University of Michigan School of Information faculty and PhD students are creating and sharing knowledge that helps build a better world. Here are some of their recent publications.

---

Wisdom of Two Crowds: Misinformation Moderation on Reddit and How to Improve this Process—A Case Study of COVID-19

Association for Computing Machinery, April 2023

Lia Bozarth, Jane Im, Christopher Quarles, Ceren Budak

Past work has explored various ways for online platforms to leverage crowd wisdom for misinformation detection and moderation. Yet, platforms often relegate governance to their communities, and limited research has been done from the perspective of these communities and their moderators. How is misinformation currently moderated in online communities that are heavily self-governed? What role does the crowd play in this process, and how can this process be improved? In this study, we answer these questions through semistructured interviews with Reddit moderators. We focus on a case study of COVID-19 misinformation. First, our analysis identifies a general moderation workflow model encompassing various processes participants use for handling COVID-19 misinformation. Further, we show that the moderation workflow revolves around three elements: content facticity, user intent, and perceived harm. Next, our interviews reveal that Reddit moderators rely on two types of crowd wisdom for misinformation detection. Almost all participants are heavily reliant on reports from crowds of ordinary users to identify potential misinformation. A second crowd–participants’ own moderation teams and expert moderators of other communities–provide support when participants encounter difficult, ambiguous cases. Finally, we use design probes to better understand how different types of crowd signals—from ordinary users and moderators—readily available on Reddit can assist moderators with identifying misinformation. We observe that nearly half of all participants preferred these cues over labels from expert fact-checkers because these cues can help them discern user intent. Additionally, a quarter of the participants distrust professional fact-checkers, raising important concerns about misinformation moderation. 

---

Direct, Orienting, and Scenic Paths: How Users Navigate Search in a Research Data Archive

Association for Computing Machinery, March 2023

Sara Lafia, A.J. Million, Libby Hemphill

Social scientists increasingly share data so others can evaluate, replicate, and extend their research. To understand the process of data discovery as a precursor to data use, we study prospective users’ interactions with archived data. We gathered data for 98,000 user sessions initiated at a large social science data archive, the Inter-university Consortium for Political and Social Research (ICPSR). Our data reflect four years (2012-16) of users’ interactions with archival resources, including a data catalog, study-level metadata, variables, and publications that cite nearly 10,000 datasets. We constructed a network of user interactions linking website landing (e.g., site entrances) to exit pages, from which we identified three types of paths that users take through the research data archive: direct, orienting, and scenic. We also interpreted points of failure (e.g., drop-offs) and recurring behaviors (e.g., sensemaking) that support or impede data discovery along search paths. We articulate strategies that users adopt as they navigate data search and suggest ways to enhance the accessibility of data, metadata, and the systems that organize each.

---

Being Creative Within (or Outside) the Box: Bridging Occupational Identity Gaps

Management Communication Quarterly, March 2023

Stephanie L. Dailey, Casey S. Pierce, Diane E. Bailey, Paul M. Leonardi, Bonnie Nardi

This study advances organizational communication scholarship by introducing the notion of an occupational identity gap as a misalignment among the personal, relational, communal, and enacted frames of identity. Despite knowledge that occupational identity gaps exist, scholars know little about how people manage them. Interviews with 31 graphic designers explain how occupational identity gaps were forged by personal frames (e.g., “I am a creative person”) that contradicted enacted (e.g., “I do boring template work”) and relational frames (e.g., “It’s the client’s decision which [design] he or she will like”). Workers managed this misalignment by employing two novel strategies—reappraising and repositioning—that bridged personal-enacted and personal-relational occupational identity gaps. Our analysis contributes to scholarship by a) theorizing these two occupational identity gap bridging strategies, (b) extending CTI research, and (c) offering a novel conceptualization of occupational identity.

---

“I Did Watch ‘The Handmaid’s Tale’”: Threat Modeling Privacy Post-Roe in the United States

Association for Computing Machinery, March 2023

Nora Mcdonald, Nazanin Andalibi

Now that the protections of Roe v. Wade are no longer available throughout the United States, the free flow of personal data can be used by legal authorities to provide evidence of felony. However, we know little about how impacted individuals approach their reproductive privacy in this new landscape. We conducted interviews with 15 individuals who may get/were pregnant to address this gap. While nearly all reported deleting period tracking apps, they were not willing to go much further, even while acknowledging the risks of generating data. Quite a few considered a more inhospitable, Handmaid's Tale like climate in which their medical history and movements would put them in legal peril but felt that, by definition, this reality was insuperable, and also that they were not the target—the notion that privileged location, stage of life did not make them the focus of government or vigilante efforts. We also found that certain individuals (often younger and/or with reproductive risks) were more attuned to the need to modify their technology or equipped to employ high and low-tech strategies. Using an intersectional lens, we discuss implications for media advocacy and propose privacy intermediation to frame our thinking about reproductive privacy.

