Quantifying the self with others

Abstract

Self-trackers collect personal data for many reasons, including generating insight about their bodies, habits, productivity, and wellbeing. Self-tracking may expose intimate facets of daily life, raising important questions about surveillance, privacy, and data ownership. In this study, we investigated an online community of self-trackers and their weekly “show-and-tell” presentations through observations of their meetings and interviews with members. Making sense of their personal data in community with others involved practical and philosophical difficulties that participants navigated by integrating competing priorities for their interactions in specific communication moves and by transcending interactional difficulties through a shared focus on an open science data imaginary. The findings contribute to the study of the datafication of health by revealing how their interactions helped them generate meaning, how they navigated the tensions inherent to making sense of personal data in community with others, and how they deliberated about the broader social issues implicated in their practice.

Advances in personal data tracking and analytics technologies have multiplied the ways in which individuals may generate and gather information about themselves. Apple Watches, ōura rings, continuous glucose monitors, mood-tracking applications, and digital food diaries are part of the increasing datafication of health–the rendering of “social activity” into “meaningful data” (Leonardi and Treem, 2020: 1603). These data can provide self-trackers opportunities to generate insights about their health and daily life, change behavior, and cultivate new pathways to knowledge of the self (Neff and Nafus, 2016). At the same time, powerful organizations control many of the devices and platforms used to collect personal data, which may be used to surveil and manipulate users (Zuboff, 2019). Self-trackers grapple with issues of automation, surveillance, privacy, and data ownership central to the “datafication” of work, play, and life more broadly (Flyverbom and Murray, 2018) and the “symbolic and imaginative work that underlie[s] coming to think of something as ‘data’” (Dourish and Gómez Cruz, 2018: 2).

Self-tracking is both a personal pursuit and social practice. Self-trackers engage in self-reflection and problem-solving surrounding the collection, meaning making, and use of intimate, personal health data with others. Their practices prompt and influence organizing (Smith and Treem, 2017), and social networks influence the health behaviors that constitute the data (Lomborg and Frandsen, 2016). Self-trackers’ data activism also shapes the practice, the technologies involved, and the social structures surrounding it (Kaziunas et al., 2018; Lehtiniemi and Ruckenstein, 2019; Sharon, 2017). As such, scholars have called for research that examines how data and devices can “materialize new forms of sociality” (Ruckenstein and Pantzar, 2017: 414). They have encouraged movement away from abstract, dichotomizing pros and cons of self-tracking and toward the study of self-trackers’ values (Sharon, 2017) and the potential of community-level participation in shaping the future of these technologies (Reich and Subrahmanian, 2021).

This study focused on LivedAnalytics (pseudonym); a web-based, global self-tracking community dedicated to advancing personal data initiatives. LivedAnalytics participants gathered online to share their personal self-tracking projects and discuss their data, devices, experiences, and how they ought to make sense of them. The LivedAnalytics website featured data processing tools and hosted data donation opportunities to support initiatives like using wearable data to predict flu infections and building human genome domains for public health. Members joined a weekly video meeting from around the world to share updates and get support over their ongoing personal data work, each with their own goals, tools, and methods.

Through observations of their weekly meetings and interviews with participants, we sought to understand this community's collective communication design work–the co-construction of their interactions and the efficacy of their communication choices as they made sense of their personal data in community with others. In sum, we found that through their interactions, they navigated tensions between the primacy of the self in personal data endeavors and the benefits of community for making sense. In doing so, they also grappled with broader social issues surrounding the ubiquity of digital surveillance and corporate control of personal data. The findings contribute to theory and practice concerning the social dynamics of self-tracking by documenting specific communication practices that helped this community cultivate useful conversations about personal health data, practices that connect their social imaginaries about data to their data activism. To frame the study, we now turn to a review of the literature on self-tracking.

Self-tracking

Research on self-tracking has grown in recent years alongside advances in the capabilities for collecting and interpreting personal data (Lupton, 2016; Neff and Nafus, 2016; Smith and Treem, 2017). We use the term “self-tracking” to refer not just to information gathering, but a rich constellation of information behaviors and meaning making practices (Mokros and Aakhus, 2002). The precarity of the self is central in self-tracking because ubiquitous data collection may empower but also may control or manipulate users (Moore, 2018). For example, self-trackers may be motivated by their own interests, or self-tracking may be imposed on them by an employer or incentivized by a health insurer. A communication perspective on self-tracking is useful because it can illuminate “meaning-making as a key aspect of the appropriation and use of self-tracking technologies” (Lomborg and Frandsen, 2016: 1018).

Self-tracking research makes clear the need to study the role of community in what might otherwise be understood as individual-centered technology use (Lomborg and Frandsen, 2016; Neff and Nafus, 2016; Smith and Treem, 2017). A thread of research on the communication involved in self-tracking has focused on “show-and-tell” talks of personal data projects (Choe et al., 2014; Sharon, 2017). Neff and Nafus (2016) studied show-and-tells at meetings initially organized by thought-leaders, writers, and co-founders of the Quantified Self (QS) collaboration, Gary Wolf and Kevin Kelly. Wolf (2020: ¶4) described show-and-tells as both a “report” and “instrument of learning” where the “act of saying what you did and what you learned operates retrospectively on the learning to crystalize or consolidate it.” The meetings let participants “discuss what they could do with their data and the pitfalls that they find” (Neff and Nafus, 2016: 42). The reflection cultivated in these meetings can help participants accomplish the difficult data collection and analysis involved in their practice and improve the analytical rigor that self-tracking has been criticized as lacking (Choe et al., 2014). This insight raises questions about how self-trackers can cultivate useful conversations about personal health data (Choe et al., 2014) and how communication constitutes self-tracking networks and communities (Lomborg and Frandsen, 2016).

