Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Everyday curation? Attending to data, records and record keeping in the practices of self-monitoring:

Everyday curation? Attending to data, records and record keeping in the practices of... This paper is concerned with everyday data practices, considering how people record data produced through self-monitoring. The analysis unpacks the relationships between taking a measure, and making and reviewing records. The paper is based on an interview study with people who monitor their blood pressure and/or body mass index/weight. Animated by discussions of ‘data power’ which are, in part, predicated on the flow and aggregation of data, we aim to extend important work concerning the everyday constitution of digital data. In the paper, we adopt and develop the idea of curation as a theory of attention. We introduce the idea of discerning work to characterise the skilful judgements people make about which readings they record, how readings are presented, and about the records they retain and those they discard. We suggest self-monitoring produces partial data, both in the sense that it embodies these judgements, and also because monitoring might be conducted intermittently. We also extend previous analyses by exploring the broad set of materials, digital and analogue, networked and not networked, involved in record keeping to consider the different ways these contributed to regulating attention to self-monitoring. By paying attention to which data is recorded and the occasions when data is not recorded, as well as the ways data is recorded, the research provides specificity to the different ways in which self-monitoring data may or may not flow or contribute to big data sets. We argue that ultimately our analysis contributes to nuancing our understanding of ‘data power’. Keywords Self-tracking, curation, users, data power, partial data, material methods Introduction: ‘data power’ and the turn to repurposed. Scholars, for example, have drawn atten- everyday monitoring tion to the commodification of these data (e.g., Ajana, 2018; Van Dijck and Poell, 2016) and their potential The growth in apps, wearables and networked technol- contribution to surveillance, allowing, for example, ogies that measure or keep track of a plethora of bodily health professionals access to individuals’ conduct states, actions and experiences, has been referenced in a (Lupton, 2012). The terms dataveillance and lateral number of key discussions within social sciences. Self- surveillance are also used in this context, signalling monitoring has been characterised as disciplining and the more diffuse network of actors among whom normalising, creating particular kinds of neoliberal, self-regulating subjects and reinforcing obligations for self-care (e.g., Lupton, 2016). For some, it is seen as part of the broader ‘datafication of health’ (Mayer- Department of Sociological Studies, University of Sheffield, Sheffield, UK Schonberger and Cukier, 2013; Ruckenstein and School of Law, Politics and Sociology, University of Sussex, Brighton, UK Schu¨ ll, 2017; Van Dijck and Poell, 2016), in which, School of Applied Social Science, University of Brighton, Brighton, UK increasingly, aspects of bodily experience are trans- Corresponding author: formed into quantified data. Self-tracking data may Kate Weiner, Department of Sociological Studies, University of Sheffield, be seen as ‘lively’ (Lupton, 2016, 2018a) as they are Elmfield, Northumberland Road, Sheffield S10 2TU, UK. aggregated, analysed, circulated and potentially Email: k.weiner@sheffield.ac.uk Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (https:// creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Big Data & Society data may be circulated, including individuals who may to mindfulness and awareness of one’s own body and willingly share their data with their own social net- experience. Here, the act of monitoring or recording works (Andrejevic, 2005; Rich and Miah, 2017). may be as, if not more important than, reviewing The foregoing scholarship has been characterised as aggregated data (Nafus and Sherman, 2014; Sharon being centrally concerned with ‘data power’ (Kennedy, and Zandbergen, 2017). Scholars have also drawn 2018). Offering some critique of this, Ruckenstein and attention to the situated and embodied way people Schu¨ ll (2017) call for more attention to everyday make sense of, or assess the value of, tracking data in engagements with data in practice: ‘Scholars who relation to other ways of knowing, as well as the way attend to the power dynamics of datafication have emotions are intertwined with these valuations been faulted for their heavy focus on the oppressive, (Lupton, 2018a; Lupton et al., 2018; Nafus and normalizing, and exploitative forces of datafication and Sherman, 2014). their lack of attention to cases of noncompliance, A third theme relates to the hidden or invisible work appropriation and existential possibility’ (256). (Star and Strauss, 1999) of making data and allowing it Kennedy (2018) similarly argues that discussions of to travel. Pink et al. (2018), for example, are interested datafication tend to leave ‘little scope for agentic in the often obscured or hidden work of mundane engagements with data’ (20). One response has been repair. Introducing the idea of ‘broken data’ and to turn to more ethnographically informed research. ‘repair work’, they argue for ethnographic attention Gaining an understanding of everyday or mundane to the constitution of digital data, describing a process engagements with self-monitoring and the data that of improvisation or repair to fill in the inevitable gaps emerges, it is suggested, is important to inform both in people’s self-tracking data. For example, people use scholarship on, and policy and commercial expecta- multiple devices or use devices in unexpected ways such tions about, the role of data in society (Gorm and as using a step counter to record cycling. In this way, Shklovski, 2019; Kennedy, 2018; Pink et al., 2017; they suggest a focus on making sure data is coherent Weiner et al., 2017). for oneself with no responsibility to provide accurate There is now a blossoming scholarship on everyday data to each device or app (Pink et al., 2018). In her or mundane self-monitoring, often addressing fitness, work on a digitised, algorithmic physical rehabilitation exercise or food tracking, but also other areas including system, Schwennesen (2019) also enrols ‘repair work’ to self-monitoring of chronic health conditions. A number describe the way patients tinker with the system to of related themes are emerging in this scholarship and make it work in practice. Other scholars draw attention here we draw attention to three in particular. The first to the broader work of engaging in self-monitoring, takes seriously people’s emotional engagements with beyond generating data, that remains invisible to its self-monitoring data (Lupton, 2017; Pantzar and proponents (Ancker et al., 2015; Lupton, 2018b, 2019). Ruckenstein, 2015; Ruckenstein, 2014), countering This