improve rapport between people, and increase their liking of social robots (Hove & Risen, 2009; Lehmann et al., 2015, Mogan, Fischer & Bulbulia, 2017). Despite demonstrating no positive effect of the light manipulation, this study raises important questions regarding the complexities of defining and measuring attachment to a robot. In Chapter 3, I delve deeper into the qualitative data collected in Chapter 2 to build a more complete appreciation of the value of open questions – particularly in terms of method validation and understanding participants’ internal experiences.

After this chapter, I shift perspective from a focus on humanoid robots (and manipulations based on human social behaviours), to human relationships with non-human companion animals. This shift was motivated by my desire to explore how non-human agents form deep and enduring social bonds with humans – as opposed to basing the thesis on human interpersonal relationships alone.

Due to the success of dogs as companions, I conducted a study in which dog owners were asked to identify behaviours that they perceived as important to the bond with their dog (Chapter 4). Seven key themes emerged from this research, indicating the importance of attunement, communication, consistency and predictability, physical affection, positivity and enthusiasm, proximity, and shared activities. In the following chapter, I implement a selection of ‘desirable’ dog behaviours within an animal-inspired robot (Chapter 5). By showing the behaviours to members of the general public, and conducting focus groups, I gained deeper insights into the polarising nature of robot animals – not only in terms of how their behaviours are perceived, but also in terms of the roles people think robots should (and should not) hold. In addition to these themes, this final empirical chapter discusses insights regarding the high expectations people place upon robots, as well as public concerns around overdependence on robots, and privacy.

By releasing these chapters to the HRI community (through publications or preprints) we sparked conversations within the HRI community – not only about the ethics of robot abuse studies, but also the potential value of qualitative approaches within the field. Our team was commended for publishing qualitative research, in a field heavily dominated by quantitative methods, and we have since been working to continue the conversations around the value of qualitative approaches. Specifically, we hosted the “Enriching HRI Research with Qualitative Methods” workshop at the International Journal on Social Robotics (2020) and launched a “Qualitative Research in HRI/HCI Discussion Group” online - allowing HRI researchers to discuss their work, and share relevant resources (e.g., events and publications).

This thesis concludes by detailing work to be done moving forwards, to enhance our understanding of human—robot social relationships, and a broader discussion of our possible future with social robots (Chapter 6). Pulling from various disciplines (including psychology, cognitive science, human—robot interaction (HRI) Studies, robot ethics, and philosophy), this section concludes with consideration of potential consequences of companion technologies – not only for the individual, but perhaps for society as a whole, as we continue to grapple with questions concerning how much of science fiction we wish to welcome into our daily lives.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Colleges/Schools: College of Science and Engineering > School of Psychology
• Supervisor's Name: Cross, Professor Emily
• Date of Award: 2021
• Depositing User: Theses Team
• Unique ID: glathesis:2021-82858
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 10 May 2022 09:58
• Last Modified: 10 May 2022 09:59
• Thesis DOI: 10.5525/gla.thesis.82858
• URI: https://theses.gla.ac.uk/id/eprint/82858

Related URLs

• Article DOI