Group Deliberation Research Articles

Conversational Agents to Facilitate Deliberation on Harmful Content in WhatsApp Groups

WhatsApp groups have become a hotbed for the propagation of harmful content including misinformation, hate speech, polarizing content, and rumors, especially in Global South countries. Given the platform's end-to-end encryption, moderation responsibilities lie on group admins and members, who rarely contest such content. Another approach is fact-checking, which is unscalable, and can only contest factual content (e.g., misinformation) but not subjective content (e.g., hate speech). Drawing on recent literature, we explore deliberation---open and inclusive discussion---as an alternative. We investigate the role of a conversational agent in facilitating deliberation on harmful content in WhatsApp groups. We conducted semi-structured interviews with 21 Indian WhatsApp users, employing a design probe to showcase an example agent. Participants expressed the need for anonymity and recommended AI assistance to reduce the effort required in deliberation. They appreciated the agent's neutrality but pointed out the futility of deliberation in echo chamber groups. Our findings highlight design tensions for such an agent, including privacy versus group dynamics and freedom of speech in private spaces. We discuss the efficacy of deliberation using deliberative theory as a lens, compare deliberation with moderation and fact-checking, and provide design recommendations for future such systems. Ultimately, this work advances CSCW by offering insights into designing deliberative systems for combating harmful content in private group chats on social media.

An exploratory study of blind spot bias in psychiatrists: The value of mindfulness and ethical skills’

IntroductionPsychiatry is an area of medicine that is particularly exposed to the risks of cognitive bias, notably because diagnosis is subjective. However, psychiatrists may develop psycho-cognitive skills that protect against bias, as the reflective approach to their patients favors the activation of system 2 (logical) as opposed to system 1 (intuitive) processes. Furthermore, the collegial practices found in psychiatry, a consequence of the ethical questions psychiatrists must address, can also mitigate the impact of bias. On the other hand, clinical ethics committees have adopted structured deliberation practices that aim to improve collective decision-making and limit the cognitive biases associated with groupthink. ObjectiveThe primary aim was to compare blind spot bias in a group of psychiatrists and a group of ‘ethicists’, experts in group deliberation. The secondary objective was to study the distribution of skills influencing the balance between systems 1 and 2 (notably, ethical deliberation, feeling of efficacy, mindfulness, and values) in both groups. The general hypothesis is that psychiatrists are less prone to blind spot bias and are more competent in the skills studied. MethodsThis descriptive, cross-sectional study was conducted from April 2022 to May 2023. Thirty-two psychiatrists were recruited from several French hospitals, along with 20 ‘ethicists’. All participants completed the following self-administered questionnaires: the Bias Blind Spot Scale (French version), the Euro Moral Case Deliberation (Euro-MCD, ethical skills), Bandura's self-efficacy scale (self and collective efficacy), the Freiburg Mindfulness Inventory (mindfulness), and the Schwartz Value Survey (values). ResultsThe ratio of self/other blind spot bias was lower in the ‘ethicists’ group than the psychiatrists’ group (P<0.001). Euro-MCD scores were higher (P<0.001) and psychiatrists tended to have a better sense of collective efficacy (P=0.08). No significant difference was found between the two groups for total mindfulness or presence and acceptance sub-scores. Scores for hedonic value were higher among psychiatrists (P<0.02). ConclusionThe results do not confirm our hypotheses. There is less evidence of blind spot bias among our sample of ‘ethicists’, who are experts in group deliberation, than psychiatrists, who have experience of the patient/physician dyad. The ethical skills that are developed within the formal deliberation process seem particularly interesting in terms of reducing cognitive bias.

Decision dynamics and landscape approach by individuals and collective groups: the case of Japanese rural land use in agricultural, forest, and boundary areas in an era of population decline

