Design Foundations for AI Assisted Decision Making: A Self Determination Theory Approach

Abstract

Progress of technology and processing power has enabled the advent of sophisticated technology including Artificial Intelligence (AI) agents. AI agents have penetrated society in many forms including conversation agents or chatbots. As these chatbots have a social component to them, is it critical to evaluate the social aspects of their design and its impact on user outcomes. This study employs Social Determination Theory to examine the effect of the three motivational needs on user interaction outcome variables of a decision-making chatbot. Specifically, this study looks at the influence of relatedness, competency, and autonomy on user satisfaction, engagement, decision efficiency, and decision accuracy. A carefully designed experiment revealed that all three needs are important for user satisfaction and engagement while competency and autonomy is associated with decision accuracy. These findings highlight the importance of considering psychological constructs during AI design. Our findings also offer useful implications for AI designers and organizations that plan on using AI assisted chatbots to improve decision-making efforts.

Do Men Have No Need for “Feminist” Artificial Intelligence? Agentic and Gendered Voice Assistants in the Light of Basic Psychological Needs

“…Nonetheless, the potential role of the BPN has been largely neglected in AI research and development. With the current study, we aimed to complement the limited literature on the role of BPN for technology acceptance ( Nguyen and Sidorova, 2018 ; De Vreede et al, 2021 ; Jiménez-Barreto et al, 2021 ). In particular, we wanted to contribute to a better understanding of how two specific design parameters of AI assistants are related to the fulfillment of BPNs and how this, in turn, affects behavioral intentions to use a system.…”

“…Interestingly, empirical research that investigates the role of BPN on AI acceptance is scarce to date. Recently, one of the first studies to draw a link between BPN Fulfillment and AI has investigated chatbot-assisted decision-making ( De Vreede et al, 2021 ). The results revealed that a stronger experience of autonomy, competence and relatedness was indeed associated with higher user satisfaction, which subsequently led to greater engagement with the chatbot.…”

“…Even though only a few studies have investigated the role BPN play in technology acceptance, recent findings suggest them to be linked to user satisfaction and engagement with a given technology ( De Vreede et al, 2021 ). As the BPNT ( Deci and Ryan, 2000 ; Ryan and Deci, 2000 ) and empirical work on BPN indicate, need satisfaction is an important foundation of behavioral motivation (e.g., to engage in a certain task or use a certain product).…”

“…However, the results of the agency conditions on Competence Satisfaction could go in both directions. Since recent research has pointed at Competence Satisfaction being linked to greater engagement with a chatbot ( De Vreede et al, 2021 ), we assume a positive impact of Competence Satisfaction on ITU. However, the high-agency conditions could either lead to lower Competence Satisfaction (higher Competence Frustration) as participants see their own competence lower in comparison with the high-agency finance coaches’ competence; on the other hand, the high-agency finance coaches could empower the participants and thus satisfy their Competence Need.…”

Moving toward AI-assisted decision-making: Observation on clinicians’ management of multimedia patient information in synchronous and asynchronous telehealth contexts

“…1 With the recent advancement of AI in various industrial sectors, it has played a significant role in individuals' decision-making by providing extracted insights from big data analytics. 2 In that sense, AI-assisted CDM in healthcare can be defined as the process of choosing among options by interacting with AI technology and its analytic suggestions. However, prior studies have documented mixed effects of AI assistance on clinicians' decisionmaking.…”