Neff and Nafus (2016) used the term community to refer to “sets of relationships in which people discuss data, whether loose or tight-knit, and whether located in the same place or not” (p. 22). Online tracking communities may be conceptualized as those digital platforms that enable talk over a shared interest, but they may also involve varied ethical and epistemic goals and the (typically) voluntary labor required to reach those goals, thereby advancing a form of digital activism (Chamakiotis et al., 2021). The key is that the community shapes and is shaped by their practice, their interaction about the practice, and their beliefs about how that practice should be carried out. Online communities negotiate and enforce norms for interaction, which in turn shape the community itself (Aakhus and Rumsey, 2010; Smith and Treem, 2017).

Self-tracking communities are particularly interesting in this respect because of the centrality of data in their interaction and organizing. “Data” may be defined in varied ways, and indeed, a central concern for self-trackers involves competing ideas of what “data” means and how “data” is or are connected to reality, interpretation, meaning, and action (Ruckenstein and Pantzar, 2017). Self-trackers may view their data collection as the identification of existing information or as a subjective process of information construction (Kitchin, 2022). How self-trackers define, access, share, withhold, and interpret their data and their social comparisons about their data influence and is influenced by their organizing (Smith and Treem, 2017).

In communities like LivedAnalytics, members operate in diverse “data assemblages” that vary in terms of political landscapes, material access, data literacy, and so forth (Kitchin, 2022; Lupton, 2016). Their differing conceptualizations of data reflect their data imaginaries, or the “political and social alternatives that different social imaginaries ascribe to the notions underlying data activism” (Lehtiniemi and Ruckenstein, 2019: 1). Social imaginaries hold normative expectations about a person's “social existence, how they fit together with others, how things go on between them and their fellows, the expectations that are normally met, and the deeper normative notions and images that underlie these expectations” (Taylor, 2002: 106). Uncovering the types of interactions self-trackers engage and hope to engage in can contribute to the ongoing scholarly dialogue that aims to connect data activism practices to users’ data imaginaries (Lehtiniemi and Ruckenstein, 2019).

Nafus and Sherman (2014) described self-trackers as enacting a sort of “soft resistance” as they engage with data as self-researchers, technologists, and meaning-makers in ways that disrupt and yet are still rooted in the extraction and categorization and, accordingly, commodification and politicization of their data. Nonetheless, their data activism–the “relational practice” wherein self-trackers might “inter-react with the arena to hijack, supplement or contest the rules or purposes of the game” (Ślosarski, 2023: 10)—may challenge, reproduce, and produce data imaginaries. These data imaginaries are “collectively held notions of desirable futures, animated by shared understandings of social aims, and attainable through advances in technology” (Lehtiniemi and Ruckenstein, 2019: 2; see also, Jasanoff and Kim, 2015; Taylor, 2002). However, the study of data activism around personal data tends to be removed from concrete social practices (Kitchin, 2022; Sharon, 2017). Understanding “the orchestration of interaction” (Aakhus and Rumsey, 2010: 82) in self-tracking communities is needed to answer questions about how practitioners marshal data imaginaries in interaction with others to make meaning, organize, and grapple with issues such as data access, ownership, and privacy.

Communication as design

In this study, we conceptualized self-trackers’ negotiation of their interaction with each other about their data as communication design work (Aakhus and Rumsey, 2010; Jackson and Aakhus, 2014). Design is “an activity of transforming something given into something preferred through intervention and invention,” (Aakhus, 2007: 112) and “communication-des­­ign-work is evident in the interventions people make to realize preferred forms of interactivity and avoid nonpreferred forms” (Aakhus, 2007: 114). Communication-as-design (CAD) research involves the study of how individuals make choices about communication and their efforts to try to enact those choices. CAD research focuses on how individuals and collectives try to discipline interaction by creating and evaluating messages, interaction processes, and communication tools and formats to accomplish their communicative goals, guided by their beliefs about how communication, organizations, and institutions work (Aakhus, 2007).

A concern for the communication design work of collectives focuses attention on how individual and collective “goals, logics, and communication techniques” are negotiated “in the ongoing flow of interaction” (Barbour et al., 2018: 347). CAD theorizing has been used to examine how people intervene in and through communication in situations like facilitation in team meetings, impasse in divorce mediation, safety in nuclear power oversight meetings, and support in online health communities (reviewed in Barbour et al., 2024). A CAD approach sheds light on how communicators manage the tensions that arise from competing goals and beliefs about how communication works or should work–goals and beliefs also embedded in their imaginaries.

In CAD, communication sophistication describes the efficacy of particular choices about communication for navigating those multiple, competing, or contradictory goals (Barbour et al., 2018). More effective approaches tend to be sensitive to the complexity and requirements of context, and they tend to find ways either to integrate the demands, fulfilling them all at once, or by transcending the demands such that they no longer matter or are no longer in conflict (Woo, 2019). Identifying or developing communication sophistication is crucial for managing the tensional, contradictory nature of organizing (Putnam et al., 2016). Against that backdrop, design focused research “should be accompanied by reflection on design practice,” which includes “questioning the societal consequences of design” (Jackson and Aakhus, 2014: 133). Doing so can help surface how people try to navigate the complexity of organizing expressed in those competing demands.

The study of self-tracking has practical relevance here because it is ripe for uncovering communication strategies deployed as practitioners perceive, share, and mobilize personal data. Communication design is particularly important in informal online support communities like LivedAnalytics because their personal data practices help constitute the community (Aakhus and Rumsey, 2010; Smith and Treem, 2017), but the practice itself can isolate practitioners and fragment them. Neff and Nafus (2016: 12) warned of the potential for datafication to lead to “a preoccupation with the personal that erodes our capacity for coordinated community action.” At the same time, collectives may intervene to strengthen their communities by focusing on “who they are and what they are doing together” (Aakhus and Rumsey, 2010: 80).