discussion of emotional engagements with and images of those who self-monitor as impartial, rational different values of data, and the work of making data, actors pursuing health aims (see Lupton, 2016, 2017; go some way to restoring a degree of agency to those Pantzar and Ruckenstein, 2015). This has included dis- who self-monitor. It helps to complicate narratives cussion of the enjoyment or pleasure derived from self- about the disciplining and normalising power of tracking, associated with for example seeing personal self-monitoring practices and about the flows of self- successes or supporting a self-identity as a fit or healthy monitoring data related to the potential for surveillance person, as well as disappointments, worry or frustra- and/or commodification. The ethnographically tion when these are not achieved (e.g., Ancker et al., informed work, in the tradition of user studies 2015; Gorm and Shklovski, 2019; Lomborg and (Oudshoorn and Pinch, 2003), therefore provides Frandsen, 2016; Lomborg et al., 2018; Lupton, empirical research ‘from below’ that helps to nuance 2018a, 2018b, 2019; Pink et al., 2017; Urban, 2017; the ‘data power’ argument. At the same time, some Whitson, 2013). of this work also considers the agency of things/devi- A second theme concerns the different values attrib- ces, which we discuss in more detail below. uted to data derived from self-tracking. In some instan- In this paper, we aim to extend important work ces, value is seen to derive from the (normalised) published in this journal concerning everyday data knowledge claims it allows, for example in the ability practices and, specifically, the everyday constitution to detect patterns, or lend credibility to facts (Fiore- of digital data (Lupton, 2018a; Pink et al., 2018, Gartland and Neff, 2015). However, data has also been 2017). Taking inspiration from and building on the shown to have communicative value, as a way to con- concepts of ‘broken data’ and ‘repair work’ (Pink nect with others, or share intimate stories (Fiore- Gartland and Neff, 2015; Sharon and Zandbergen, et al., 2018), we adopt and develop the idea of curation 2017). In other instances, self-tracking may be linked in relation to self-monitoring, using material from our Weiner et al. 3 study of the everyday practices of monitoring blood emotions associated with tracking and the value of the pressure and/or body mass index (BMI). data. In our discussion above, we have illustrated how people may gain pleasure or satisfaction and are able to communicate particular stories about themselves Adding a curatorial lens through the (hidden) work of curating their records. Curation is multivalent. Davis (2017) offers a theoret- In this context, curation helps to bring together the ical treatment of digital curation, describing curation as three emerging themes we identified relating to self- a theory of attention, concerned with how people allo- monitoring, linking the hidden work of making data cate and control attention. Drawing on examples relat- with the emotional aspects and the value of the data. ing to social media, she suggests that curation How does curation relate to the notions of repair ‘broadly .. . refers to the discriminate selection of mate- work (Pink et al., 2018) and filtration work (Nielsen, rials for display [online]’, where ‘productive curation’ 2015)? All these concepts help to bring to light the involves deciding what to ‘document, make, share, and hidden work of making data. While curation signals with whom’ and is integral to performances of self for the possibility of selectivity, repair work is suggestive oneself and for others (Davis, 2017: 771, 772). of an ultimate hope of completeness. Yet it does While there has not been a thoroughgoing applica- involve putting materials together for one’s own satis- tion of the notion of curation to self-monitoring, a faction. Where curation may be broadly communica- sense of this selectivity is present in some emerging tive, part of identity construction for oneself and studies of everyday self-monitoring practices. Kent conveying this to others, filtration work in Nielsen’s (2018: 67), in a study of how self-trackers represent (2015) account is solely orientated to others. It is con- ‘health’ through social media, discusses the way her cerned with opening up or closing down particular con- participants construct an appropriate self-tracking per- versations with specific actors. In this paper, we would sona ‘through careful inclusion and exclusion of certain like to propose curation as an overarching concept, health information’. Studies of calorie and of fitness where repair work and filtration work offer particular tracking have documented the way participants may examples of this concept. Curation helps to illuminate manipulate data input, for example not recording every- the hidden or underarticulated work of producing and thing consumed on days of excess (Didziokaitee _ t al., sharing self-monitoring records. It, thus, helps to bring 2018; Lomborg et al., 2018), or not saving ‘unflattering’ the agency of those who monitor into view. Yet the runs (Esmonde, 2020: 84). They might also engage in concept of curation does not only illuminate the work ‘episodic use’ (Gorm and Shklovski, 2019) of tracking of human actors, but may also acknowledge the work technologies, recording calories or wearing fitness track- of materials. ers only on days when they anticipate good, interesting or useful numbers (Didziokaite_ et al., 2018; Esmonde, Curation as socio-material practice 2020; Gorm and Shklovski, 2019; Lupton et al., 2018). In this way, participants are selective in the records they In her discussion of curation, Davis (2017: 775) attends create either imagining an external audience, or support- to the agency of materials through making a distinction ing their motivation and protecting themselves from dis- between ‘human’ and ‘machine’ curation, discussing appointing outcomes. how the design of platforms and algorithms shapes In Nielsen’s (2015) work, the external audience is and constrains the way users produce and consume particularly important for patients, who she suggests online content. Pink et al. (2017: 3) make a similar undertake ‘filtration work’ when making entries to a move in relation to self-monitoring data, wanting to new e-health system. This involves being selective in rela- take ‘users’’ perspectives seriously but also to ‘decentre tion to what information to provide and has a particular, the human’, suggesting that ‘personal data’ is ‘consti- dialogic, orientation; patients imagined the receiver, and tuted and experienced between human and digital/algo- shaped their entries in line with conversations they hoped rithmic devices and processes’. The breakages in data to pursue or avoid. Work on the development of a clin- they describe, when devices are not charged or lack connection, when software updates make existing devi- ical self-monitoring system for diabetes similarly showed how patients might decline to share data or respond to ces redundant, or devices track some activities