Digital platforms commonly ask customers to select interest categories (e.g., genres/topics) as input for personalized recommendations. Twelve main studies and two pilot studies (total N = 8,824) reveal that customers share less diverse preferences with algorithms (versus human curators or when merely listing preferences for themselves); they focus on core preferences while omitting tangential ones, a phenomenon termed preference filtering . It is driven by customers’ expectation that algorithms weigh their preferences more uniformly than human curators (i.e., expected uniformity). A mathematical model, as well as interviews and a survey with practitioners show that preference filtering appears rational ex ante yet leads to negative consequences for both customers and firms ex post. The authors examine key design dimensions of the preference elicitation task— when preferences are elicited, how customers articulate them, what purpose is made salient, and who customers believe they are interacting with—that determine the extent to which customers engage in preference filtering. Two studies on self-developed video-streaming websites show that alleviating preference filtering can boost engagement and enhance customer reviews of recommendation services. These findings offer valuable insights for firms that rely on algorithms to engage customers.

Supporting decision-making is a central goal in visualization research, yet evaluating how effectively visualizations aid decision tasks remains difficult. Objective metrics for decision quality as a baseline for comparative evaluations are rare, with previous work proposing the consistency between a decision made and self-reported subjective preferences. However, the impact of preference elicitation design on such metrics is not well understood. Our focus is therefore on understanding how elicitation methods shape the sensitivity of decision quality metrics, not on comparing visualization techniques themselves. We report a preregistered study with 548 participants examining how varying the expressiveness of preference elicitation affects the sensitivity of choice consistency metrics when comparing parallel coordinates and tabular visualizations. Our baseline condition replicates a previous study using simple preference elicitation with significantly more participants and confirms its inconclusive findings that suggest a comparable performance of parallel coordinates and tabular visualization for the tested decision task. We further test two additional conditions to investigate if more expressive preference elicitation increases the metric's sensitivity. Our results suggest that more expressive elicitation may improve sensitivity in detecting performance differences, but further increases offer no clear additional benefit in our study. We observed these sensitivity differences even though the experimental setup, with its comparable baseline performances, was not optimized for detecting visualization performance differences. Our results provide initial evidence that elicitation design can affect the informational value of decisioncentric visualization studies, informing both the development of metrics and the interpretation of comparative studies. A preprint of this paper, along with the pre-registration, data analysis scripts, and all supplemental materials, is available at https://osf.io/u63zw/?view\_only=b8b04a7e505c45c89ae957674248c494.

ABSTRACT Electric micromobility (e-micromobility) has gained significant popularity in cities worldwide, yet urban infrastructure often fails to support adequate access to and use of these modes for a wide range of users. This systematic literature review applies the Capabilities Approach to identify key physical features of the built environment necessary to foster e-micromobility as an equitable and inclusive travel option. We analyse how personal, social, political, and economic conversion factors interact with physical features of the built environment to impact the translation of e-micromobility resources into three capabilities: e-micromobility access, mobility and accessibility. Drawing from the review of 31 studies, we discuss the interactions between conversion factors, propose a definition of equitable e-micromobility, and identify infrastructure elements to be prioritised by planners to foster the three capabilities. Key priorities include (i) dockless parking; (ii) dense and connected dedicated lanes, physically segregated from motorised vehicles and pedestrians; (iii) affordable fares of shared systems; and (iv) regulations to reduce conflicts between modes. The findings suggest the need for research that approaches e-micromobility equity from a systems perspective: focusing on the relationships between the conversion factors and on the interactions of e-micromobility with other travel modes. The explicit comparison of how differently e-bikes and e-scooters serve vulnerable groups and the direct elicitation of perceptions and preferences within these groups are also important gaps that should be filled to guide the design and implementation of more equitable e-micromobility systems.