Establishing a decision-making scheme for multiple land use types in a shrinking society is an urgent and persistent task. In particular, residents and governments face the complexity of cross-sectoral visions over concerned landscapes. This study examined the preference of residents for future visions and the effects of group deliberation in a Japanese rural community to enhance the current understanding of decision and learning processes on cross-sectoral land use types (i.e., agricultural land, forest land, and boundary areas) across ages and groups. This study posed the following questions: (i) what types of interactions do residents find among options on cross-sectoral land uses? and (ii) how do group deliberation influence the individual preferences of residents for land use? In the survey, the residents disclosed their preferences for the prepared visions over three land use types. The survey also enabled the comparison of choices before and after group deliberation. Regarding question (i), the residents’ choice prior to the deliberation demonstrated thematic coherence through land use categories. Smart technology and potential productivity improvement were selected across agriculture and forests. However, regarding question (ii), thematic coherence decreased after the group deliberation. The residents could communicate disagreements during group deliberation and consider the issues relatively independently over land use types, which separately views agriculture lands and forests. The choice after group discussions became increasingly embedded and influenced by residential areas and land ownership of residents. These findings provide unique insights into the group discussion on how individual decisions are affected. In summary, the preferences of cross-sectoral land use types did not converge but tended to diverge.

Qualitative insights into the effectiveness of a targeted nursing research support program: Understanding and experiences of support recipients and providers

AimThe aims of this study were to examine the effectiveness of a targeted nursing research support program for clinical nurses. BackgroundNursing research capacity is increasingly essential to clinical nurses and currently relatively low. Therefore, effective and systematic nursing research training programs are urgently needed to improve the scientific research abilities of nurses. MethodsQualitative research was conducted to investigate the effectiveness of a targeted nursing research support program. The program was formulated by considering the research training requirements of nurses and standard nursing research procedures, through literature review and group deliberations. The program was implemented for 973 nurses using a “plan–action–observation–reflection” learning cycle. The research outcomes achieved by nurses were evaluated and thematic analysis conducted to assess the perspectives of nurses and teachers regarding the research support program. ResultsNurses participating in the targeted nursing research support program collectively accomplished 195 research proposals and authored 332 original research articles. Nurses shared their rich experience as “understanding my needs and achieving my potential”, including: (1) systematic procedures and coherence; (2) easy to learn, easy to use; (3) a sense of belonging and mutual support; (4) self-confidence growth; and (5) high expectations. Further, the experiences of teachers were summarized as “helping others is helping myself”, including: (1) teaching is learning; (2) the happiness of being needed; and (3) the importance of scientific teaching. ConclusionThis study evaluated the experiences of nurses and educators involved in a targeted nursing research support program and assessed its preliminary effectiveness. The findings revealed that the program, grounded in scientific and systematic research principles, was beneficial to both nurses and teachers. Based on our findings, we recommend that nursing educators should prioritize comprehensive, practice-integrated research training programs and create supportive environments, to effectively enhance the research capacity of nurses.

“Justly and Without Bias”

This paper investigates the possibility that one purpose of consultation is the mitigation of cognitive biases in individual participantsand in the group as a whole. After exploring the nature of cognitive biases through the lens of evolutionary psychology, the paper surveys existing research on effective methods of “debiasing” individuals. This research suggests that the most effective environment for mitigating bias is a deliberative group, in which individual participants may be asked to justify their reasoning in a social environment of diverse perspectives. Bias mitigation diminishes over time, requiring repeated exposure to the debiasing environment. This model for debiasing strongly resonates with Bahá’í consultation, a conclusion that can enrich Assemblies’ and other consultinggroups’ perspectives on, and expectations of, consultation.

Anchoring as a Structural Bias of Deliberation

AbstractWe study the anchoring effect in a computational model of group deliberation on preference rankings. Anchoring is a form of path-dependence through which the opinions of those who speak early have a stronger influence on the outcome of deliberation than the opinions of those who speak later. We show that anchoring can occur even among fully rational agents. We then compare the respective effects of anchoring and three other determinants of the deliberative outcome: the relative weight or social influence of the speakers, the popularity of a given speaker’s opinion, and the homogeneity of the group. We find that, on average, anchoring has the strongest effect among these. We finally show that anchoring is often correlated with increases in proximity to single-plateauedness. We conclude that anchoring can constitute a structural bias that might hinder some of the otherwise positive effects of group deliberation.

Proxy Preferences and the Values of Children's Health States.