but not clinicians’ messages (Piras and Miele, 2017). All of the others, draw attention to the way devices and platforms studies discussed so far illustrate selectivity in records shape the production of self-monitoring records. As we made or shared, suggesting there is value in the concept have discussed, their work also documents the way of curation in relation to self-monitoring. users may attempt to get around material constraints In considering the value of curation as a conceptual through their repair work (Pink et al., 2018). In previ- lens in this context, we need to acknowledge that cura- ous research we have drawn attention to the multi-user torial work is suffused with and inseparable from the functionality of some devices for measuring blood 4 Big Data & Society pressure and weight, to highlight the way these shape, who generated them. Ultimately we propose curation or script (Akrich, 1992) particular ways of recording can therefore be helpful in interrogating concerns with and sharing data (Williams et al., 2020). These sorts data power. of socio-material analyses illuminate the way platforms and devices shape the production and management of Methods self-monitoring records without resorting to technological The paper is based on a UK study involving interviews determinism. They allow space for both users and tech- with people who self-identified as monitoring their blood nologies (and their developers) to have agency (Henwood pressure or BMI/weight. Our engagement with self- and Marent, 2019; Lupton, 2018a; Oudshoorn and monitoring stemmed from our broader interest in every- Pinch, 2003). day health practices, the use of health technologies in Yet, in considering the material dimensions of cura- domestic settings and the way these might redistribute tion we would like to draw attention to the kinds of health work between the home and the clinic (see self-monitoring so far discussed in critical scholarship. Henwood and Marent, 2019; Weiner et al., 2017; This has, with notable exceptions (e.g., Lupton and Weiner and Will, 2018; Williams et al., 2020, forthcom- Smith, 2018; O’Riordan, 2017), tended to focus on dig- ing). Home blood pressure monitoring and BMI moni- ital and networked types of self-tracking involving, toring offer particularly interesting cases in the way they especially, fitness and diet apps and wearables. Yet, blur the boundary between the clinic and the home. as Neff and Nafus (2016: 98) note: ‘self-tracking tools In the UK there are well established consumer mar- do not have to be fancy’ and might include low tech kets for both blood pressure and BMI monitoring. A materials such as pen and paper. Indeed, Fox and range of devices are available to purchase in supermar- Duggan’s (2013) oft-cited research reported that the kets, pharmacies and online retailers, such as digital majority of Americans who track a ‘health indicator’ blood pressure monitors, digital and analogue weighing did this with pen and paper or ‘in the head’. Rather scales and digital body analysis scales. These products than equating self-tracking with digital and networked include both stand-alone and networked devices and self-tracking, we think it is important to consider the may be accompanied by proprietary apps, but also wider materials and technologies and their place in paper booklets or diaries for recording readings (see the practices of self-monitoring. What, for example, Williams et al., 2020, forthcoming, for further analysis are the implications of different materials for data of this market). There are also other apps to calculate/ flows? Further, since curation is concerned with allo- track BMI or track blood pressure, such as cating and controlling attention (Davis, 2017), are MyFitnessPal and Apple Health, where data may be there material dimensions to paying attention, avoiding entered manually or pushed through from networked noticing or being inattentive to self-monitoring? devices, as well as websites providing online BMI calcu- In sum, we propose that a curatorial lens facilitates lators. Both forms of monitoring have clear links to the exploration of the way self-monitoring data is con- clinical interests. Monitoring blood pressure is well stituted in practice, illuminating the work of both established in clinical practice and self-monitoring is humans and materials. Further, the idea of curation increasingly sanctioned as one response to white coat helps to link the work of making data with the emo- hypertension (doctor-induced high blood pressure) tional aspects of self-monitoring and the value of the (National Institute of Health and Care Excellence data. In our analysis, we adopt this lens to develop a (NICE), 2011). Clinical concern with BMI and weight socio-materialist account (Henwood and Marent, 2019; relate to obesity, and to risks of diabetes and cancer and Weiner and Will, 2018; Williams et al., 2020, forthcom- forms part of public health messages (Gatineau et al., ing) of everyday data practices relating to self- 2014; Hooper et al., 2016). In sum, both have clear clin- monitoring, exploring what records people keep, what ical relevance and established self-monitoring markets. materials are involved and whether and how records While our study involves voluntaristic self-monitoring, are shared. We suggest that this curatorial approach we acknowledge the non-innocence of self-monitoring helps to clarify the relationship between self- monitoring and the accrual and flow of data. By technologies and their links with broader socio-political paying attention to which data is or is not recorded, contexts. Notwithstanding the contested history of BMI, as well as the ways data is recorded, the research pro- the measure links with weight management which is asso- vides specificity to the ways in which self-monitoring ciated with strong narratives of personal responsibility, may or may not contribute to big data sets in different guilt and shame (Lupton, 2013). This and other forms ways. It allows reflection on the ‘liveliness’ (Lupton, of tracking intended to work on the body relate to gen- 2018a) of self-monitoring data, in terms of their poten- dered norms of beauty and fitness as well as to health tial to be circulated, reconfigured and monetised, and (Esmonde, 2020). Relatedly, there are clinical/psycholog- do so in ways that might act back on the individuals ical concerns about the possible links between Weiner et al. 5 food-tracking apps, such as MyFitnessPal, and eating In this paper, we focus on self-monitoring data prac- disorders (Lupton, 2018b). Discourses relating to tracking tices and the materials this involves. It is not our inten- are also infused with assumptions about people’s capacity tion to provide a definitive definition of self-monitoring to incorporate tracking which do not chime with gen- and we do not see an obvious difference between this dered, classed or marginalised experiences of daily life and self-tracking. Resonating with other research (e.g., or work routines (Ancker et al., 2015; Esmonde and Lupton, 2019) we followed an emic approach, keeping