Digital Health Technologies (DHTs) are expanding rapidly, offering new opportunities to support care delivery. Their adoption, however, depends on how well they match patients' needs and expectations. Accurately assessing patient preferences is challenging due to diverse user profiles and varying methods used to measure preferences. A synthesis of current evidence is needed to clarify what patients value in DHTs. This study synthesizes evidence on patient preferences for DHTs-including eHealth, telehealth, telemedicine, and mHealth-and examines the methods used to elicit these preferences, highlighting opportunities to improve adoption and design. We conducted a scoping review of literature published from 2000 to 2026 following PRISMA-ScR guidelines. Searches were performed in PubMed, EMBASE, CINAHL, Scopus, and Web of Science. Two reviewers independently screened titles, abstracts, and full texts for eligibility. Data were charted and narratively synthesized, with study characteristics categorized by methodology (qualitative, quantitative, mixed methods) and preference elicitation techniques. Of 2,419 records identified, 115 underwent full-text screening and 85 met all inclusion criteria: 27 qualitative studies (31.76%), 45 quantitative studies (52.95%), and 13 mixed-methods studies (15.29%). Quantitative studies primarily applied attribute-based methods (e.g., Discrete Choice Experiments, Conjoint Analysis, Best-Worst Scaling), and six studies used the Contingent Valuation method to estimate total willingness-to-pay for DHTs. Qualitative studies employed thematic analysis, deductive, inductive, and immersion-crystallization approaches. Across studies, patients consistently emphasized cost, privacy, convenience, and personalization of DHTs as key concerns. Only a few studies estimated the relative importance of the attributes or the marginal willingness to pay for the device's features. Considerable heterogeneity in preferences was observed by DHT type, patients' health condition and age. Quantitative, qualitative, and mixed-method approaches provide complementary insights into how patients perceive and value DHTs. Patients consistently seek personalized, easy-to-use, secure, and affordable solutions, which is particularly important for older adults managing chronic conditions. Given the diversity of patient preferences, it is essential to consider these differences when developing digital health technologies to ensure they effectively support patients and address their needs.

A deck of cards-based co-constructive approach for modeling higher-order uncertainty in fuzzy decision-making

ABSTRACTRationale, Aims and ObjectivesMost methods for elicitation of values and preferences (V&P) ask respondents directly to make explicit numerical trade‐offs across outcomes, such as judging how many strokes equal one death. Although conceptually straightforward, these tasks are cognitively demanding, uncomfortable for many patients and panellists, and prone to instability when multiple outcomes must be compared. This paper proposes an integrated framework that starts with minimally important differences (MIDs), links them to acceptable regret, and then converts them into relative values (RVs) for use as V&P in decision‐analytical models.MethodsWe describe a three‐step approach. First, respondents identify the smallest absolute change in outcome frequency that would be important enough to justify a different decision. Second, acceptable regret is used to interpret and calibrate these thresholds as the amount of utility loss from a wrong decision that patients can tolerate. Third, MIDs are transformed into RVs on a common scale anchored to a worst outcome, usually death, and entered as V&P into a weighted disutility expected utility model.ResultsThe resulting framework is designed to replace difficult, cognitively demanding elicitation of V&P with simpler threshold judgements, maintains proportional relationships among outcomes regardless of the chosen anchor, reduces the number of required judgements, and yields internally consistent weights for multiple benefits and harms. A worked example shows how MID thresholds for death, stroke, myocardial infarction, major bleeding and brain bleeding are converted into RVs and then incorporated into a transparent benefit‐harm calculation.ConclusionsIntegrating MIDs, acceptable regret and RVs offer a coherent approach to eliciting values and preferences for clinical decision‐making, guideline development and health policy. The approach is easy to explain, well aligned with human decision processes and readily applicable at both individual and population levels.

This paper investigates how group deliberation changes individual distributional preferences. We experimentally assess the relative contribution of persuasion, social identity, and social comparison to shifts in preferences following deliberation. In a controlled setting, participants engaged in ten minutes of non-binding written group deliberation about distributional choices. Post-deliberation preferences became significantly more egalitarian than pre-deliberation ones. This within-subject preference shift is supported by a between-subject comparison showing that group deliberation has a larger egalitarian effect than individual deliberation. What explains this egalitarian shift? Our findings suggest that social identity formation is the primary but not unique driver of the change in preferences. Social identity appears to largely explain the pronounced egalitarian shift among participants who lose from equality, while persuasion and social comparison seem to account for the preference changes among those whose material payoffs are unaffected by the distributive outcome. These findings have important implications for the elicitation of distributional preferences and for the design of communicative institutions that precede collective decision-making.

