Date Presented Accepted for AOTA INSPIRE 2021 but unable to be presented due to online event limitations.
This study explored office workers' perspectives on including artificial intelligence (AI) in their office workspace. Following an iterative analysis of six focus-group interviews with a total of 45 participants, three constructs emerged. Rich discussions demonstrated how acceptability of an AI workstation is complex and affected by the person, context, and their occupations.
Primary Author and Speaker: Yoko E. Fukumura
Contributing Authors: Julie McLaughlin Gray, Gale Lucas, Burcin Becerik-Gerber, and Shawn C. Roll
PURPOSE: Advances in machine learning that can inform artificial intelligence (AI) have opened a world of possibilities for promoting worker health and performance. In particular, AI-supported technologies could be a breakthrough support for office workers who are at risk for work-related musculoskeletal and psychosocial issues due to barriers in changing poor behaviors, habits, and routines (Soriano, Kozusznik, & Peiró, 2018). With machine learning capabilities, an AI office workspace can facilitate healthier work behaviors, such as improvement in posture, and prevent negative health impacts. There is limited knowledge of how implementation of AI would alter the occupational context for workers; thus, we aimed to explore office workers' perspectives introducing AI into their workspace.
DESIGN: To gain a rich description of office workers' experiences and preferences, we conducted a qualitative descriptive study involving focus group interviews. We used purposive network sampling to recruit office workers within personal professional networks in the U.S. (in-person) and U.K. (via video conferencing).
METHOD: Six focus group interviews were conducted by two to four moderators, each using a semi-structured interview guide. A total of 45 participants were interviewed. Focus group interviews were selected to facilitate rich discussion in novel topics (Kamberelis & Dimitriadis, 2011). Additionally, group discussions can support interactivity across a broad range of viewpoints and insights. All interviews were audio recorded and transcribed for analysis. Data analysis was completed by two researchers. Both read all transcripts and completed coding through a systematic text condensation procedure (Malterud, 2012). First, both researchers independently read the transcripts and identified and labeled all meaningful words and phrases. Next, transcripts were re-read and secondary codes were identified through an iterative process based upon clusters identified among first-level codes and dialogues. Finally, after developing initial themes, findings and implications were synthesized inductively after discussion with a third researcher with knowledge of the overall project.
RESULTS: Participants reported working at a desk for the majority of their workdays. Three constructs emerged from the transcripts. First, participants shared perspectives related to their preferences and concerns regarding communication and interactions with the technology. Second, numerous conversations highlighted the dualistic nature of a system that collects large amounts of data; the potential benefits for behavior change and the pitfalls of trust and privacy. Finally, numerous thoughts were shared relative to future AI solutions that could enhance the workplace. An overarching discussion related to the complexity of worker performance was noted within all three constructs. Rich discussions revealed how acceptability of an AI workstation is complex and affected by the person, context, and their occupations.
CONCLUSION: Through this exploratory qualitative study, we were able to understand the complexity of participants' experiences as office workers as well as their preferences in an AI workstation. One of the unique advantages of an AI workstation is its ability to develop and react to both the individual and the environment. Like transactions within daily occupations, the AI system should transact with the individual, occupation, and environment to capitalize on its machine learning capabilities (Dickie, Cutchin, & Humphrey, 2006). As technology continues to evolve, AI will shape the future of work. The findings from this study call for the use of OT and OS frameworks to understand human computer interaction and inform the development of an AI workstation.
References
Kamberelis, G. & Dimitriadis, G. (2011). Chapter 33. Focus Groups: Contingent Articulations of Pedagogy, Politics, and Inquiry. In N. K. Denzin & Y. S. Lincoln (Eds.), The SAGE handbook of qualitative research: 4th ed (pp. 545-562). Thousand Oaks California, CA: SAGE.
Malterud, K. (2012). Systematic text condensation: A strategy for qualitative analysis. Scandinavian Journal of Public Health, 40(8), 795-805.
Soriano, A., Kozusznik, M. W., & Peiró, J. M. (2018). From office environmental stressors to work performance: the role of work patterns. International Journal of Environmental Research and Public Health, 15(8).
Dickie, V., Cutchin, M., & Humphry, R. (2006). Occupation as transactional experience: a critique of individualism in occupational science. Journal of Occupational Science, 13(1), 83–93.