Jette, 2018; Lupton, 2018b). At the same time, there are our recruitment material broad and allowing people to concerns that fitness tracking may be pushed or imposed identify themselves as engaging in self-tracking in order (Lupton, 2016) by healthcare insurers or employers to study what this involves for them. Lupton (2016) (Ajana, 2018; Esmonde and Jetter, 2018; Lupton, 2016). proposes that self-tracking entails ‘practices in which We note, however, the relevance of these concerns is lim- people knowingly and purposively collect information ited in the UK context, where healthcare is largely about themselves which they review and consider apply- accessed through a universal, national, government- ing to the conduct of their lives’ (2). In our analysis, we funded system. Even so, self-tracking is likely to be explore the distinctions and relationships between these linked with uneven and differentiated experiences and different potential aspects of self-monitoring focussing effects. on three main themes: the relationship between taking In our study, we made efforts to recruit a diverse and recording measures, how and where records are sample. Following institutional ethics approval, we made, and storing and reviewing records. advertised on email lists at three UK universities and noticeboards across campuses, at older people’s groups and at community centres in less-advantaged areas. Findings The advert sought people who identified themselves as ‘measuring and keeping track’ of either their blood 1. The relationship between taking and recording pressure or BMI. In this paper, we draw on 67 inter- measures views conducted with 81 people, including 14 inter- a. No records views with couples. Participants varied in terms of We start our analysis by considering the approxi- age, sexuality, ethnicity, socio-economic background and health. All had acquired monitoring devices for mately one-quarter of our participants who took meas- themselves and no one reported acquiring these from ures but did not record these. Understanding curation employers or clinicians. While we were alive to issues of to be concerned with attention, we consider what diversity, we did not find these significant in the current people are attending to in these cases. In other words, analysis, although they are more central to other if they are not recording their data, what are they doing themes (see Will et al., 2019). when they self-monitor? Sometimes participants did In interviews, we asked people how they came to this for reassurance, just wanting to know if their monitor or acquire a device, what they do or do not blood pressure or BMI was in the normal range and do with it and who else might use it, how this may have they were able to recall this without needing to remem- changed over time and with whom data is shared. The ber the precise number or to keep records. People limitations of ‘conventional’ social science methods talked of monitoring ‘to keep an eye on something’ such as interviews for researching everyday life are or ‘for peace of mind’, illustrating the emotional reso- well rehearsed (Martens et al., 2014: 3). People may nance of the practice. For example, Gary explains he find it difficult or are unable to talk about certain ele- has anxiety issues and uses his blood pressure monitor ments of their everyday practices, in particular embod- for reassurance: ‘I need to know if there’s something ied, tacit and affective aspects (Martens et al., 2014; wrong you know ... So if I think I’ve got a bit of a Martens and Scott, 2004). The use of material objects headache or I get some palpitations I’ll check it’ or photos in interviews can provide an aid to memory (Gary, 45, school administrative officer, white British). and reflexivity that interviews alone cannot elicit (e.g., Gary does not record his readings, and cannot recall Harper, 2002; Woodward, 2016). In our interviews, we the precise numbers from the last time he used his monitor invited participants to demonstrate their monitors and (four weeks before the interview), but knows they were talk through any records they kept and where these ‘under the 140 and 90’ which he called ‘the bench mark’. were stored. This helped both to prompt reflection For other participants, monitoring was concerned and tie practices to particular time periods and with managing day-to-day conduct. Linda, for exam- events. We analysed the interviews thematically ple, does not keep a record of her weight and uses BMI (Hammersley and Atkinson, 1995), collaboratively as trigger to take action when she sees herself ‘creeping’ developing a coding frame, which synthesises our the- oretical grounding with emergent themes. near to the boundary between normal and overweight: 6 Big Data & Society If I see myself creeping .. . I haven’t actually got to the I don’t write down the good numbers, I only write point of going into the next category .. . so it’s sort of down the bad numbers. So when it’s fine I don’t time to take some action in the sense of, you know, just bother, but when it’s bad I think I probably should, cutting back on what I’m eating, being more careful because I’ve got a rubbish memory I think I probably about portion sizes, that sort of thing. (Linda, 67, need to keep a record of that. (Annie, 45, university retired Further Education teacher, mixed heritage) administrative officer, white British) Occasionally, blood pressure measures also resulted in So in contrast to Ayo, who only records her measures immediate action such as drinking some green tea or when they go in the right direction, Annie only records trying to relax. ‘bad’ numbers. Participants measuring blood pressure In these cases, people were not seeking to under- also discussed processes of selecting or averaging mul- stand patterns in their data, but to attend to their tiple readings for recording, e.g. best of three. We also immediate bodily status for reassurance or potential encountered occasional stories of participants curating actions. Self-monitoring helps to address questions charts to make them more meaningful or pleasing, for such as: how am I today? Am I stressed? Do I need example by removing outlying data points or choosing to go to the doctors? Should I eat less today? This helps the span of time and the right axis. Gareth, for exam- to explain why some participants cannot recall precise- ple, showed a graph of his weight on his Google Fit ly or do not record or review their data. app, illustrating how the axis changes when he selects different years, and how the falling line pleases him: b. Discerning work and partial data ‘That’s me overall graph. I’m quite pleased with how that’s steadily falling. Especially when I put that year in Yet among those who did record data, participants it puts a different axis on it and it’s whoosh’ (Gareth, described selective recording, including not recording 58, property maintenance engineer, white British). particular readings. Ayo weighs herself on stand- The foregoing accounts illustrate how participants alone digital scales, and records this into her are