IntroductionNon-pharmaceutical interventions (NPIs) such as mask use, physical distancing, business or school closure, and isolation have been central to controlling respiratory viral infections, including influenza, COVID-19, and respiratory syncytial virus (RSV). Understanding the preferences of the public and stakeholders for these interventions is critical to ensure their acceptability, uptake, and effectiveness. Discrete choice experiments (DCEs) provide a robust method to quantify how individuals value different attributes of NPIs and the trade-offs they are willing to make. However, evidence from DCEs is fragmented across diseases, populations, and methodological approaches. This scoping review aims to systematically map and synthesise evidence from DCEs examining public and stakeholder preferences for NPIs used in the prevention and control of influenza, COVID-19, and RSV. Specifically, it will: (1) summarise the attributes and attribute levels used in DCEs, and the methods employed to select them; and (2) identify which NPI attributes are most valued by the public and key stakeholders, including patients, healthcare workers, and policymakers.Methods and analysisThe review will follow the PRISMA-ScR framework. Comprehensive searches of electronic databases (PubMed, Scopus, and Embase) will identify DCEs evaluating NPIs for influenza, COVID-19, and RSV. Data extraction will capture study characteristics, target populations, attributes, experimental design, and analytical methods. Reporting quality will be appraised using the DIRECT Checklist. Given anticipated heterogeneity, findings will be synthesised narratively. By providing a structured overview of existing DCE evidence, this review will inform the design of future preference studies and guide the development of acceptable and evidence-based NPI strategies for respiratory viral infections.Ethics and DisseminationNo ethical approval is required. The completed review will be shared through peer-reviewed journals and conference presentations.Open Science Framework Registration https://doi.org/10.17605/OSF.IO/H36UC

Shopkeeper robots face a critical trade-off when recommending products to first-time customers. Robots must balance recommendation efficiency with unconstrained natural language interaction. Sparse modeling techniques like Lasso regression offer efficient feature selection from limited samples. However, they typically demand numerical evaluations from customers, degrading the social experience. We define this dilemma as the Qualitative-Quantitative Divergence Problem and systematically compare the impact of numeric and non-numeric (natural language) recommendation dialogues on the HRI experience. As a verification platform, we implemented Impression-SmartClerk. This system extracts recommendation parameters from natural utterances via LLMs, rather than forcing customers to adapt to mathematical models, and ensures robustness through intent-adaptive conversation repair. Results from a physical robot experiment with 20 participants demonstrated that the non-numeric approach significantly improves the freedom of self-expression and the robot's likeability, while maintaining recommendation accuracy equivalent to numerical input methods without increasing cognitive load. Ultimately, this article demonstrates the resolution of the Qualitative-Quantitative Divergence Problem, and through an empirical comparison, provides valuable interaction-level insights into the UX trade-offs among expressiveness, robustness, and recommendation quality.

Background/Objectives: Chronic Kidney Disease (CKD) is a prevalent condition requiring ongoing patient counseling and engagement, yet little is known about how physicians communicate with patients about CKD in routine clinical practice. We conducted a qualitative study to examine physician communication approaches related to CKD and to assess how these approaches align with Picker’s principles of patient-centered care framework. Methods: Semi-structured interviews were conducted with primary care physicians and nephrologists practicing in community and safety-net settings. Using directed content analysis, we identified patterns in how clinicians describe educating patients, contextualizing clinical information, and deferring aspects of counseling to other providers. Results: Physicians predominantly emphasized information-giving and the use of laboratory data to explain disease status. In contrast, practices such as explicit patient preference elicitation, addressing fear, anxiety, or physical comfort, and involving family or support persons were infrequently described. Mapping these communication behaviors to patient-centered care principles highlighted specific elements that are routinely enacted and others that remain underutilized in everyday CKD counseling. Conclusions: These findings identify concrete, feasible opportunities to strengthen patient-centered communication through brief, practice-ready strategies such as plain-language explanations, teach-back, values checks, and shared decision-making prompts. Enhancing these communication practices represents a pragmatic opportunity to improve the quality and patient-centeredness of CKD care.