---

viz2viz: Prompt-driven stylized visualization generation using a diffusion model

arXiv, April 2023

Jiaqi Wu, John Joon Young Chung, Eytan Adar

Creating stylized visualization requires going beyond the limited, abstract, geometric marks produced by most tools. Rather, the designer builds stylized idioms where the marks are both transformed (e.g., photographs of candles instead of bars) and also synthesized into a 'scene' that pushes the boundaries of traditional visualizations. To support this, we introduce viz2viz, a system for transforming visualizations with a textual prompt to a stylized form. The system follows a high-level recipe that leverages various generative methods to produce new visualizations that retain the properties of the original dataset. While the base recipe is consistent across many visualization types, we demonstrate how it can be specifically adapted to the creation of different visualization types (bar charts, area charts, pie charts, and network visualizations). Our approach introduces techniques for using different prompts for different marks (i.e., each bar can be something completely different) while still retaining image "coherence." We conclude with an evaluation of the approach and discussion on extensions and limitations.

---

A Bibliometric Review of Large Language Models Research from 2017 to 2023

arXiv, April 2023

Lizhou Fan, Lingyao Li, Zihui Ma, Sanggyu Lee, Huizi Yu, Libby Hemphill

Large language models (LLMs) are a class of language models that have demonstrated outstanding performance across a range of natural language processing (NLP) tasks and have become a highly sought-after research area, because of their ability to generate human-like language and their potential to revolutionize science and technology. In this study, we conduct bibliometric and discourse analyses of scholarly literature on LLMs. Synthesizing over 5,000 publications, this paper serves as a roadmap for researchers, practitioners, and policymakers to navigate the current landscape of LLMs research. We present the research trends from 2017 to early 2023, identifying patterns in research paradigms and collaborations. We start with analyzing the core algorithm developments and NLP tasks that are fundamental in LLMs research. We then investigate the applications of LLMs in various fields and domains including medicine, engineering, social science, and humanities. Our review also reveals the dynamic, fast-paced evolution of LLMs research. Overall, this paper offers valuable insights into the current state, impact, and potential of LLMs research and its applications.

---

Perceptions of telehealth among older U.S. adults during the COVID-19 pandemic: A national survey

Journal of Telemedicine and Telecare, April 2023

Kathleen Y Li, Liz B Marquis, Preeti N Malani, Erica Solway, Matthias Kirch, Dianne Singer, Jeffrey T Kullgren, Melissa A Plegue, Lorraine R Buis

Introduction: COVID-19 necessitated a shift from in-person to virtual care for all patients, particularly older adults. It is unknown how older individuals’ views of telehealth changed during this time and how this may affect their future use of telehealth services.

Methods: We used data from a cross-sectional online survey of a nationally representative sample of 2074 U.S. adults ages 50–80 who were participants in the National Poll on Healthy Aging. We performed a descriptive and multivariable analysis of individuals’ perspectives on past and future telehealth visits, sociodemographics, and health status.

Results: Before March 2020, 5.8% of respondents had used telehealth, compared to 32.0% by June 2020. Of telehealth users, 36.1% indicated their most recent telehealth visit used audio-only (i.e., without video) technology. In multivariable analysis, those who never used video technology compared to those who were “very comfortable” (average marginal effect (AME) 49%, 95% CI: 36–63), identified as Hispanic (AME 19% vs White, non-Hispanic, 95% CI: 5–32), or were female (AME 9%, 95% CI: 1–17) were more likely to report audio-only use. Concerns remained about the inability to conduct physical exams (75%) and telehealth quality of care (67%), though most (64%) older adults indicated an interest in future telehealth visits.

Discussion: Telehealth use increased substantially among older U.S. adults during the early months of the COVID-19 pandemic; however, many reported using audio-only telehealth, an important consideration for policymakers and providers. Addressing older adults’ concerns about and barriers to telehealth visits is needed to ensure telehealth does not exacerbate disparities in their care.