selective in the way they compile data into records. Samsung Health app, which calculates her BMI. We propose that they are undertaking discerning work She told us she only records her weight when it had (rather than ‘repair work’), making judgements about gone down: which readings are useful, or worth remembering or drawing attention to, and how to process or clean read- Ayo: When I weigh and it’s more and I’ve put on ings to make best sense of them. Further, rather than weight I don’t enter into the app to update my the metaphor of ‘broken data’, we propose the idea of BMI .. . I only do it when I lose weight partial data may be more apposite in these instances. Partial here has a double meaning, understood in the Interviewer: So how come you don’t put it in? sense that only some of the data get recorded, but also in the sense of interested or partisan, in contrast to Ayo: Because it makes me sad ... The fact that I’ve put impartial or neutral. The readings written down or on weight, which is not what I want. I want to lose entered into apps may only be a subset of the readings weight .. . So that’s sad for me so I don’t bother putting taken and may be selected for very particular reasons. it onto .. . each time I found out I’ve lost weight then I add my weight here ... I want to see that I’m losing c. Intermittent measures and partial data weight on my app’. (Ayo, 33, university researcher, black African) In other interviews, participants told us of intermit- tent measurement which lead to intermittent records. In Ayo’s account again underscores the affective reso- interviews relating to blood pressure, participants some- nance of self-monitoring records, which have the times compiled records for time-limited periods specifi- capacity to make her ‘sad’ if they go in the wrong direc- cally to take to clinical consultations. For example, Fred tion. This chimes with studies of calorie and fitness records his blood pressure for one month prior to his tracking (Didziokaite et al., 2018; Esmonde, 2020; appointment compiling a spreadsheet which he prints to Gorm and Shklovski, 2019). take to his doctor’s appointment. Such intermittent Other participants reported, in a more pragmatic records resulting from intermittent measuring might be vein, that there was no point noting down weight or understood as a second form of partial data, in the sense blood pressure when it was stable, only noting when that it is compiled from time to time. there is change. Annie records her blood pressure on Fred’s account is useful in reinforcing an important scraps of paper and in a booklet. She told us she only point of our analysis by demonstrating the consider- notes this down when her reading is high: able work of making records. Fred told us that he Weiner et al. 7 records his blood pressure on pieces of paper by hand we have described, or in the sense that they measure, (Figure 1), transcribing these data into a spreadsheet and therefore record, intermittently. From a user per- which he compiles specifically for his appointment spective, records may not be intended to be comprehen- (Figure 2). He labels the spreadsheet with the name of sive or continuous, so here there is no ‘broken data’ and his GP practice and the date of his appointment. This therefore no need for ‘repair work’. This means that, helps to illustrate the curatorial work involved in beyond the data breakages identified by Pink et al. making data ready to share. In moving from hand- (2018), there are other reasons why data may not flow written slips of paper to a neatly presented spreadsheet, seamlessly from measurement to an individual’s records Fred demonstrates the skilfulness and probity of his self- to be aggregated by third parties. All of this should act monitoring. The materiality of the spreadsheet commu- as a gentle corrective to expectations about the exploi- nicates ‘I am a responsible patient, you can take my tation of such data in existing literature (see Ruckenstein readings seriously’. and Schu¨ ll, 2017; Van Dijck and Poell, 2016). So far we have shown that the practices of taking a reading do not always lead to keeping a record of those data. For some of our participants, self-monitoring did 2. How and where to record not involve making any records. Where people do keep records, the readings entered into apps or recorded The analysis so far has mostly concerned which measures people record, but has also touched on the importance of elsewhere may only be a subset of the readings taken. We have suggested the term discerning work to describe how records are made and presented, particularly in our the skilful judgements people make about which data discussion of Fred. We now turn to this theme in more to record, which to omit, and how to process and pre- detail, to expand on issues around how and where people sent records that differentially draw attention to suc- record and the materials involved. We will consider the cesses, warning signs or the credibility of the person socio-material arrangements in this curatorial work that making the record. We have also suggested that partic- draw attention to or deflect attention from different ipants may create partial data both in the sense that aspects of self-monitoring. they may choose only to record some of the readings they take, selecting these through the discerning work a. Visibility and being reminded Figure 1. Fred’s handwritten notes, retrieved from the waste paper bin during the interview. 8 Big Data & Society Figure 2. Fred’s spreadsheet compiled for his doctor’s appointment, practice name blanked out. While half of the participants in the study had expe- with her sister and sister’s wife. They met on Saturday rience of using an app to track BMI, the visibility of mornings to record their weight and kept a joint record paper records emerged as an important theme. For on a sheet of paper. The record had been set up on a example, Becky told us she was losing weight together spreadsheet, but this was printed off and weights Weiner et al. 9 written on by hand (Figure 3). When asked why they resonates with some studies of digital self-tracking did not simply enter the data onto the spreadsheet, (Lomborg et al., 2018; Pink and Fors, 2017, cf. Kent, Becky responded: ‘I think because it was going to be 2018), placing into question expectations of widespread a group thing that we could all jot it down while we lateral surveillance (Rich and Miah, 2017). Here we were together. So I think that’s why I have a physical extend the existing analysis by considering not only sort of ... ’ (Becky, 36, charity researcher, white with whom participants share, but also the materiality of sharing. British). The visible emplacement of monitoring devices in In contrast to the digital co-presence discussed by Pink and Fors (2017), where people who are physically particular domestic spaces, for example close at hand separate share data and are present together online, on a table next to a favourite armchair, may encourage self-monitoring in our study involved physical co- people to monitor (Weiner and Will, 