To assign values to the health states of children, some health economists have suggested relying on the 'proxy' preferences among the health states of children expressed by a random sample of the adult population. These preferences have been elicited in several ways, with respondents sometimes asked to express their (adult) preferences among the health states of children, and sometimes asked to imagine themselves as children and to express what they think their preferences would be. This essay discusses three grounds for eliciting the preferences of a random sample of adults that have been suggested as ways to assign values to the health states in the EQ-5D-Y, and criticizes the first two: (1) the evidential ground: the preferences of the population sample are good evidence of how good or bad the health states of children are; (2) the 'taxpayer' ground: the adult population has the authority to assign values to health states, therefore their preferences are determinative; and (3) the pragmatic grounds: surveying is straightforward and shifts the responsibility from health economists to the population. I argue that instead of surveying a random sample of the population, health economists should rely on deliberative groups that include older children, experts on children's health and development, as well as members of the population at large. These groups should engage with the reasons that lie behind preferences among health states.

Ambivalences of Technological Literacy and Cognitive Growth: A Learning Classroom Dynamic in the Disruptive Era

This research aims to reveal the ambivalence of using technology in the learning process. The research methodology employed is qualitative. The research approach employed in this study is a case study. The data utilised in this study comprises statements extracted from a Forum Group Discussion (FGD). The data source consists of the FGD transcripts. Two ELT lecturers, one concentrates in English Writing and another in Translation, were put in sharing ideas session during the Focus Group Discussion (FGD). The data collection technique employed in this research is focus group deliberations. The data gathering method utilised in this study is the Single Focus Group (FGD). The analytical approach employed in this work is thematic analysis. From the FGDs conducted, two axes of answers were found. The first is related to how the utilisation of technology, such as AI (Artificial Intelligence), is a must. There are two points related to the necessity of technology integration in learning: 1) Improving Access and Learning Opportunities and 2) Increasing Student Motivation and Engagement. The second, on the other hand, the utilisation of technology in the learning process also reflects a very ironic phenomenon. There are two main reasons why the utilisation of technology in learning is very dilemmatic at the moment: 1) Balanced Potential Benefits and Risks, and 2) Educational Paradigm Shift. The details of these findings can be traced from the elaboration of key points from the excerpts of discussions with lecturers when sharing teaching experiences in this disruptive era.

Collective Wisdom and the Fermi Method: Improving the Accuracy of Deliberative Groups

Collective Wisdom and the Fermi Method: Improving the Accuracy of Deliberative Groups

Communicating CC(U)S and UGS technologies: The case of deliberative group interviews in North-Western Poland

The paper presents the results of deliberative group interviews with the representatives of local communities in West Pomeranian Province (Poland) on the opportunities and challenges related to the theoretical location of combined CO2 and natural gas underground storage (CC(U)S and UGS) in their immediate vicinity. The research has been conducted within AGaStor Project that focuses on innovative usage of captured CO2 as a cushion for stored natural gas. Therefore, there are two possible benefits: active limitation of CO2 emissions (Poland is and will be dependent on conventional energy sources) and reduction of gas storage costs (by using waste CO2 as so-called cushion gas). However, previous experiences in these types of projects show the key role of local public opinion in the final shape of the investment.CC(U)S projects are usually located in less developed, less populated areas, with smaller, more rural local communities facing serious socio-economic challenges. This is the case presented in this paper. In such communities, socially expressed uncertainty and dissent are mechanisms of emancipation and articulating community's subjectivity, and a strategy for articulating fears of distrust and inadequacy of empowerment. The presented results of the research lead to four significant mechanisms of building trust for CC(U)S and UGS in West Pomeranian Province communities: a) communicating safety, b) communicating the lack of nuisances, c) communicating the benefits and d) transparent and competent knowledge transfer.

Competence or Experience

Competence or Experience

An Indonesian Secondary English Teacher’ Strategies to Accommodate Computer Mediated Communication for Group Discussions: An Interview Study

Numerous scholarly inquiries have been dedicated to exploring the advantageous consequences of implementing computer-mediated communication (CMC) for the purpose of orchestrating online group discussions. However, there is a paucity of empirical evidence elucidating how educators adapt to the use of CMC in the context of online group deliberations. The present research delves into the experiences of an English teacher in an Indonesian secondary school who employed CMC as a pedagogical tool. Grounded within the qualitative research paradigm, this study conducted unstructured in-depth interviews and subsequently subjected the interview transcripts to thematic analysis. The resultant findings shed light on several pivotal aspects of the teacher's role in CMC-enhanced online group discussions, which encompass the activation of students' creative faculties, the fostering of students' self-regulation of learning, the cultivation of a conducive learning environment, the advocacy for technological integration, and the effective management of assessment procedures in the online group discussion setting. Conclusively, this research posits that the incorporation of computer-mediated communication in online group discussions bears substantial educational advantages, as it affords students with alternative avenues to harness their skills and technology, thereby facilitating interpersonal exchanges among students and between students and educators.