Shared decision-making (SDM) is required for lung cancer screening (LCS) and is intended to support informed, patient-centered decisions. However, SDM implementation in primary care remains inconsistent, particularly in rural settings where LCS uptake is low. This study examined which conversation elements within clinical interactions are associated with LCS completion and may support SDM discussions. We conducted a cross-sectional survey with screening-eligible adults residing in rural Appalachian communities. Survey items were adapted from the DECISIONS instrument to assess conversation elements within LCS-related clinical interactions, information sources, content discussed, decision attributes, and post-conversation perceptions. Associations between conversation elements and LCS completion were examined using chi-square tests, t-tests, and multivariable logistic regression. Socioeconomic factors were explored to contextualize findings in a rural population with documented health disparities. Provider initiation of the LCS conversation and provision of educational materials were the two conversation elements independently associated with screening completion. Additional conversation elements, including elicitation of patient preferences and clear provider recommendations, were associated with screening completion in bivariate analyses. Use of educational materials and decision support aids was strongly associated with patients' perceptions of being informed and confident, which were in turn associated with LCS decisions. Socioeconomic indicators were associated with variation in several conversation elements relevant to SDM. Findings highlight that discrete conversation elements that reflect SDM interactions are differentially associated with behavioral and perceptual outcomes, with provider initiation and educational materials linked to screening completion and decision support and clear recommendations linked to feeling informed and confident. Identifying and prioritizing a small set of key conversation elements that support SDM interactions may inform LCS delivery in rural primary care.

ABSTRACT The Flexible and Interactive Tradeoff Elicitation (FITradeoff) method is a Multi‐Attribute Decision‐Making (MADM) approach designed to capture the preferences of a Decision Maker (DM) while minimising cognitive effort. To reduce the frequency of interactions and optimise the preference elicitation process, this paper introduces an innovative FITradeoff method integrated with Stochastic Multi‐Attribute Acceptability Analysis‐2 (SMAA‐2). The proposed method follows six steps: (1) It identifies the central weight vectors of each potentially optimal alternative obtained through SMAA‐2. (2) It formulates pairwise tradeoffs based on their ratios and selects the most informative ones based on their probability of identifying potentially optimal alternatives. (3) It selects the most informative pairwise tradeoff and its ratio based on the minimum number of potentially optimal alternatives. (4) It engages the DM to express a preference relation. (5) It constructs an updated weight space with the identified pairwise tradeoff constraint and iterates the previous steps until an optimal alternative is identified. (6) A Genetic Algorithm‐based Linear Programming (GA‐based LP) model is developed to evaluate the robustness and efficiency of our approach. To prove the feasibility and validate the effectiveness of the proposed approach, a case study on the selection of Battery Energy Storage Systems (BESS) is conducted. Additionally, a comparative analysis with the traditional FITradeoff method is included; the results demonstrate that the proposed method identifies the optimal solution while reducing the DM's cognitive burden, highlighting its potential to improve decision‐making processes.

To examine gynecological cancer patients' preferred role in decision making, their experienced involvement, and their advice to patients and clinicians on preparing for and supporting SDM in clinical consultations. Two validated questionnaires, the Control Preference Scale (CPS) and CollaboRATE, were used to assess patients' preferred roles in SDM and their perceived level of involvement in medical decisions. Two open-ended questions were included to gather descriptive advice from patients, intended for both patients and clinicians. The sample for this survey was drawn from participants in a Danish patient advocacy group. In total, 117 patients completed the CPS, with 90% (n = 105) indicating a preference for an active role and 10% (n = 12) indicating a preference for a collaborative role in decision making. Mean item scores for CollaboRATE (n = 114) were 6.6 (SD=2.1) for explanation, 5.2 (SD = 2.6) for preference elicitation, and 5.5 (SD = 2.5) for integration, with 7.9% of the respondents giving a top score. The most repeated theme of advice to patients was to ask questions. Empathic communication, including active listening and the provision of clear, easy-to-understand information, was the most frequently emphasized advice to clinicians. Patients with gynecological cancer express a strong desire for involvement in SDM; however, many report that their experienced level of participation falls short of their preferences. They recommend an active role by asking targeted questions, thoroughly understanding benefits and risks of options, and remaining persistent in expressing needs. Patients emphasized the importance of empathetic communication, active listening, and the provision of clear, easy-to-understand information by clinicians. Our findings offer actionable recommendations to bridge the gap between patients' preferred role in SDM and their actual experiences across the care journey. By incorporating these recommendations, both patients and clinicians can adopt practical strategies to facilitate more personalized, patient-centered decisions aligned with individual preferences and needs.

Large language model-enhanced home energy management with dynamic user preference elicitation and hierarchical data-sharing

The equivalent income is a preference-based, interpersonally comparable measure of well-being. Although its theoretical foundations are well established, empirical applications remain limited, primarily due to the detailed data requirements on individuals' preferences across various well-being dimensions. This paper reviews the literature on preference elicitation methods with a focus on estimating equivalent income. We examine several survey-based methods, including contingent valuation, multiattribute choice or rating experiments, and life satisfaction regressions. The review highlights the advantages and limitations of each method, emphasizing the considerable scope for methodological improvements and innovations.