2018). In the same presence where different materials come to the fore. way, the emplacement of self-monitoring records, such A paper record in this account appears to allow these as a chart pinned to a wall or a record on a mobile three women to participate together and to attend to phone that is always to hand, might act as a reminder their data collectively. in different ways. Participants told us that leaving It was striking that records for sharing with part- paper records and charts somewhere visible within ners, relatives and friends tended to be paper, charts the home helped to remind them to monitor or and/or DIY forms of digitally networked communica- helped to keep commitments in mind. Becky, for exam- tion such as texting or setting up WhatsApp groups. ple, told us that the shared record she made with her We encountered very little discussion of sharing sister and sister-in-law was pinned to her notice board through broader social media or proprietary self- in her home office. She explained that this was visible monitoring apps that would facilitate sharing with enough to remind herself she was trying to lose weight, wider social networks or publics, even where partici- but not so public that visitors to the house would read- pants had devices with the capacity to do so. The mate- ily see it (compared with, for example, pinning it to the rials our participants discussed appeared to allow them fridge in the kitchen). It is placed to hold her attention to do things together with limited, selected others while avoiding drawing the attention of visitors. This (friends, family), and offer each other encouragement, concern with the emplacement of the chart suggests while precluding broader attention. This finding that even paper records have the capacity to act as Figure 3. Becky’s ‘group thing’, names blanked out. 10 Big Data & Society ‘indiscreet technologies’ (see Oudshoorn, 2012), she continued to record her calorie intake in the app, she making public aspects of identity or practice that had stopped recording her weight in this because she people would rather keep to themselves. In Becky’s found it contributed to her becoming ‘obsessive’ about case, the materiality and emplacement of the record it. At that point she told us she preferred to record her lend themselves both to monitoring in a group and to weight on a weekly basis in a notebook: keeping a project in mind, while allowing the record to remain relatively private. So although I weigh myself at the minute, I’m not put- Others told us that they recorded on the phone ting into MyFitnessPal because I found I was getting because it is always with them, unlikely to be forgotten maybe a bit too – I started weighing myself every day or because it formed a convenient mode for transport- and I may have got a little bit too obsessive about ing records. Bella, for example, related that it was not it .. . I felt I’ve gradually started being calmer about until she got her smartphone and found a free blood it .. . So I thought I’m going to gradually start doing pressure app that she started to record her readings. all the things that I used to do again which is weigh Before that she had not made a record: myself once a week. (Andrea, 27, university adminis- trator, white British) because it was a pain in the neck .. . Because I had to keep writing it down and then remember to write it While MyFitnessPal is designed to allow and encour- down and find the piece of paper, like that. So I was ages daily recording of weight, Andrea’s ‘weekly weigh really happy when I found the app. (Bella, 57, charity in’ notebook is quite literally scripted for weekly administrator, white British) recording (Figure 4). In Andrea’s case, she recounts being overly concerned with, or over-attentive to, her In contrast to a ‘piece of paper’, Bella’s phone is always weight and in changing to a different way of recording nearby. Yet recording on phones did not always tries to regulate this over-attentiveness. involve using proprietary tracking apps, as participants In this section, we have paid particular attention to also told us they used note apps or Google Sheets to the materiality of records, looking at the different ways record self-monitoring data. Samuel, for example, these contribute to paying and regulating attention to talked about recording his blood pressure readings in self-monitoring. Like others (Lomborg et al., 2018; a note app, to take to his doctor’s appointment: Pink and Fors, 2017), we found little sharing through proprietary self-monitoring apps, but the employment Interviewer: you said you’d put the records on your of other materials – paper, spreadsheets, WhatsApp phone for a time. Was there any reason for that? groups – which work to limit attention to small, select- ed groups of people. The visibility and emplacement of Samuel: Only for ease of transport, I knew I’d have my paper records may facilitate collective practice within phone with me. I haven’t kept them as a record, it was the home and help to remind people to monitor or keep just a way of transporting information to the surgery commitments in mind. The nearness of phones facili- with me. tates the recording of readings. Yet in one case, the immediacy of a mobile app was associated with over- Interviewer: Okay, so was it an app for the blood attentiveness, and a paper record helped to remedy pressure? this. More baldly, the diverse materiality of self- monitoring records, including, but not limited to pro- Samuel: No, it was just a note. (Samuel, 62, university prietary self-monitoring apps, suggests that self- counsellor, white British) monitoring data may not always be readily compiled or harvested by third parties, placing brakes on the The constant presence of phones facilitated the record- liveliness (Lupton, 2019) of this data. ing of measures and made sure records were always at hand. 3. Storing and reviewing records a. Broken data and repair work b. Not being reminded Curation involves both what people make and what they keep and display. This relates to what they would While a few participants told us they had moved to like to remember or be reminded of. We found occasion- apps from other ways of recording, in one instance a al stories of people going to efforts to retrieve data from participant had moved away from an app, precisely different sources in order to retain a complete record. because of the attention it demanded. Here again the emotional resonance of tracking comes into view. These may be understood as examples of the repair Andrea told us she used MyFitnessPal and that, while work and broken data proposed by Pink et al. (2018). Weiner et al. 11 Figure 4. Andrea’s ‘weekly weigh in’ note book. John (55, IT support, white British) had recorded his b. Dormant records weight and BMI on a weekly basis for the last decade using a number of different platforms. Initially he used a In considering self-monitoring through the lens of weight loss website called Weightloss Resource, because curation, we have so far discussed record keeping prac- tices in fairly deliberate