Aging societies are facing rising rates of amnestic mild cognitive impairment (aMCI), which calls for digital assistance that preserves autonomy rather than reinforcing dependency. We propose augmented living spaces, a concept leveraging Spatial Augmented Reality (SAR) and context-aware decision support. Such environments provide in-situ, object-anchored cues that adapt and fade as competence grows. Based on Vygotsky's Zone of Proximal Development as our kernel theory, we have developed a set of prescriptive design principles that operationalize scaffolding, intersubjectivity, and semiotic mediation for everyday cognitive support. Using a theory-driven BAUSTEIN configuration within the Design Science Research paradigm, we derive design principles from kernel theory mechanisms, practical foundations (including focus group and observational insights), and relevant literature. Then, we instantiate the design principles in a prototypical scenario for evaluation with experts and our target group of aMCI individuals. A two-strand evaluation with ex-ante and ex-post episodes indicates that projected, just-in-time guidance is perceived as intuitive, nonpatronizing, and cognitively relieving. The set of design principles is also judged useful, clear, and transferable beyond the prototype. Our contribution is design knowledge grounded in mechanisms for autonomy-supportive digital health decision support systems, comprising concise design principles that can guide SAR-enabled, context-aware assistance toward safety without surveillance and help without dependency, thereby providing a practical path to extend independent living.
Integrating machine learning (ML) into decision support systems (DSS) has improved the efficiency and quality of decision making. Yet historical data can embed hidden biases that may lead to unfair outcomes, making algorithmic bias an important societal concern. To address this issue, we develop a fairness-aware DSS based on adversarial learning (AL). The system uses a dual-network architecture: a prediction network that learns task-relevant representations and predicts target outcomes, and an adversarial network that infers protected attributes and thereby helps mitigate bias. As a computationally oriented design science artifact, our DSS embeds fairness norms as executable mechanisms within the model training phase, supporting fairer decisions in practice. We evaluate the model on a Chinese P2P lending dataset and the German Credit dataset. The results show that the proposed approach performs well with respect to four standard fairness criteria, namely unawareness, independence, sufficiency, and separation, while achieving higher predictive accuracy than widely used ML and AL baselines. Overall, this study advances the design of fairness-aware DSS and provides evidence that adversarial learning can reduce algorithmic bias and support fairer decision making.
Gamified virtual reality (VR) is an increasingly powerful tool for engaging users with experiential products. Yet little is known about the cognitive mechanisms through which gamified VR shapes behavioral intention, thereby limiting the development of effective decision support systems. Across three experimental studies employing both desktop and head-mounted VR devices, we built on the cognitive absorption theory to examine the synergistic effects of VR interaction and gamification on users' behavioral intention. Further, we show that different gamification design orientations (achievement- vs. narrative-oriented) selectively reinforce distinct dimensions of cognitive absorption, and that users' motivational goals (pursuit of meaning vs. pleasure) moderate the effectiveness of these designs. Our findings enrich cognitive absorption theory by revealing its dimension-specific pathways in immersive environments, introduce motivational goals as critical boundary conditions for gamified interaction effectiveness, and extend IS research on VR decision support by highlighting interaction as the critical driver of cognitive engagement. For practitioners, the results offer evidence-based guidance for tailoring VR gamification strategies to foster more effective decision-making processes and product promotion.
With the penetration of social media into our daily life, video sharing platforms have become the mainstream venue for information exchange and content dissemination. However, the mechanism by which nonverbal cues drive viewer engagement in user-generated videos remains unclear. Drawing on emotional contagion theory, we examine the curvilinear relationship between the speakers' vocal cues (pitch, intensity, jitter and shimmer) and viewer engagement and the moderating effects of visual cues (speaker presence). Leveraging 40,742 observations obtained from Bilibili and an online experiment, the results indicate that a U-shaped relationship between pitch/intensity and viewer engagement and an inverted U-shaped relationship between jitter/shimmer and viewer engagement, both mediated by emotional arousal. Speaker presence strengthens the U-shaped effect of intensity and the inverted U-shaped effect of shimmer. This study contributes to user engagement literature by revealing the unique effects of nonverbal cues in user-generated videos and offering practical guidance for content creators on social media.
While advertising sustains social media platforms, its intrusive nature threatens both user experience and platform sustainability. Social newsfeed advertising (SNA) intrusiveness denotes users' psychological reactance elicited when unsolicited ads disrupt their processing on newsfeeds. Beyond delivering product information, SNA embeds social cues-such as Likes, Shares, Ratings, and Sales-into users' social contexts. However, prior research provides limited insights into how heterogeneous social cues mitigate intrusiveness, particularly regarding the mechanisms and conditions underlying these effects. Drawing on attitude function theory, this study examines how distinct social cues reduce intrusiveness via sociability and informativeness, and delineates the product and user conditions that amplify these effects. Two between-subjects experiments (N = 304; N = 306) on a custom-built platform show that proximal cues reduce intrusiveness for social products and high selfmonitors by enhancing sociability, whereas remote cues benefit utilitarian products by increasing informativeness. Attitude-based cues primarily foster sociability, whereas action-based cues strengthen informativeness. These findings clarify how social cues operate within SNA and offer actionable guidance for reducing intrusiveness.