Integral to self-tracking is the idea that health is knowable and mutable. At the same time, the push to capture the self in data may reflect dubious assumptions that humans lack the objectivity and predictability that only technologies and algorithms can provide, and that such an objective gaze is critical for self-understanding (Lupton, 2016). Self-trackers bring competing orientations not just to their own health, but also to the social dimensions of their practice as they navigate the issues related to automation, surveillance, ownership, objectification, and social and personal evaluations of their data (Nafus and Sherman, 2014; Zuboff, 2019). Taking a CAD approach focuses analytical attention on how they navigate such tensions in their choices about interactions, what Aakhus and Rumsey (2010: 69) described as the “design struggle.” Guided by the CAD framework, this question motivated our study of LivedAnalytics: how do self-trackers design their communication to make sense of data in community with others?

Methods

Our research team initially set out to find a community of self-trackers and to study their interactions about their personal data work and the communication problems they encountered, if any. After completing a review for the protection of human subjects (The University of Texas at Austin Institutional Review Board Protocol Number 2018-04-0124), the research team recruited participants by posting descriptions about the research interest of interactions around self-tracking in various online forums. A leader of the LivedAnalytics community responded, introduced the researchers to the community, and invited us to attend weekly LivedAnalytics virtual meetings.

LivedAnalytics was a nonprofit, open-source project dedicated to enhancing individuals’ access to and practices with data. The LivedAnalytics website hosted chat forums, blog posts and informational resources, and a tool that let members upload and analyze their data from common wearables, applications, and services such as Apple Health, 23andme, GPS trackers, and others. LivedAnalytics meetings were open to the public and welcomed new members regardless of their expertise or data literacy. Members joined through the website, which was promoted in various self-tracking online forums. The website hosted hundreds of members, and all were welcome to join the weekly self-tracking meetings. The bulk of meeting time focused on “show-and-tells” in which members shared and discussed their self-tracking projects.

We took a grounded practical theory (GPT) approach that oriented us first to the communication problems members sought to solve, and then the techniques they deployed to address those problems, and finally how their techniques revealed their situated ideals about their communication practice (Craig and Tracy, 2020). Our goal was to uncover the communication choices (techniques per GPT) that they saw as facilitating or constraining organizing around data. We arrived at an integration of GPT and CAD following previous research (Aakhus and Rumsey, 2010; Barbour and Gill, 2014) during our study as we watched members’ conversations about their self-tracking projects and their conversations about how they should interact about them. GPT and CAD integrate well because GPT provides a framework for collecting and analyzing data about the negotiation of communication practices aimed at addressing difficult puzzles in interaction that can reveal participants’ underlying situated ideals about the practice.

Participants

The number of attendees in any given meeting varied. Typical meetings consisted of about ten attendees. New people joined occasionally but there were fifteen members who attended regularly during our fieldwork. Membership in the community was fluid. Individuals attended meetings as often as they desired and shared and received feedback about their personal data projects during meetings. Members acknowledged the absence and presence of frequent attendees at the start of meetings, prioritized attention on new attendees, and shared status updates about ongoing conversations and collaborations, which contributed to a community-level understanding of their membership and shared knowledge. We refer to those members whom we observed and interviewed as participants.

Participants differed in their goals, methods, and years spent tracking, which ranged from less than a year to decades of tracking. The public nature of the group and the sensitivity of the topics necessitated that we obscure identities by using pseudonyms and limiting participant details. During meetings, various participants described themselves as persons of color, as persons with diverse gender identities, and as neuro diverse. They mentioned joining from Switzerland, France, Canada, the United Kingdom, and the United States. The most frequent attendees demonstrated technical sophistication, and many had formal training in technology or research as academics, software engineers, or journalists. The full range of participants across the span of our observations had more varied backgrounds and expertise. Examples of participants’ projects included optimizing sleep by manipulating environmental factors in the home, testing the impact of generic and brand-name versus placebo medication on illness, and capturing the day-to-day experiences of ADHD to formulate personal behavioral interventions. Methods included tracking with pen and paper, commercial health trackers, and devices they built themselves. Their projects required varied financial investments and time commitments. Alongside personal projects, they collaborated on initiatives like building a repository for wearable data collected during illness.

Data collection

Observations focused on participants’ interactions during weekly meetings. The first author attended twenty-three meetings, from December of 2020 until September of 2021, amounting to approximately twenty-four hours of observation. She assumed a participant-as-witness role: She disclosed her research interests at the start of each meeting, tracked her own health data related to her pregnancy for much of the observation timeframe, and shared her experiences with the group. She took fieldnotes that built a narrative by connecting dots across observations and interviews through ongoing memoing (per Tracy, 2020: 140). Note taking focused on participants’ stories of their self-tracking, their interactions with each other, and comments about how they ought to be interacting and organizing. She kept track of each participants’ project and their comments that stood out for later follow up during interviews.

Fieldnotes and memos totaled 89 typed, single-spaced pages (44,478 words). Fieldnotes were not verbatim but captured the flow of the interaction and verbatim phrases in jottings elaborated later. Observations supported researchers’ understanding of participants’ accounts by allowing interviews to focus on specific moments from meetings. The first author conducted all ten interviews and the third author participated in one interview and joined one meeting to sensitize him to the research context. The combination of interview and observational data was particularly useful for CAD and GPT research as interviews gave participants the opportunity to reflect aloud about their communication choices.

Interviews focused on participants’ experiences with self-tracking and the LivedAnalytics community. The first author recruited ten interviewees by describing the project during the go-around portion of meetings. Interviewees received a $50 gift card. One interviewee was a newcomer who joined halfway through data collection, and another joined at the very end of data collection. The remaining eight interviewees attended consistently across the observation timeline. Two interviewees, whom we refer to as Hugo and Ralph, facilitated meetings. Hugo also managed the community's website. Interviews were audio recorded and transcribed (209 single-spaced pages and 83,361 words). They typically lasted 54 min (range = 40–67 min).