terms. We have portrayed self- his wife was already subscribed to this. He ended this monitoring records as being created and shaped subscription in 2014, downloading his data to a Google Sheets spreadsheet, and moved to MyFitnessPal, which through a combination of the discerning work of he used for 10 months before getting a Fitbit. He told us humans and the materiality of the devices and broader that to export his 10 months of data from MyFitnessPal technologies involved. Our final brief section provides a would incur a fee, which he was not prepared to pay, caveat to this view, suggesting that sometimes the although he lamented the ‘gap’ in his data. human and material elements combined in such a We asked him at different points in the interview way that participants found it difficult to keep track why he had downloaded his data and if this was impor- of their records. For example, Tony keeps records of tant to him. His responses suggested an emotional con- monitoring his blood pressure in a rather ‘haphazard’ nection to graphs as ‘comforting’. They also posed a fashion on various slips of paper and backs of enve- link between records and biography – ‘a reminder of lopes. He told us he stored most of his records in a bag where you’d been and where you’d come to’. John but that he threw a lot of these away: related this to one particularly significant time in his life, during which one of his daughters was diagnosed Interviewer: Do you normally keep the readings in with and treated for a serious illness. For John, retain- that? ing and looking over his records appeared to be both a way to celebrate his successes in weight control and to Tony: Yeah in a rather haphazard fashion. On bits of notes .. . I just used to write them down on bits of paper remember how he and his family had come through his daughter’s illness. This underscores the emotional and and shove them into this bag. And then I was packing communicative aspect of these records. this all away one day and suddenly thinking gosh 12 Big Data & Society there’s an awful lot of ancient results here, I’m never through the concept of curation, we illuminate both the going to do anything with these and I remember I human work involved in making and retaining records, chucked a load away. (Tony, 54, electronic engineering while, at the same time, taking seriously the role of lecturer, white British) materials. Understanding curation as a theory of atten- tion, we have analysed the different ways both humans The emplacement of records ‘shoved’ into a bag means and materials are implicated in drawing attention to, or that they do not seem to hold Tony’s attention. Like detracting attention from, the practices of self- Tony, participants were often not trying to discern pat- monitoring and the data these create. terns in their measures in any sustained way, nor did In thinking through the work of curation, we have they look over them to derive comfort or pleasure. proposed the concepts discerning work and partial data People talked of losing records and in some cases in relation to self-monitoring. In suggesting these we they rediscovered records during the course of the have been influenced by Pink et al.’s (2018) concern interview that they did not remember making or keep- with the way (digital) data is constituted in everyday ing. While these stories were most notable in relation to situations. We find that their ‘concept metaphor’ of paper records, participants also talked of difficulties broken data and focus on the ‘work of repair’ do locating and retrieving digital records. Terry (83, useful analytic work, although they describe only a retired credit controller, white British), for example, small amount of the curatorial work we encountered recounted that he plugs his digital blood pressure mon- in our study. The ideas seem to imply an aspiration for itor into his computer every six months or so to look at completeness which we find often absent. We think that the data, but that when he did this recently, prompted discerning work in the context of self-monitoring pro- by receiving an invitation to participate in the study, he vides a broader term for describing the work that was unable to locate previous readings, telling us ‘I people do to create self-monitoring records. We have must have saved it somewhere, and I can’t find it any- shown how people do not necessarily record all the where’. He attributes this to having acquired a new readings they take, but make decisions about which computer. to record. In this way, records may be selective where One way to interpret these accounts of lost or inac- people record only the data they are happy with, or cessible records is through the lens of broken data, that they feel they need to be reminded of. Here, data characterised by ruptures in people’s records. Yet, in may be partial, but not necessarily broken, in the sense these cases, these ruptures were not accompanied by of representing an incomplete set of the data created efforts to repair the records except perhaps for the pur- and capturing the selectivity or interestedness of the poses of our interviews. Pieces of paper were stashed data recorded. We have also suggested that data may away in bags or with devices, computers and phones be understood as partial when monitoring is undertak- were upgraded, and old ones were discarded or moved en intermittently, perhaps with specific purposes in to transitional spaces in the home such as the loft, just mind (e.g., for a doctor’s appointment) or in seemingly as old diaries were stored in the cellar. We find parallels less patterned ways. We recognise that all data is par- within the sociology of consumption in Sophie tial (Gitelman and Jackson, 2013), but think the notion of partial data, in contrast to broken data, helps to Woodward’s (2015) notion of ‘dormant things’. This keep hold of this sense of selectivity and intermittency. references the accumulation of things not currently A second contribution of this paper is our analysis of being used which may be stored deliberately, but may the material dimensions of curation in relation to self- also be forgotten. Drawing on Woodward, we propose monitoring. Rather than figuring self-monitoring as these accumulated records might be considered dor- exclusively digital or networked, we have documented mant. While an important focus of our analysis has the variety of materials associated with records and been to highlight people’s agential engagements with pointed to the way different materials help to hold or the constitution of data, the notion of dormant records regulate participants’ attention. The visibility of paper helps to acknowledge disengagements and lack of records may facilitate people to monitor together when intentionality. Records may have been created and they are physically co-present. Notebooks or charts stored deliberately but become dormant when they no prominently emplaced might also help participants to longer hold participants’ attention. remember to monitor or keep a commitment in mind. Self-tracking and other apps such as