---

Do "bad" citations have "good" effects?

arXiv, April 2023

Honglin Bao, Misha Teplitskiy

The scientific community discourages authors of research papers from citing papers that did not influence them. Such "rhetorical" citations are assumed to degrade the literature and incentives for good work. While a world where authors cite only substantively appears attractive, we argue that mandating substantive citing may have underappreciated consequences on the allocation of attention and dynamism in scientific literatures. We develop a novel agent-based model in which agents cite substantively and rhetorically. Agents first select papers to read based on their expected quality, read them and observe their actual quality, become influenced by those that are sufficiently good, and substantively cite them. Next, agents fill any remaining slots in the reference lists by (rhetorically) citing papers that support their narrative, regardless of whether they were actually influential. By turning rhetorical citing on-and-off, we find that rhetorical citing increases the correlation between quality and citations, increases citation churn, and reduces citation inequality. This occurs because rhetorical citing redistributes some citations from a stable set of elite-quality papers to a more dynamic set with high-to-moderate quality and high rhetorical value. Increasing the size of reference lists, often seen as an undesirable trend, amplifies the effects. In sum, rhetorical citing helps deconcentrate attention and makes it easier to displace incumbent ideas, so whether it is indeed undesirable depends on the metrics used to judge desirability.

---

Home Office: Working from a Private Place

Business and Information Systems Engineering, April 2023

Claudia Mueller, Alexander Maedche, Gerhard Schwabe, Mark Ackerman, Volker Wulf 

The labor world is currently undergoing a major transformation. New forms of digitalization are giving rise to new socio-technical working arrangements. The COVID-19 pandemic has acted as a burning glass, favoring new models of working from home suddenly and without notice.

Thus, more research on the “home office” phenomenon and resulting scientifically grounded knowledge is urgently required that helps practitioners to advance future arrangements of good (hybrid) working conditions. Even beyond COVID-19, many office employees will continue to work, at least partly, from home. Although the degree of virtualization will continue to increase, a combination of physical and virtual presence can be expected. Hybrid forms of working will become part of our future private and professional life.

Working from home goes hand in hand with great potential, but at the same time also with risks. Recent experiences with home schooling and home office have triggered intensive discussions on potential positive and negative outcomes of co-locating work with private and schooling activities. There is a need to design digital technologies appropriately to support individuals, groups, and organizations in ways that increase productivity and wellbeing towards better work and life. Thus, pursuing a socio-technical paradigm for understanding and designing for the home office is essential.

This makes it all the more important to now closely analyze the experiences and lessons learned during the COVID-19 phase and to set the research agenda and development of working practices from home for the future.

---

Bridging Nations: Quantifying the Role of Multilinguals in Communication on Social Media

arXiv, April 2023

Julia Mendelsohn, Sayan Ghosh, David Jurgens, Ceren Budak

Social media enables the rapid spread of many kinds of information, from memes to social movements. However, little is known about how information crosses linguistic boundaries. We apply causal inference techniques on the European Twitter network to quantify multilingual users’ structural role and communication influence in cross-lingual information exchange. Overall, multilinguals play an essential role; posting in multiple languages increases betweenness centrality by 13%, and having a multilingual network neighbor increases monolinguals’ odds of sharing domains and hashtags from another language 16-fold and 4-fold, respectively. We further show that multilinguals have a greater impact on diffusing information less accessible to their monolingual compatriots, such as information from far-away countries and content about regional politics, nascent social movements, and job opportunities. By highlighting information exchange across borders, this work sheds light on a crucial component of how information and ideas spread around the world.

---

Catch, Corroborate and Communicate: The Contours of Professional fact-Checking In India

CSCW, June 2023

Arshia Arya, Soham De, Saloni Dash, Joyojeet Pal

The work of professional fact-checking involves three broad processes - selection of the issues that need to be fact-checked, verifying a story that is flagged as being a falsehood, and finally disseminating a fact-checked version of a story, each of which have unique challenges. We conducted in-depth interviews on their professional practice with fact-checkers of five major organizations that work in debunking stories daily, alongside conducted a large-scale analysis of the archives of fact-checked articles and studied the social media footprint of fact-checkers from these organizations. Our findings show that the outcomes of work with each of the three processes influence what stories get debunked by organizations, how much public impact they have, and the depth of debunking in each story, all of which in turn play a role in the scope and breadth of individual organizations’ impact. Our findings have implications from both the supply side of how organizations can place themselves within the fact-checking ecosystem, and from the demand side of how consumers can consider the work and positioning of their preferred fact-checkers.