The interview protocol was open-ended and semi-structured. The interviews focused on understanding the challenges and successes of their practices. We asked about (a) their story of their personal experiences with health data, (b) current self-tracking projects and what, if any, automation they implemented, (c) if and how they communicated with others about their practice, and (d) future goals for tracking. Probing questions were framed in terms of difficulties and strategies and included: “what do you think makes a meeting effective,” “what do you find difficult about communicating about your tracking,” and “when you want to ask for help, what do you do?” The data analyzed in this article cannot be shared publicly to protect the privacy of individuals that participated in the study. Data that may be shared such as anonymized vignettes and code lists will be provided on request.

Data analysis

Observations and interview recruitment continued until the iterative, ongoing analysis of the data suggested that a rich, comprehensive, and diverse set of experiences had been gathered from folks who had a range of tenures but a shared interest in the group. The first author began open coding field notes and interview transcripts prior to the completion of data collection, utilizing first-level codes to understand “the what” of talking about self-tracking practices with others (Tracy, 2020: 232).

It became clear early on that the LivedAnalytics meetings played an important role in the progress of members’ projects. Members described the importance of meetings for advancing their personal data pursuits. We observed how the meetings, while consistently focused on “show-and-tell” as a format, were crafted to meet the needs of members through comments about how the meetings should be. Second-level coding involved synthesizing “the what” codes using “why” and “how” questions guided by an iterative process of theoretical categorization (Tracy, 2020: 232), during which the first author returned to the relevant literature on self-tracking, communication as design, and grounded practical theory.

Axial coding focused first on the communication dilemmas that participants navigated (GPT's problem-level) and the techniques they seemed to deploy in response to them (GPT's technical level). The first author shared the data and codes with the second and third authors who also generated additional ideas for codes and categorizations. The research team made and revised analytical tables to refine the clustering of choices and beliefs about choices. This process produced a second, intermediate summary table that listed participants’ communication choices and their rationales for those choices. The first author revisited the data with this table to check for its exhaustiveness. She revised the axial codes by looking for examples that did not fit. Tensions between the priorities of the self and the community emerged through this process as it became clear that participants’ accounts included competing ideals.

We clustered the resulting codes in terms of participants’ communication choices about topic, manner, and timing. We treated those choices as revealing the designable features of their meetings as they understood them, and the next section organizes our findings according to those designable features and their rationales for their choices about communication. We also draw attention to their descriptions of how certain communication choices connected to certain beliefs about data. These choices were largely consistent across participants unless otherwise noted.

Findings

This study sought to uncover how self-trackers interacted over data in LivedAnalytics meetings and how participants saw particular communication choices producing particular outcomes. In sum, participants described making choices about the topic, manner, and timing of their interactions to help them engage their intimate, personal data in community with others. They made choices about presenting their data and responding to others’ that integrated tensions between (a) realizing the benefits of community as a generative tool for refining and making meaning of their practice, and (b) honoring each person as the final arbiter of meaning-making about their data. They also transcended competing demands specific to the self, their community, and the big data ecosystem through a shared data imaginary of open science. Their shared data imaginary was a resource for their communication design and the data-intensive nature of their practice shaped their communication as members made choices to live out their ideals for the future of self-tracking in their interactions. In the following sections, we describe the value participants placed on community and open science practices and then we provide examples of how these ideas were evident in their choices about how to interact.

Community and open science

Participants reported valuing the community that LivedAnalytics provided. They explained (and we observed) that show-and-tells helped them check their practice, learn from others, and rethink their self-research questions. As members reflected on the value of LivedAnalytics, they indicated that self-tracking could be isolating. For instance, Jonah described LivedAnalytics as “rewarding” because he “did not know anyone in his direct circle” with whom he could discuss tracking. Hakim described LivedAnalytics as the first place he felt “genuinely connected socially on these topics.” Self-trackers are “really isolated,” he explained, and perceived as “weird to everyone around them.” He said that members reminded one another, “No, no. We are amongst friends. We are all into this.”

Members also expressed a shared commitment to the community's embodiment of “open science” and “open source” methods for designing, accessing, and developing self-tracking tools and community resources. Open science practices are based on the assumptions that (a) scientific knowledge should be equally distributed and accessible to the public, and (b) scientists, politicians, and citizens should all be involved in advancing science via “open access, intellectual property rights, open data, and open code” (Miedema, 2021: 188). Members’ usage of these terms signaled their shared open-science data imaginary. For example, Hugo explained that he participated in show-and-tells because he was a “big fan and proponent of open science.” Hakim indicated that openness supported democratic processes and sustained innovation around personal data initiatives because, “if you document it well, then it probably will not die, and you can work with other people on it.” The value members placed in community and their shared commitment to open science influenced their choices for how to talk about their data.

Choices made in show-and-tells

We categorized participants’ choices into three clusters: (a) the topics they sought to talk about or avoid, (b) the manner in which they sought to present their work, question others, and represent themselves, and (c) the timing in which they sought to present, ask questions, and engage topics. As we detail below, these choices helped them orient around the goals of each individual and the community by integrating competing priorities in their interactions or by transcending the contradiction towards higher-order imaginaries about data. The following descriptions detail their choices as they put them into practice.

To describe these choices, we use a mix of participants’ reflections about them during interviews and examples of members talking about choices during meetings themselves. We focus on communication choices that most participants seemed to be negotiating to varying degrees. During meetings, we observed what seemed to be consensus about how they should be interacting. Their shared approaches to interaction were enforced in various moments by all members as opposed to being enforced by leaders in the flow of meetings; although, to convey the variation in our data, we also describe an example of a facilitator intervening in a show-and-tell that threatened the group's normative beliefs about data access. Taken together, the findings revealed their ideas of how their communication practice ought to be and “what missteps would invalidate the practice” (Taylor, 2002: 106).

Topics

In sum, participants described trying to emphasize particular topics in show-and-tells while avoiding others. They explained that show-and-tells should be for (a) giving updates about personal lives; (b) asking questions that interrogated data appropriateness and accuracy; (c) proposing high-level ideas; and (d) avoiding discussion that supported black box algorithms and technology.

