Social engineering attacks are successful due to the exploitative nature of human psychology. Although security solutions, such as 2FA, aim to reduce the severity of attacks, these approaches tend to be insufficient considering sophisticated phishing attacks. In the present paper, we investigate how the design of the 2FA method may influence the likelihood of a login to a malicious website. Through an online experiment with N=94 participants, we show that contrary to common assumptions, warning designs prompting to proceed with login might have a contradicting effect. In contrast, designs that prompt users to abort the login have minimal desired effects. Moreover, we identify that involvement in login activity and confidence in the decision made had further significant effects on the likelihood of login. Our exploratory results contribute to the knowledge of susceptibility to phishing attacks and potential misconceptions about the effects of opinionated design in the context of 2FA.

Algorithmic news ranking and large language models (LLMs) increasingly mediate how citizens receive political information. While personalization is often criticized for reinforcing echo chambers, ranking has also been proposed as a lever for depolarization by shaping exposure to cross-cutting viewpoints. We test this claim in an online experiment (N=100, Prolific) using five LLM-generated news articles on gun legislation spanning pro–gun freedom to pro–gun control. All participants read the same articles; the only manipulation was article order, instantiated as a random baseline and six stance-aware depolarization-oriented strategies (counter-narrative sandwich, balanced alternation, and directional gradients). Pre–post questionnaires measured ideological self-placement (feeling thermometer) and affective evaluations of gun-control and gun-freedom advocates. Across six research questions, we find no statistically reliable evidence that any ranking strategy reduces ideological or affective polarization relative to the baseline, nor that stance-conditioned assignment improves outcomes. These results suggest that, under the constraints of a short, single-session, single-topic exposure study with LLM-generated article stimuli, article ordering alone did not produce detectable depolarization effects. The findings motivate future work to design longer-term, multi-topic, and more interactive solutions for mitigating affective polarization.

Vibrotactile feedback often supports mobile microinteractions, yet designers still apply it as a static enhancement. We test when vibrotactile feedback adds value versus becomes redundant by manipulating tactile sensitivity. In a controlled study (n=60), participants completed mobile microinteractions with and without vibrotactile feedback while wearing touchscreen gloves or using bare hands. With gloves, vibrotactile feedback compensated for reduced sensitivity and improved efficiency, accuracy, and satisfaction. With bare hands, feedback mainly reduced errors and did not improve completion time or overall experience, suggesting limited benefit under high tactile certainty. These results position haptic feedback as an adaptive interface element that should respond to inferred tactile certainty and context, motivating user modeling and personalization of feedback strategies.

The increase in energy and water consumption of artificial intelligence (AI) services, such as GPT, poses significant sustainability challenges. To overcome these challenges, user-centered strategies are necessary to reduce the environmental impact without compromising the user experience. This study investigates how visual nudges influence users’ decision making in AI interfaces and explores the trade-offs users make between processing time, output quality, and environmental impact. A between-subjects online experiment (𝑁 = 93) using mock-up AI interface scenarios and a choice-based conjoint design was conducted. Participants selected among processing options, generating a specific result of a prompt, with one group exposed to a visual nudge and the other not. The options to choose varied in processing time, output quality and environmental impact. Results show that the visual nudge increases the selection of environment friendlier options, particularly trading off output quality rather than processing time. The study also found evidence of a compromise effect, where users frequently chose moderate options to balance sustainability with usability. Environmental attitudes, measured via the New Ecological Paradigm (NEP) scale, significantly predicted sustainable choices, though they did not moderate the nudge’s effect. These findings contribute to sustainable human-computer interaction (HCI) by demonstrating that visual nudges can support greener computing behaviors, especially when aligned with acceptable performance trade-offs.

Personalized news systems often adapt ranking to user preferences, which can reinforce selective exposure. We investigate an alternative personalization strategy: stance-aware incentive shaping. Instead of modifying content ranking, we use a minimal per-topic stance user model to adapt reward gradients, awarding more points for engaging with counter-attitudinal content. In a between-subject experiment (N = 98), we compare a non-adaptive baseline with two adaptive incentive framings (levels vs. leaderboards). Stance-aware gamification increased behavioral engagement (clicks and time), while subjective engagement and comprehension did not differ reliably. Only the level-based framing produced significant pre-post increases in stance change and open-minded thinking, with effects varying by topic. We position stance-aware incentive shaping as a lightweight user-model intervention that adapts motivational feedback rather than ranking, offering an alternative pathway for diversity-aware personalization in recommender systems.