---

Data Work on Frontline Care Workers: Practices, Problems, and Opportunities in the Context of Data-Driven Long-Term Care

Proceedings of the ACM on Human-Computer Interaction, April 2023

Yuling Sun, Xiaojuan Ma, Silvia Lindtner, Liang He

Using data and data technologies to support healthcare has drawn significant attention recently. While CSCW and HCI have largely celebrated the tremendous promise of 'data-driven healthcare' in reforming the healthcare sector, this paper reveals 'labor-driven reality' of this promised data-driven future. Drawing from a qualitative study in a real-world data-driven long-term care (LTC) facility in China, we demonstrate how data-driven technologies work in practice, and especially how frontline workers, as the crux of this data-driven configuration, conduct a tremendous amount of "data work" to make data-drivenness work. This data work, we argue, goes beyond the "clerical work" and functions as a labor of maintenance, articulation, and repair, that both guarantees data technologies' functionalities and acts as an interface between stakeholders. We conclude by discussing the practices, problems and opportunities of this data work in a broader socio-cultural context.

---

Calibrated Uncertainty: How In-Vitro Fertilization Patients Use Information to Regulate Emotion

Proceedings of the ACM on Human Computer Interaction, April 2023

Melody S.Y.Ku, Mark Ackerman 

The acquisition of information and its use in decision making and coping by people with health concerns have garnered much attention in CSCW. This study investigated patients' information behavior during a critical treatment, in vitro fertilization (IVF). Based on in-depth interviews with 29 IVF patients, this study uncovered several underlying drivers and mechanisms accounting for patients' information behavior. Their behavior is shown to be driven by coping concerns - specifically, dealing with the unpredictability of the treatment outcome, overcoming feelings of powerlessness and the sense of being out of control, and managing difficult emotionality. These factors shape patients' information needs and drive their behaviors in seemingly irrational but ultimately logical and adaptive ways. In contrast to the conventional wisdom that patients typically seek information that can help them fill knowledge gaps to resolve treatment uncertainty and foster their positive emotions, we discovered that in response to the desire to control their perceptions of irresolvable uncertainty and the difficult emotional needs of the moment, IVF patients frequently used ?calibrated uncertainty" - a psychological mechanism or state driven by the simultaneous seeking of varying levels and contradictory valences of certainty - to actively conduct targeted searches and actively use the information sought. This behavior has not always been understood by the IVF clinicians, who have assumed that information was primarily for knowledge transfer. This study shows that coping, emotion regulation, and information-seeking are inextricably bound together in the patient experience, and that this intertwining must be considered in the clinical setting for physician-patient communication and for patient-facing information. The findings have several valuable design implications for improved health informatics technology and service delivery systems.

---

Racial Capitalism Online: Navigating Race Among Creative Professionals of Color

Proceedings of the ACM on Human-Computer Interaction, April 2023

Jaleesa Rosario Turner, Julie Hui

Racial capitalism, which describes how people in power extract value from the racial identity of others, has been a constant driver in the creative industries. As social media becomes the widespread avenue for sharing creative works and building professional reputation, the effects of racial capitalism become further amplified online. In fact, many creative professionals of color find that the biases they face offline are only replicated, and sometimes magnified, on digital platforms. In this study, we interview creative professionals of color who heavily rely on digital platforms to promote their work in order to understand how racial capitalism shapes their experience and performance of race online. Creative professionals describe seeing their work appropriated and shared for little to no recognition, while at the same time, feeling pressured to present themselves in a palatable way in order to meet the expectations of a dominant consumer audience. Creatives also worked within these capitalistic expectations by building communities with similar others in order to exchange advice, serve as role models, and share resources. Our data uncover how expectations on social media, fueled by racial biases, burden creative professionals of color, thus informing alternative futures that could compensate their work more equitably and build more inclusive spaces for creative professional growth.

---

Putting a Teaspoon of Programming into Other Subjects

Communications of the ACM, May 2023 

Mark Guzdial, Emma Dodoo, Bahare Naimpour, Tamara Nelson-Fromm, Aadarsh Padiyath

Using teaspoon languages to integrate programming across myriad academic disciplines.