Give personal updates

LivedAnalytics emphasized the individual in a way that built community by dedicating time and space at the start of meetings to check in with each member and encouraging one another to share details about how tracking fits into the context of their personal lives. Hugo explained that “people actually get to know each other” in LivedAnalytics. He differentiated their show-and-tells from formal talks where “people show up, listen to it, and then leave afterwards, and they might have gained something from the talk, but not necessarily any connection to others.” Members characterized self-tracking as involving imperfect processes where being “amongst friends” made it easier to be vulnerable. Rather than the formal, one-off events held by other communities, they described their informal, frequent meetings as key to cultivating a community that also honored everyone's data journey.

Question accuracy and appropriateness

Participants emphasized the importance of asking the right sorts of questions of show-and-tell presenters. When presenters struggled to find answers, the community asked questions that scrutinized the device, tracking methods, and phenomena being tracked. For example, Hugo gave a presentation on changes in his heartrate data across his use of various apps like Zoom and Netflix and wondered how accurate it was. Members asked questions like, “Do you get nervous during Zoom presentations?” and “What are you watching on Netflix that might explain your increased heartrate?” Others asked about the wearable itself, suggesting that data from wearables he cannot code himself are inherently problematic. Hugo concluded that he needed to “decompose the data further” and consider more variables. In interviews, participants explained that these exchanges capitalized on community expertise by helping the presenter scrutinize their work and refine their methodological and analytical strategies. Indeed, the community was less deferential to the individual in this form of questioning, but participants reported that the questioning was useful because it helped them generate insights. In this moment, they integrated the tensions between prioritizing self and community-level expertise by asking questions to make sense together of the data and its trustworthiness given the obscurity of the data processing system in Hugo's wearable. In line with their hopes for improving agency and accessibility, the community deliberated about data inconsistencies and interventions for addressing them.

Talk about big ideas

Participants explained that the meetings gave them space to talk about the future of self-tracking, including broader social concerns involving datafication, surveillance, and data ownership. Specific topics and comments connected show-and-tells to more abstract policy conversations. For example, Arthur lamented how hard it is to make sense of “black box magic” and argued that collectives like LivedAnalytics needed to push back against “data-extractive economies” and the “failures of the medical system.” In interviews, Arthur explained that meetings let him gauge the community's reactions to these ideas. In one exchange, Hakim described how using multiple sensors communicated more meaningful occurrences such as how the amount of light in a room affects one's immune system. Arthur agreed that these “effects” are the points everyone wants but are limited by ownership. He argued, “It's what we’re all pioneering for, these models of logic and reasoning. The cause and effect… but because we don’t own [the technology], we’re playing second fiddle to everything. We need to be at the front seat.” Hakim agreed, and pondered, “What could we do to support the ecosystem? I can’t do it alone, but I have friends!” These examples demonstrate the topical focus on broader social issues surrounding big data that participants explained should be part of LivedAnalytics meetings. Members integrated individual and community-level interests by honoring the diversity of members’ opinions and passion around specific technologies. They also grappled with the social implications of technologies by characterizing the future of self-tracking as malleable and their community interaction as a space for shaping that future.

Avoid Talk that Supports “Black Boxes.”

In contrast, participants emphasized that they should avoid advancing projects that contradicted their open ethos such as those that developed “black box” algorithms and technologies. For example, an application-development entrepreneur, John, came to participate in a show-and-tell and present data he tracked using his proprietary application. Hakim asked what the data export process looked like, a common concern of the group. John replied that, for now, he is willing to give users database dumps if they ask for them, but that eventually the business needs to stop “giving away the secret sauce.” At this point, Micah, a founder of LivedAnalytics and meeting facilitator, interjected: “I don’t want to get too deep into aggregate data sharing plans. It is important we focus on the interests of self-research for this group.” This interjection ended the discussion. This moment highlights a contradiction between John's individual interests to promote his application and the community's interests in data access. The community's commitment to open science obviated the contradiction, transcending it in the moment, by underscoring that open access and data transparency served everyone.

Manner

Participants also sought to shape how they communicated about their projects and questioned the work of others. Their choices emphasized that in show-and-tells, members should (a) share work that is imperfect and unfinished, (b) speak with humility, (c) ask reorienting questions, and (d) use persuasive questions to make arguments.

Share imperfect and unfinished work

During each show-and-tell, after a quick round of introductions, the group opened the floor. Participants explained they did not want to formalize expectations that might discourage members from presenting, and instead encouraged sharing imperfect and unfinished work. Presentations varied, including data organized in slides, screenshares of excerpts of raw or aggregated data, and showcases of physical hardware. Hugo explained that they arrived at the format of the show-and-tells guided by individual members’ needs, which vary in “speed” and “rigor.” He described how the community abandoned initial plans for a “structured and guided process” for meetings with “synchronized” project efforts.

Participants explained that this informal approach helped them make sense of self-tracking as it unfolded, pushing past the neat surface of stories. For example, Hakim explained the informality of show-and-tells was beneficial for his questions about data aggregation because they tend to be “half-baked and not thought through.” He elaborated that because he was trying “to solve aggregation problems and generalization problems,” he tended toward less polished questions regarding a particular phenomenon. Hakim also explained that, while his ideas were not “verbalized in a rigorous manner,” discussing them informally made them accessible and relevant to more people in the group who had different expertise or interests. Emphasizing the informality of show-and-tells made them more open to a diverse community and kept conversations from becoming too granular or personalized.

Speak with humility

Related to the emphasis on sharing imperfect ideas and projects, participants described in interviews the insecurities they grappled with to share their personal health data and associated projects. Participants described being mindful of differences in technological acumen and tracking experience. They described how one another's expertise was an invaluable collaborative resource, especially when their projects demanded advanced skills in data processing, analytics, or hardware development that were “incompatible” with their own skills. At the same time, participants described how they took care not to boast or focus on expert attributions that did not directly contribute to knowledge generation. For example, Yara reflected on how she has stopped describing herself in terms of her professional identity and her leadership in self-tracking initiatives to the group, and instead chooses to refer to herself based on her self-research project or the problem she is trying to solve: “I’m just Yara and I’m just doing the resting heart rate.” Focusing on the problem helped participants overcome the complexities of authority over knowledge in a group with diverse skillsets and experience.