From individual decisions to team decisions under risk: evidence from a field experiment

This paper investigates how individual decisions are aggregated to team decisions under risk in field experiment with a sample of Ethiopian farmers and the team decision-making process behind this aggregation. In an experiment structurally similar to the Gneezy-Potters' risk game, subjects make decisions first individually and then jointly with a random anonymous partner. In the aggregate, teams make higher allocation into the risky asset compared to the average of individual decisions. However, teams are neither polarizing nor simply averaging individual preferences. Instead, team decisions are consistent with an outcome of group bargaining and deliberation process based on intensity of individual preferences. But more risk-taking subjects have a stronger influence on team decisions. This influence of more risk-taking members on the group decision is leveraged by their better education levels. Analysis of the team decision-making process reveals several interesting insights. About 54% of the teams do not reach immediate agreement with initial allocations. Both less risk-taking and more risk-taking team members are equally likely to disagree with initial allocations but for different reasons. Teams that disagreed with initial allocations reached final team decisions significantly different from disagreed initial allocations. Less risk-taking subjects are more willing to concede to allocations proposed by more risk-taking subjects to reach at an agreed team decision. Demanding messages in group communications have a stronger effect on outcomes of group decision-making. Finally, teams with greater differences in willingness to take risk among members are more likely to disagree with initial team allocations, take more rounds of deliberations to come to a decision, and make choices further away from the average of individual decisions. Our results permit to better characterize the process of group decision-making beyond differences between individual and group decisions.

What do other people think he deserves? Social influence on utilization of mitigating information regarding a violent offender's unfortunate life history.

The blameworthiness of an offender is often discussed in groups. Yet, the research literature overwhelmingly examines individuals assessing blameworthiness in isolation. To address this gap in the literature, the present study examines group deliberations about blameworthiness, with a particular focus on how group deliberations impact utilization of mitigating information about an offender's unfortunate life history. Participants from introductory psychology courses at a U.S. university were placed in groups of two or three and each group also included a confederate who followed a script. Groups were randomly assigned to one of four conditions. In one condition (deed only), groups learned only about the offender's heinous crimes. In the three remaining conditions, participants also received a historicist narrative regarding how the offender's unfortunate history deformed his moral character. These conditions differed in terms of the confederate's arguments: Neutral arguments, arguments to ignore the narrative, or arguments to give great weight to the narrative. Results showed that the historicist narrative was particularly effective at reducing outrage and increasing compassion when the confederate argued for its utilization. The reduction in outrage mediated a reduction in spiteful punitiveness toward the offender. Interestingly, the confederate who urged fellow deliberators to ignore the historicist narrative had no impact on outrage or compassion. We also examined mediation of the impact of historicist narratives on outrage and compassion. We found that when the confederate remained neutral the impact of historicist narratives on outrage and compassion was mediated via diminished perceptions of the offender's control of self-formation. This mirrors what is typically found in prior work focused on individual judgments. In contrast, when the confederate argued that great weight should be given to the narrative, reductions in outrage were mediated via diminished perceptions of offender freedom of action. This pattern of mediation is not typically found but has been found in one previous study where participants received social encouragement to mitigate blame. Results are discussed in terms of how social influence might alter the inferences draw from historicist narratives. Suggestions for future research on social influence in the context of blame are presented.

Whakawhanaungatanga-Building trust and connections: A qualitative study indigenous Māori patients and whānau (extended family network) hospital experiences.

