Engaging Community Capabilities with Technology: A Mixed-Method Study of Human-Robot Social Learning for People with Disabilities

Background and Purpose: Socially Assistive Robots (SAR) promote autonomy, quality of life, and extend the lives and capabilities of individuals across the lifespan.  They provide assistance ranging in complexity from performing activities of daily living to participating in complex social roles (Birnbaum, Mizrah, Hoffman & Sass, 2016; Rabitt, Kazdan & Scasselati, 2014). It is critical to understand how best to benefit from these new advancements in technology as innovative approaches to learning and best practices are developed. Learning how users employ SAR and problem solve performance barriers can lead to human-robot collaborations that enhance mastery and capabilities (Bligard, & Osvalder, 2014). This study identifies factors that can be designed into SAR to better meet the needs of users with a disability and address emerging ethical considerations.

Methods: Sequential mixed-methods design with interviews and focus group were implemented followed by quantitative survey to test the measures assessing user preferences of SAR. Ten individual interviews and a focus group was held over the winter of 2015. Underlying themes and constructs were coded and used to inform the design of the survey applied to 244 participants with disabilities in examining technology exposure, preferences, usage patterns and criteria for sharing information among assistive technology users. Cross-tabs were used to understand user preferences based on demographics. Exploratory factor analysis was used to confirm the underlying constructs. The factorability of each item and the resulting dimensions were confirmed using the correlation matrix. Reliability of the constructs were assessed using the Cronbach’s alpha and cross-checked with the coded themes from the qualitative analysis.

Results: Key themes emerged from the qualitative data analysis show that users are concerned with design, learning tasks, functionality and confidentiality issues pertaining to SAR. Participants were keen on design features that reduces stigma, turnaround time in task, along with dynamically engaging with their SAR. They also raised concerns about confidentiality issues pertaining to stored data on SAR usage.

Survey result showed that a majority (63%) of all age groups (18 to 70) agreed that their SAR should engage dynamically with them on their tasks [χ² (20) =.38.542, p<.01]. More than 90% of men and women agreed that the users should have the final say over what their SAR share with others. Exploratory factor analysis with the Kaiser-Meyer-Olkin measure of sampling adequacy ≥.5 resulted in three rotated components with more than 50% of the variances explained. The Bartlett's test of sphericity was significant for compatibility [χ² (21)=405.39, p<.001], functionality [χ² (21)=639.51, p<.001] and personable [χ² (21)=407.37, p<.001].  All factors had Cronbach alpha of 0.80.

Conclusions and Implications: This self-reported study results indicate that users are eager to engage with SAR in their daily living. However, they would also like to have a say over how they engage with their SAR. Understanding user perceptions of their assistive technology can improve human-robot collaborations that enhance mastery and capabilities. This study contributes toward addressing the Grand Challenges for Social Work focusing on harnessing technology for advancing social justice and work towards enhancing user capabilities and inclusive environments.