Along similar lines, participants used hedging language when asking presenters about their project. We observed members say, “I’m no expert…” and, “You surely know more about this…” before posing their question, which appeared to put people at ease as other members scrutinized their projects. Yara pointed this out by highlighting that another long-term member and experienced self-tracker, Ralph, prefaced his questions by acknowledging his ignorance of a topic. Despite Ralph being a known expert in this space, he used humility to position presenters as the sole arbiters of their data. This approach integrated competing goals by framing each individual as the expert and authority on their data while taking care not to overshadow their shared identity with personal identity.

Ask reorienting questions

Questions interrogated accuracy and appropriateness (topic), and also helped reorient presenters (manner). Hugo explained in his interview that question-asking served as a proxy for more formalized methods of coaching common to other show-and-tell talks. He described this process as an “implicit negotiation” wherein “people ask questions, which show that they couldn’t follow, where people then need to adapt.” Asking reorienting questions rather than giving direction let community members help individuals make sense of their data without the community imposing meaning on it. For example, Thomas presented his dream data and noted a concern about the dwindling volume of his hand-written reflections. The group asked if Thomas's ability to engage with the manual, reflective part of his tracking may be compromised by other automated areas of his data. Sam said that “a tool you don’t trust can be really toxic to your reasoning process,” and the ability to “focus on the phenomenon you’re reasoning on will be limited by this skepticism around the data your device is giving you… is it me? Is it the device?”

Questioning to reorient helped presenters reconsider their work at different levels of abstraction. For example, Ralph described asking participants to step back from interpretations of their data and focus instead on the raw data to reorient them to their goals. He pressed members on their initial motivations, asking what would count as evidence relevant to those goals. “We have the tendency to get lost in these symbolic or rhetorical expressions about what we’re experiencing,” Ralph explained. “Show-and-tell forces members to ask and answer different questions of the data and move beyond initial representations or assumptions.” In one instance, Bob asked about Jonah's efforts to automate his diet tracking. “What is the hypothesis that you’re trying to investigate?” he asked, noting the difficulty of manually tracking diet. “Maybe you don’t need to track every micronutrient but create a broader hypothesis to simplify the tracking?” Questions like these helped members reorient their focus and rethink their self-tracking practices. In these moments, the community used questions to avoid imposing meaning while also challenging how automated technologies generated meaning.

Ask questions to advocate

Related to this point, participants explained and we observed that they also asked questions to advocate for social issues related to datafication that prompted deliberation in a way that still deferred to individuals. For example, during Hakim's presentation of sleep data from an öura ring, Fitbit, and in-bed sensor, Arthur asked a reflective question about data ownership: “Are you living the QS life or living the life of this black box algorithm that has pushed you toward consuming all these products you’re marketed on Instagram?” Following this, Arthur proposed a collaboration: “We could work as a community to deploy your own algorithm onto the hardware…[We] ought to dedicate our time to create personal algorithms. I don’t see how we can get to that phase of quantified self if we don’t own our algorithms,” he said. In an interview, Hakim reflected on how Arthur's question opened up conversations and opportunities for advocacy and prompted a collaboration between them.

Along similar lines, Colin used questioning to advocate for tracking genome data. He believed that population-wide genome sequencing was an important component of an open science future. We watched him ask other members about their own genetic data or if they have “been sequenced yet,” to suggest it may help them find answers about their health. He explained in his interview that a key motivation for attending meetings was to “sell members on the idea of getting themselves sequenced,” and asking questions was a way to do so “very gently.” He recognized the differences between the meaning he attributed to genome data and that of the broader group, and he communicated in ways that allowed for those differences.

Other members asked questions that indicated differing interest and trust in automated versus manual tracking technology. For example, Micah asked the group to reflect on why manual tracking is often considered “harder” for people. Hakim responded that automated tracking has difficulties too and described feeling like he's “trying to drink the ocean” because of all the data aggregation he wants. Micah echoed this sentiment and framed the issue of automation as: “Everyone talks about data flowing, but data is like luggage: You have to pack it and carry it in a suitcase, and then open it and unpack it. And you always realize once you unpack it what you’ve forgotten to pack!” Engaging such issues through questioning did not force consensus. They blended priorities of community and self as members brought one another into conversation about their data ontologies. Individuals varied in their tools, type of data, and use of automation. They advocated for the value of these choices by asking questions that prompted deliberation about the future of open science.

Timing

Participants also made choices about how often and when they should talk to support generative discussion and help them accomplish personal milestones. They emphasized that communicators should (a) set deadlines, (b) delay questions at times, and (c) avoid spending too much time discussing any one topic.

Set deadlines

Participants mentioned that having deadlines helped them overcome insecurities about when to share. For example, Hakim mentioned that members “don’t want to impose” or “don’t want to bother” other members with their problem. He referred to this as “analysis paralysis,” delaying sharing because “once I know more, I can be more confident.” Others echoed these comments and identified deadlines as helpful. They organized research showcases where people could commit to presenting in advance to provide a “kind of accountability or deadline to work towards.” Hugo said that without deadlines attached to projects, “we see that we are not making progress.” They explained that establishing deadlines for presenting gave members a collective goal, helped them move projects forward, and increased their inclination to share.

Delay talking

Participants also described making choices about the timing of their questions. They delayed asking questions and at times remained silent during the Q&A portion as presenters reflected on a problem. Yara described how remaining silent is “very powerful” because it makes presenters explain, and by explaining “…they come to some conclusion themselves.” She described this process as “empowering the person to restate,” and that by “talking and using this interaction” they could “find an answer together.” Intentional silence demonstrated the value of the community as an engaged audience but prioritized the individual presenters’ insights.

