Older adults as users of conversational agents in health and social care: designers' and developers' motivations and images of ageing.

Abstract

Purpose The term conversational agents (CA) encompasses various other terms and is often used interchangeably with chatbots, virtual assistants (such as Apple's Siri, Microsoft's Cortana, or Amazon's Alexa), smart agents, and more. CA are used in various fields today, including healthcare and social care. While older adults are often considered as late adopters of digital technologies, they have been identified as a central user group of CA in in these fields (Merkel & Schorr, submitted) as CA are believed to offer "more advantages than challenges." (Even et al., 2022). On one hand, the technology offers a seemingly perfect match between the purportedly low digital skills and competencies of older adults and the ease of use provided by natural language processing, which requires no knowledge of graphical interfaces. On the other hand, there is evidence that the CA and AI are seen as a “techno-fix” for ageing societies (Higgs & Gilleard, 2021). The design and the use cases of CA are influenced by images of ageing (e.g. Neven, 2010; Endter, 2017; Merkel, Kucharski & Schorr, 2022). Against this background, this article aims to analyse the motivations of designers and developers of CA. The focus is on two questions: how the development of the devices is framed and what the underlying images of ageing are.

Method   In February and March 2023, a scoping review was conducted on smart speakers as a type of CA. Three databases (PubMed/Medline, Scopus, and Sociological Abstracts) were searched, along with the first 20 pages of Google Scholar. The search was conducted using a combination of multiple keywords, including the technology used, the brand, and the setting. Only publications that met the inclusion criteria were included. Two reviewers screened the title and abstract, and if they matched the criteria, the full text was also screened. To extract information from the publications, we used a data extraction template that includes general information such as the title and year of publication, as well as specific details like the setting and target group. 

Results and Discussion A total of 28 publications were included in the scoping review, twelve of which addressed older adults as a user group, mostly in the setting of their private households. This makes older adults the main user group identified, whether directly or indirectly e.g. through their relatives and family members. Our analysis shows that the ease of use and the distribution of the devices among private households are the main motivations cited. Furthermore, evidence suggests that designers and developers draw on negative stereotypes of ageing to identify use cases and justify their development. This supports the observations of previous research on other technologies such as robots (e.g. Neven, 2010) or Ambient Assisted Living (e.g. Endter, 2017). 

References

Endter, C. (2017): Assistiert altern. Die Entwicklung eines Sturzsensors im Kontext von Ambient Assisted Living [Assisted ageing. The development of a fall sensor in the context of Ambient Assisted Living]. In: Biniok, P., Lettkemann, E. (eds) Assistive Gesellschaft. Öffentliche Wissenschaft und gesellschaftlicher Wandel [Assistive Society. Public science and social change]. Springer VS, Wiesbaden, 167–181. https://doi.org/10.1007/978-3-658-13720-5_8.

Even, C., Hammann, T., Heyl, V., Rietz, C., Wahl, H.-W., Zentel, P., & Schlomann, A. (2022): Nutzen und Herausforderungen konversationeller Agenten für ältere Menschen: Ein Scoping-Review [Benefits and challenges of conversational agents in older adults: A scoping review]. Zeitschrift Für Gerontologie Und Geriatrie, 55(5), 381–387. https://doi.org/10.1007/s00391-022-02085-9.

Higgs, P., Gilleard, C. (2022): Techno-fixes for an ageing society, Aging & Mental Health, 26:7, 1303-1305. doi: 10.1080/13607863.2021.2008308

Merkel, S., Kucharski, A., Schorr, S. (2022): “Computer, how do smart speakers support aging in place?” – A feature analysis focusing on smart speaker applications in Amazon’s skill store for older persons. Gerontechnology, 21(1), 1-11. https://doi.org/10.4017/gt.2022.21.1.592.12

Merkel, S., Schorr, S. (submitted): Smart speakers in healthcare and social care. A scoping review. JMIR.

Neven L. 'But obviously not for me' (2010): robots, laboratories and the defiant identity of elder test users. Sociology of Health and Illness 1;32(2), 335-47. doi: 10.1111/j.1467-9566.2009.01218.x.

Keywords: Conversation agents, healthcare, social care, images of ageing

Address: Ruhr-University, Bochum

Email: sebastian.merkel@rub.de