Investigated the experiences of Māori (the Indigenous peoples of Aotearoa, New Zealand) patients and whānau (extended family network) engaging with acute hospital inpatient services and their priorities for a Māori-centred model of relational care. A qualitative Māori-centred research design using a Thought Space Wānanga (learning through in-depth group discussion, deliberation and consideration) approach. Two wānanga were conducted between May 2022 and June 2022, with 13 Māori patients who had been acutely hospitalized within the past 12 months and their whānau members. The first wānanga utilized storytelling and journey mapping to collect data. The second wānanga refined the initial themes. Wānanga were audio-recorded and then inductively coded and developed into themes. Thirteen patients and whānau attended the first wānanga, while 10 patients and whānau participated in the second wānanga). Four themes were developed: (1) Whakawhanaungatanga (establishing connections and relationships), (2) Whakamana (uplifting the status and esteem of Māori), (3) Whakawhitiwhiti kōrero (the importance of communicating, discussing and deliberating) and (4) Kotahitanga (working together with purpose) all provide insights into the importance of effectively engaging and connecting with Māori patients and whānau when acutely hospitalized. The experiences and priorities of Māori patients and whānau affirm the international literature, suggesting that Indigenous relational concepts are critical to building relationships, connections and trust. Despite existing healthcare models for working with Indigenous peoples, their poor application contributes to sub-optimal healthcare experiences at all points of their healthcare journey. A relational mode of practice focused on engagement and forming connections better meets the needs of Indigenous peoples engaging with inpatient health services. Despite the existence of Indigenous models of care, Indigenous peoples consistently report a lack of engagement and connection when accessing inpatient health services. Without establishing relationships, applying models of care is challenging. What problem did the study address? Internationally, healthcare systems are consistently ill-equipped to deliver culturally safe care for Indigenous and marginalized peoples, evident in ongoing health inequities. Like other reports of Indigenous experiences of health services, Māori express dissatisfaction with care delivery in an acute inpatient setting. This study investigated Māori patients and whānau experiences engaging with acute hospital inpatient services and their priorities for a Māori-centred model of relational care. What were the main findings? Māori patients and whānau recounted negative experiences with healthcare professionals lacking effective relationships and trust. Satisfaction occurred when engagement with health care professionals resembled Indigenous cultural rituals of encounter that considered their holistic, collective and dynamic worldviews. Previous models of relational care, while helpful, are not Indigenous and so do not address their needs, such as engagement as a mode of practice (how) to achieve this. Where and on whom will the research have an impact? This research impacts Indigenous peoples' health outcomes, particularly Māori, and nurses and clinicians working and interacting within acute inpatient and other hospital settings. Indigenous research methods support co-constructing knowledge for translation into practical outcomes through transformational practices, policies and theory development. We used the Consolidated Criteria for Strengthening the Reporting of Health Research Involving Indigenous Peoples (CONSIDER) statement (see File S2-CONSIDER Checklist) and the Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines (see File S3-COREQ Checklist). Māori patients and their whānau interviewed about their experiences were involved in data interpretation.

Can moral case deliberation in research groups help to navigate research integrity dilemmas? A pilot study

There is an increased focus on fostering integrity in research by through creating an open culture where research integrity dilemmas can be discussed. We describe a pilot intervention study that used Moral Case Deliberation (MCD), a method that originated in clinical ethics support, to discuss research integrity dilemmas with researchers. Our research question was: can moral case deliberation in research groups help to navigate research integrity dilemmas? We performed 10 MCDs with 19 researchers who worked in three different research groups from three different disciplinary fields at a university in the Netherlands. We analyzed the dilemmas and values discussed, sent out a survey questionnaire to assess self-perceived moral competencies, and conducted in-depth interviews. We found research integrity dilemmas pertained to authorship disputes, supervision of junior co-workers, and questionable handling of data. Participants perceived the majority of moral competencies to a higher degree during the MCD when compared to perceiving them in daily practice afterward. Interviewees told us that they felt most comfortable discussing dilemmas among peers with whom they were not closely affiliated. We conclude that MCD sessions could be relevant in navigating research integrity dilemmas, but that revisions to ensure commitment and safety are required.

DeliData: A Dataset for Deliberation in Multi-party Problem Solving

Group deliberation enables people to collaborate and solve problems, however, it is understudied due to a lack of resources. To this end, we introduce the first publicly available dataset containing collaborative conversations on solving a well-established cognitive task, consisting of 500 group dialogues and 14k utterances. In 64% of these conversations, the group members are able to find a better solution than they had identified individually, and in 43.8% of the groups who had a correct answer as their final solution, none of the participants had solved the task correctly by themselves. Furthermore, we propose a novel annotation schema that captures deliberation cues and release all 14k utterances annotated with it. Finally, we use the proposed dataset to develop and evaluate two methods for generating deliberation utterances. The data collection platform, dataset and annotated corpus are publicly available at https://delibot.xyz.