Spend time, not too much

Participants also emphasized that show-and-tells should not focus too long on any one topic. They explained that the diversity of capabilities and interests meant they had to be careful not to alienate members by getting too “in the weeds” in their conversations. They avoided this by gauging members’ interests and limiting the time they spent on any one topic. For example, after Ralph presented his new heartrate tracker, he proposed creating a messaging channel dedicated to deepening the observational possibilities of this tracker to “get deeper into the weeds of the data.” His comments underscored a shared sentiment that making choices about how members should talk was itself worth their time even if the conversation that followed did not occur during his show-and-tell or in the meetings at all. When others brainstormed about his proposal, Ralph interjected: “There are so many details here! I am using all my power not to get into everything on my mind right now, but I don’t think it's fair to take this whole group in that direction.” Ralph explained that he expected no more than “3 or 4 people to be willing to keep up with this space,” but “if you’re on this meeting and one of those people, you have a right to determine how we talk about this.”

Discussion

These findings elucidate how self-trackers gathered and made meaning of personal data in community with others. In sum, participants aimed to cultivate particular sorts of conversations about self-tracking shaped by the data-intensive nature of their practice. Their choices sought to preserve individuals’ authority in their self-tracking while building and taking advantage of the benefits offered by community. They did so by (a) integrating competing goals for their communication in discrete interactions with each other and by (b) transcending the contradictions between competing goals in the collective. Grounding communication choices in the tenets of an open science data imaginary helped members do the difficult work of organizing around multiple, diverse personal data projects and associated ontologies.

This study makes three contributions to theory and practice: First, the study contributes to research on datafication, societal debates about surveillance, privacy, and ownership, and the potential of community action as data activism. Participating in LivedAnalytics meetings meant members made choices that grappled with the data-intensive nature of their practice while fostering deliberation about these issues. LivedAnalytics's deliberations embodied broader interactional difficulties in datafication (Flyverbom and Murray, 2018). By describing how their data imaginaries functioned in practice we illuminated the issues at stake for self-trackers, how they navigated them, and the role of data imaginaries in organizing around data.

Second, this work provides an exemplar of how participants collectively designed show-and-tells as spaces to generate knowledge about their personal data in community with others. These insights have value in that they build a practical theory of how to solve communication problems that arise as groups reflect about personal data. The practical theory is captured in those statements about how communication should work; it should question accuracy and appropriateness, talk about big ideas, include imperfect and unfinished work, be humble, ask reorienting questions, and so forth. The findings build a grounded practical theory of self-tracking by connecting specific patterns in the negotiation of tensions between goals (integration and transcendence) with the kind of choices being made and implemented about communication (discrete, interaction-specific choices and global choices about an imagined future).

Finally, this research advances the study of communication design work and CAD by challenging existing conceptualizations of how designs for communication may attend to multiple demands at once (cf. Barbour et al., 2018). It also bridges the gap between work on data activism and specific practices (Kitchin, 2022; Sharon, 2017) by illuminating how collectively held values may help members transcend competing priorities between individual and community-level interests. We now elaborate on these contributions.

Data imaginaries in practice

The first contribution centers on the interactional processes that undergird social imaginaries about self-tracking. Members’ communication gave them space not just to connect but to exercise control over how they connected and by extension set their own personal health data policy and engage in dialogue with big data politics (Nafus and Sherman, 2014). Existing scholarship has highlighted the problematic nature of the datafication of health and criticized totalizing, closed corporate structures. This study profiled a community making its own alternatives (cf. Kaziunas et al., 2018). LivedAnalytics offered an empowering model of engagement with technologies, corporate interests, and human decisions that may otherwise capture and commodify life. Examining the transformative role of ideologies around data access and ownership for the negotiation of rules, expectations, and knowledge through discreet interactions contributes to ongoing conversations about the relational dynamics of grassroots organizing and activism around data (Ślosarski, 2023) by answering calls to document the concrete social practices involved (Chamakiotis et al., 2021; Kitchin, 2022; Sharon, 2017).

Indeed, members used meetings to experiment with their ideas about how self-tracking ought to progress. Their interactions informed the tools they used, the resistance they exercised against interests they saw as abusing their personal health data, the questions they asked of their data, and the meaning they made of their practice. At the same time, their ontological orientations towards data were multiple and diverse, and the interaction choices they made embodied the sort of openness they imagined for society. For example, members’ competing orientations to incorporating artificial intelligence in their practices fostered their deliberations about it. This finding disrupts traditional models of information behavior that rely too heavily on the “givenness of practice,” instead orienting us to see how interactions may be designed around data to “appreciate uncertainty and contradiction as a resource” (Mokros and Aakhus, 2002: 305). It also joins scholarship challenging assumptions about data as inherently neutral or transparent (Dourish and Gómez Cruz, 2018; Leonardi and Treem, 2020).

Communities that can engage with uncertainty and competing ontologies may advance the public good by challenging exploitative technologies. Zuboff (2019) highlighted the dangers of treating fitness trackers as toys due to their poor security record, the sensitivity and intimacy of data collected, their tendency to collect unnecessary and unspecified data, and opaque sharing (and selling) practices. This community's discussions helped them resist the built-in interests of wearable manufacturers and platform owners hoping to nudge users towards a specific set of behaviors or beliefs about their bodies (Lupton, 2016). Conversations about an imagined future for self-tracking helped members unpack the automated collection of data, question algorithmically determined meanings, and craft alternatives, comprising the sort of “soft resistance” that enabled them to “partially yet significantly escape the frames created by the biopolitics of the health technology industry” (Nafus and Sherman, 2014: 1784). Making space for their own narratives moved them closer to what has been called the “qualified self,” defined by Sharon (2017: 115) as “a more accurate expression of the entanglements and negotiations between metrics and interpretive schemes that characterize the quest for self-knowledge.” Future research that aims to understand how to foster such deliberation should incorporate the communication documented in this study that challenged and contested data-intensive surveillance while still using data-intensive technologies.