A variety of therapies are used in stroke rehabilitation to help patients regain abilities that they may have lost following a stroke. Virtual hand exercises present a low-cost approach to enabling stroke patients to embark on immersive journeys to do hand movement via immersive virtual reality (IVR). With technological advancements, ongoing evaluation and integration of the latest IVR innovations will be crucial to maximizing their potential in enhancing the quality of life for individuals with mobility disorders and stroke survivors. A total of 68 participants took part in our study, followed by a design thinking process. Using Unity software, we developed a VR gamified system that allows users to enter a museum and feel, enjoy, track, and explore a VR-based painting (Mona Lisa Frame and The Art Basel Banana) as game elements. These lead to moving or holding the VR-based painting as a hand exercise. We tested the VR gamified system on relevant stakeholders (patients and their relatives and physiotherapists), where they filled out a qualitative survey and a quantitative SUS (System Usability Scale) questionnaire. The result of the study showed that the usability effect of the game elements was positive in the context of user satisfaction in supporting virtual hand exercises.

Selective exposure in online news is often attributed to personalization mechanisms and user modeling. Recent work proposes interface-level interventions that visualize reading balance or frame cross-cutting exposure as progress. However, we lack empirical evidence on whether alternative representations of user-model feedback meaningfully influence engagement with belief-opposing content. We conducted a between-subject experiment (N = 84) in a controlled news environment comparing two representations of diversity feedback: (1) an analytic bias visualization summarizing viewpoint balance and (2) a metaphorical growth visualization framing cross-cutting exposure as personal development. Across behavioral and attitudinal measures of open-minded engagement, neither feedback representation increased engagement relative to control, and the two designs did not differ reliably. Our results suggest that lightweight representations of diversity signals, without adaptive personalization or structural changes to recommendations, may be insufficient to alter selective exposure in single-session settings. We discuss implications for designing user-model feedback and depolarization objectives in recommender systems. 

Large language models (LLMs) are becoming a common feature in content creation tools, prompting important questions about how design choices influence user trust and engagement in AI-assisted journalism. Beyond output quality, factors such as prompt specificity, model choice, and authorship disclosure are themselves interaction design parameters that influence how users interpret and evaluate AI contributions. Yet, little is known about how these design decisions affect reader perceptions in journalistic contexts. To address this gap, we conducted an experiment with 150 participants who evaluated news articles on the sensitive topic of assisted suicide. The articles systematically varied in authorship (human-written, AI-edited, or AI-generated), stance (pro- or anti-legalization), and prompt specificity (vague, moderate, or highly detailed). Participants rated each article on engagement, subjectivity, and perceived AI involvement, and also provided open-ended justifications for their authorship judgments. Our findings show that prompt specificity and model choice significantly influence perceptions of authorship, underscoring how technical design decisions in AI tools can shape public trust in journalism. 

The built environment plays an important role in human well-being, performance, and daily activities. Lighting, one of the primary indoor environmental quality factors, not only influences vision but also physiological and psychological processes associated with autonomic regulation, with respiration frequency serving as a measurable indicator. The feasibility of using millimetre-wave radar for 1) occupancy detection, 2) respiration frequency estimation, and 3) exploration of respiration frequency as an indicator for emotion-aware systems was investigated using two experimental frameworks. First, a controlled setup using two custom-built Controlled Airflow Thorax robots (CAT-bots) simulated consistent chest movements to systematically evaluate occupancy detection and respiration frequency estimation across scenarios with zero, one, or two occupants at varying distances from the radar. The second experiment included human participants (individually or in pairs) seated at different distances from the radar, watching emotion-evoking videos. Radar performance was evaluated with real occupants and examined whether radar-estimated breathing frequency varied across different emotional states (high/low valence/arousal). The results showed reliable occupancy detection, with weighted F1-scores of above 0.90 for both experiments. Respiration frequency estimation achieved 80 % within a clinically meaningful tolerance of ±3 bpm. Repeated-measures ANOVA showed that radar respiration frequency was significantly different for high versus low arousal videos, while no significant differences were observed across valence conditions. The findings demonstrate the feasibility of radar-based occupancy and respiration monitoring in indoor environments with varying occupancy; however, the small arousal-related differences relative to the estimation accuracy indicate that further investigation is required before respiration frequency can support emotion-responsive lighting systems. 

Reflection emerges when familiar scenes are revisited from a new angle, yet such moments do not always arise on their own during routine encounters. We present LudiScape, an ambient interface developed as a design probe to examine how staged visual augmentation shapes reflective experience through a five-level Visual Transformation Spectrum. LudiScape is instantiated in two metaphorical versions: Mirror for private reflection and Frame for asynchronous shared reflection. In a controlled within-subjects study (N=18), reflective experience increased relative to the baseline level in both versions. However, the most supportive level differed, peaking at Level 2 in Mirror and Level 4 in Frame. The findings suggest that degrees of reflective cue explicitness may support private and shared reflection differently, with lighter prompting appearing more effective in Mirror and more explicit traces appearing more effective in Frame.