Patient Values and Preferences for Managing Acute Dental Pain Elicited through Online Deliberation.

Patient values and preferences (PVP) are among multiple sources of information panelists synthesize when developing clinical practice guidelines (CPG). Patient and public involvement (PPI) can be critical for learning PVP; however, the methodology for engaging patients in CPG development is lacking. Deliberative engagement is effective for obtaining public views on complex topics that require people to consider ethics, values, and competing perspectives. Elicit comprehensive understanding of PVP concerning oral analgesics for managing acute dental pain consecutive to toothache and simple and surgical dental extractions, with consideration of associated outcomes, both desirable and undesirable. Multistage engagement involving 2 electronic surveys and a 90-min online small group deliberative engagement. Adults who have experienced acute dental pain deliberated about 3 hypothetical scenarios stratified according to expected pain intensity, completed a postdeliberation survey, and validated a PVP statement developed by researchers based on review of qualitative data from deliberations and quantitative data from surveys. Participants affirmed the PVP statement reflected their small group deliberations and their individual views. Most indicated that pain relief is critical to deciding which pain relief medicine they would want regardless of expected pain level. Most also identify as critical concerns about substance abuse or misuse, although many believe it unlikely that they will experience these outcomes over the brief prescription timeframe for acute dental pain. Participants identified agency in decision-making, consultation including "better communication" of options, and treatment actions tailored to life circumstances as key values. Participants preferred nonprescription and nonopioid pain relief options. As expected pain levels increased, more participants expressed willingness to accept opioids, but more also mentioned rescue analgesia as a third outcome critical to decision-making. Online deliberative method provided opportunities for obtaining informed perspectives. Guideline developers and policymakers may find online deliberations useful for eliciting PVP related to health outcomes. Study results informed the US Food and Drug Administration-funded clinical practice guideline on the management of acute dental pain. Findings may be a resource for clinicians in decision-making conversations with patients regarding expectations for pain relief and positive and negative outcomes of differing pain relief medications. Further research should pursue applicability of online deliberative engagement as a method to elicit patient values and preferences.

Biased Processing and Opinion Polarization: Experimental Refinement of Argument CommunicationTheory in the Context of the Energy Debate

We combine empirical experimental research on biased argument processing with a computational theory of group deliberation to overcome the micro–macro problem of sociology and to clarify the role of biased processing in debates around energy. We integrate biased processing into the framework of argument communication theory in which agents exchange arguments about a certain topic and adapt opinions accordingly. Our derived mathematical model fits significantly better to the experimentally observed attitude changes than the neutral argument processing assumption made in previous models. Our approach provides new insight into the relationship between biased processing and opinion polarization. Our analysis reveals a sharp qualitative transition from attitude moderation to polarization at the individual level. At the collective level, we find that weak biased processing significantly accelerates group decision processes, whereas strong biased processing leads to a meta-stable conflictual state of bi-polarization that becomes persistent as the bias increases.

A Dynamic Dyadic Systems Perspective on Interpersonal Conversation

ABSTRACT Conversations between people are where, among other things, stressors are amplified and attenuated, conflicts are entrenched and resolved, and goals are advanced and thwarted. What happens in dyads’ back-and-forth exchanges to produce such consequential and varied outcomes? Although numerous theories in communication and in social psychology address this question, empirical tests of these theories often operationalize conversational behavior using either discrete messages or overall features of the conversation. Dynamic systems theories and methods provide opportunities to examine the interdependency, self-stabilization, and self-organization processes that manifest in conversations over time. The dynamic dyadic systems perspective exemplified by the articles in this special issue (a) focuses inquiry on the turn-to-turn, asynchronous exchange of messages between two partners, (b) emphasizes behavioral patterns within and the structural and temporal organization of conversations, and (c) adapts techniques used in analysis of intensive longitudinal data to identify and operationalize those dynamic patterns. As an introduction to the special issue, this paper describes a dynamic dyadic systems perspective on conversation and discusses directions for future research, such as applications to human-computer interaction, family communication patterns, health care interventions, and group deliberation.