Designing self-tracking communication

Indeed, this study also contributes a grounded practical theory of self-tracking communication and show-and-tells in particular: The communication techniques and situated ideals through which LivedAnalytics accomplished the work of describing, optimizing, and making meaning of personal health data. Show-and-tells cannot be fully understood as isolated performances or as merely useful communication formats for sharing about personal data (Lomborg and Frandsen, 2016; Neff and Nafus, 2016). Taking a CAD approach to analyzing LivedAnalytics’ show-and-tells as complex processes of ongoing communication design helps us understand the problems self-trackers manage by seeing how their deliberations unfold in situ (Sharon, 2017). This study revealed how show-and-tells bolstered their practice and let them engage with what could be possible in self-tracking (Reich and Subrahmanian, 2021).

As a normative account of self-tracking communication, this grounded practical theory would hold that this communication should help generate insights about personal data because it navigates tensions between self and community and between the interests of the individuals and the potentially exploitative technologies in multiple ways. Their meetings encouraged extemporaneous, low-stakes sharing of “half-baked” ideas and asked questions in ways that took care to avoid imposing on the interpretations of others’ data. They shared knowledge but avoided talk of their own expertise when they suspected it might compromise their egalitarianism. Their talk gave them tangible opportunities for collaboration and experimentation without letting their diverse capabilities get in the way. Capturing approaches to the topic, manner, and timing of communication–their techniques per GPT–answers calls to concretize abstract political “themes” of the quantified self, including transparency and self-experimentation (Ruckenstein and Pantzar, 2017: 413).

Furthermore, datafication complicated members’ communication design because their focus on personal health data made the contradictions between self and community particularly fraught. Rather than coaching individuals toward specific conclusions about their data, the community asked questions to reorient individuals’ thinking and help them come to conclusions. This community cultivated useful conversations about personal health data, the sorts of participatory “reflection and dialogue” needed to achieve “a shared vision” for these technologies (Reich and Subrahmanian, 2021: 39), through specific communication techniques that integrated personal and community goals or transcended them by marshalling an open science data imaginary.

These findings expand our understanding of self-tracking as communication especially when users’ connectedness over data is not afforded by a shared tool or fitness endeavor (Lomborg and Frandsen, 2016), but rather, a shared data imaginary. Members of LivedAnalytics possessed different tools, goals, data literacies, and data ontologies, and yet their shared imaginary for open science practices motivated, generated, and deepened their communication. Their shared data imaginary did not homogenize the practices and values that functioned within their data activism. This community's approach to communication may be particularly useful for communities that organize around personal data and health conditions. Aligning with Chamakiotis et al. (2021), the particular technologies mattered less here than their ideas for interacting about the technologies.

Future research should investigate the extent to which users’ specific approaches to communication might inform future interventions. For example, wearable-based healthcare interventions should examine not just immediate health effects of adopting a technology or practice for the individual but the conversations and communities that form around those wearables. The community organizing that occurs around data-driven personal health interventions for stigmatized illnesses is an especially important context for study because these self-trackers may find it even more challenging to navigate individual and community-level interests as they interact.

Building theory of communication sophistication

Finally, these findings should prompt a reconceptualization of communication sophistication in CAD research. Members’ reliance on a shared data imaginary to transcend competing and contradictory goals pushes back on theorizing that complex situations demand more complex communication design (cf. Barbour et al., 2018). This past work has conceptualized communication sophistication as about the efficacy for coping with those tensions and contradictions and held that more complex communication moves would fare better. These findings suggest that communication sophistication should also include how it marshals shared beliefs apart from the complexity of the communication. In these data, participants’ shared data imaginary obviated contradictions.

These insights about their design work help researchers answer persistent questions regarding how individuals and collectives make communication choices that manage complex situations defined by multiple, overlapping, and contradictory goals. Communicators commonly grapple with tradeoffs especially as they attend to complex individual and collective-level goals in organizing (Aakhus and Rumsey, 2010). In these data, the communication that integrated tensions did so at the level of discreet, microsocial, interactive moves. Transcendent communication called on the overarching data imaginaries–a macrosocial move. This finding–that integration and transcendence may operate in microsocial or macrosocial ways–has broad value for tensional approaches to the study of organizing (Poole and Van De Ven, 1989; Putnam et al., 2016; Woo, 2019). Future research should assess this potential association between integrative and transcendent approaches and their micro- or macrosocial grounding.

Limitations

The value of these findings notwithstanding, this study also has limitations that should be considered. First, the economic costs of self-tracking can operate as a barrier to participation. The most common tools and platforms are not free, and even the use of free or low-cost tools require time and digital access and literacy to participate in communities online. Next, LivedAnalytics members were more technically sophisticated and knowledgeable compared to the typical self-tracker. The distinctiveness of this community may limit the degree to which insights translate to other contexts. Future research should focus on the usefulness of the strategies identified for communities that are more vulnerable to data extractive technologies. At the same time, the LivedAnalytics case is useful in part because it featured interactions between people both within and external to data professions, allowing for conversations that engaged both technological and socio-critical imaginaries about data (Lehtiniemi and Ruckenstein, 2019). Thus, the insights here might be applicable to contexts where diverse practitioners work together to make sense of data. Although most users were more knowledgeable than the typical person surveilled through the datafication of daily life, their practice offers a window into a future where more and more of us have access to such tools and a model for how we might deliberate about them.

Conclusion

Understanding the interactions of LivedAnalytics members and their efforts to shape them revealed difficulties inherent to data-intensive conversations. These difficulties center on issues of privacy, ownership, meaning-making, and technological change. Members drew on an open science data imaginary as they balanced competing priorities to honor individual authority over data and engage with a diverse community of trackers to support self-research. They aimed to improve the future of their community and society as well as their own lives through the work they put in each week, and they sought to instill and amplify these commitments through the design of their interactions. LivedAnalytics managed to honor the individual and capitalize on their organizing in their deliberation about the relative merits of human and machine labor.