In this article, we investigate using explanations to improve decision making for robotic scientific data collection missions. We propose the Preference Elicitation with contrasting Feature-based eXplanations (PrEFeX) method, which combines preferences with contrasting explanations focused on a single explanatory feature. We first provide a verification of the contrasting explanations by themselves using a planner prediction user study with 16 expert participants (Phase 1). This study showed that there was no increase in understanding of robot plans using contrasting explanations without preferences. To elucidate what information the autonomous measurement selection system was missing to be useful, we interviewed 4 planetary scientists and 2 oceanographers (Phase 2). We found that scientists focused heavily on understanding the objectives of the system and wanted explanations that (1) grounded the explanations to tradeoffs the system made, (2) connected the explanations to an ability to modify the behavior of the decision making, and (3) attached the explanation system's features to comparable features the scientists considered. To this end, we propose combining explanations with user preference learning of the reward function in an iterative design process of the measurement plans with scientists in the loop. We tested our proposed preference and explanation system in a field deployment with planetary scientists on Mt. Hood, Oregon and performed a post data collection survey on the quality of the plans with 22 experts (Phase 3). We found the experts preferred the measurement plan selected by our proposed PrEFeX method over a baseline without explanations or preferences.

Risk Preference Elicitation in Finance: Survey vs. Experiment

Building on the framework for moral artificial intelligence (AI) proposed by Schaich Borg, Sinnott-Armstrong and Conitzer (SSC), we discuss what would be required for AI as a moral decision-making aid in the context of medical decision-making. SCC outlines a five-step approach that centres on how to best handle the training data for AI: survey people’s moral views, use preference elicitation methods to ascertain the weights of different considerations, idealise preferences to avoid the problem of bias based on ignorance, aggregate individual preferences to group judgements, and model moral decision-making. While their framework is plausible and implementable, we argue that it rests on three problematic assumptions about the three pairs of similar but distinct concepts. The aim of this article is to use the SSC framework as a starting point for developing an account of moral AI that preserves the strengths of their model while adding further features suitable for assisting moral decision-making. In order to outline a more comprehensive model of moral AI, we emphasise three conceptual distinctions: (1) preferences versus reasons, (2) rankings versus deliberation and (3) predictions versus judgements. The resulting approach focuses on the latter concepts of these pairs and suggests a version of moral AI that retains the virtues of the SSC approach while avoiding some of the potential pitfalls.

Valuing Child and Adolescent Health States to Derive Utilities for Use in Economic Evaluation: A Good Practices Report of an ISPOR Task Force.

Abstract In this work we are concerned with the design of efficient mechanisms while eliciting limited information from the agents. First, we study the performance of sampling approximations in facility location games. Our key result is to show that for any $$\epsilon > 0$$ ϵ > 0 , a sample of size $$c(\epsilon ) = \varTheta (1/\epsilon ^2)$$ c ( ϵ ) = Θ ( 1 / ϵ 2 ) yields in expectation a $$1 + \epsilon $$ 1 + ϵ approximation with respect to the optimal social cost of the generalized median mechanism on the metric space $$(\mathbb {R}^d, \Vert \cdot \Vert _1)$$ ( R d , ‖ · ‖ 1 ) , while the number of agents $$n \rightarrow \infty $$ n → ∞ . Moreover, we study a series of exemplar environments from auction theory through a communication complexity framework, measuring the expected number of bits elicited from the agents; we posit that any valuation can be expressed with k bits, and we mainly assume that k is independent of the number of agents n . In this context, we show that Vickrey’s rule can be implemented with an expected communication of $$1 + \epsilon $$ 1 + ϵ bits from an average bidder, for any $$\epsilon > 0$$ ϵ > 0 , asymptotically matching the trivial lower bound. As a corollary, we provide a compelling method to increment the price in an English auction. We also leverage our single-item format with an efficient encoding scheme to prove that the same communication bound can be recovered in the domain of additive valuations through simultaneous ascending auctions, assuming that the number of items is a constant. Finally, we propose an ascending-type multi-unit auction under unit demand bidders; our mechanism announces at every round two separate prices and is based on a sampling algorithm that performs approximate selection with limited communication, leading again to asymptotically optimal communication. Our results do not require any prior knowledge on the agents’ valuations, and mainly follow from natural sampling techniques.