Google Sheets, and Discussion phone memos or notes helped to retain attention This paper introduces and develops an analysis of self- through their emplacement, always present and unlikely monitoring through the lens of curation. In doing so, it to be forgotten. Yet, as exemplified by one participant, builds on and extends a now growing scholarship on the permanent presence of smartphones might risk everyday self-monitoring. By analysing data practices people becoming over-involved in self-monitoring and Weiner et al. 13 the relatively static emplacement of paper records might into continuous monitoring and recording (see also enable monitoring to be kept at a distance. Further, Gorm and Shklovski, 2019), and therefore the data when people shared self-monitoring records, these were they produce may be limited, even if it is in a material mostly in the form of paper records and digital DIY form that can easily flow. The different materials networks. Compared with tracking apps, we suggest enrolled for making and sharing records might further these are perhaps more straightforwardly discreet dampen expectations about the potentials for data because sharing is more readily limited. flows. We propose that curation, as a theory of attention, To what degree does our analysis stem from cases helps brings together different aspects of self- we chose? Blood pressure monitoring and weight/BMI monitoring discussed in the more ethnographically currently involve measuring devices that may be, but informed scholarship. It links the work of making are often not, networked. Whether and how to keep records (e.g., Pink et al., 2018) with the emotional records is relatively open. However, devices are likely aspects of self-monitoring (e.g., Ancker et al., 2015; to become increasingly networked or even wearable, Gorm and Shklovski, 2019; Lomborg et al., 2018; suggesting a move from manual data input to system- Lupton, 2018a, 2018b, 2019; Pantzar and generated records and more continuous measurement. Ruckenstein, 2015; Pink et al., 2017; Ruckenstein, Yet the discerning work in the creation of records we 2014; Urban, 2017) and what scholars have discussed and others have observed suggests that people may still as the different values of self-tracking data (Fiore- exercise a degree of agency over their self-tracking Gartland and Neff, 2015; Lupton, 2018a; Lupton data. Further, as Lupton et al. (2018), Esmonde et al., 2018; Nafus and Sherman, 2014; Sharon and (2020) and Gorm and Shklovski (2019) amply illustrate Zandbergen, 2017). In undertaking curation, people in relation to activity tracking, people may still remove constitute records that are pleasing or communicate devices or delete unwanted data points, or only moni- aspects of their identity or biography (e.g., a trustwor- tor on days that they think are likely to show desirable thy patient, a successful dieter). Materials may help to results. Moreover, as we have discussed, the materiality distance self-tracking so as to reduce obsessiveness or of records is entwined with the practices of monitoring. anxiety, or may act as a reminder of a commitment. In People’s willingness to use specific technologies may this way, we have shown that curation complements depend on the levels of visibility and discretion they other research on how people make sense of or evaluate offer and the degrees to which they are suited to the tracking data (Lupton, 2018a), by underscoring the types of individual or collective practices of monitoring way these valuations may prefigure and shape the gen- we have described. This means that, even when they eration of data in the first place. could use them, people may sometimes eschew digi- While our analysis finds space for the agency of those tal/networked technologies and use analogue/non- who self-monitor in creating records, we have illustrated networked forms of monitoring and recording instead. the difficulties some participants had marshalling unruly What does this all mean for the generation of big materials, as they decided what to keep or tried to remem- data and our understandings of data power (Kennedy, ber if or where they had stored records. Following 2018)? Adopting a curatorial lens helps to unpack pre- Woodward (2015), we have suggested the term dormant cisely which data points are recorded and omitted from records to account for records that have been stored in self-monitoring records, and the ways in which these case of potential future use, as well as those that are data may or may not travel beyond the people who unattended and those that have been forgotten. generate them to be aggregated into big data sets Our analysis has pointed to the way people engage and/or used by other actors. It thus adds specificity and disengage with self-monitoring and the data that it to discussions about data that does not become ‘big’ produces. In this sense, it helps to put data and records and lends nuance to our understanding of the poten- in their place. Accounts of discerning work and partial tials for data flows in practices of self-monitoring. data return a degree of agency to users of self-tracking Acknowledging the importance of discerning work, technologies in the creation and circulation of data, the partiality of data, the varied materiality of self- while being attentive to the constraints imposed by monitoring and the dormancy of some records suggests the diverse materialities involved. Like others we should temper expectations about data flows, data (Didziokaite et al., 2018; Esmonde, 2020; Gorm and power and claims about surveillance and exploitation Shklovski, 2019; Lomborg et al., 2018), we have linked to these. shown that, even where people do record their data in ways that might be compiled by third parties (i.e. Acknowledgements through apps), they do not necessarily give up all We would like to thank all the people who volunteered to their data, and may be selective in what they record. participate in the research for their time and interest. We are They may also not be ‘hooked’ (Lomborg et al., 2018) grateful to the editors of Big Data & Society, three 14 Big Data & Society anonymous reviewers and colleagues for their helpful com- Ontologies of Self-Tracking Practices. London: Emerald, ments, questions and suggestions on earlier versions and pre- pp.137–155. sentations of this article. Esmonde K (2020) ‘There’s only so much data you can handle in your life’: Accommodating and resisting self- surveillance in women’s running and fitness tracking prac- Declaration of conflicting interests tices. Qualitative Research in Sport, Exercise and Health 12 The author(s) declared no potential conflicts of interest with (1): 76–90. respect to the research, authorship, and/or publication of this Esmonde K and Jette S (2018) Assembling the ‘fitbit subject’: article. A Foucauldian-sociomaterialist examination of social class, gender and self-surveillance on fitbit community message Funding boards. Health: An Interdisciplinary Journal for the Social Study of Health, Illness and Medicine 1–16.Epub ahead of The author(s) disclosed receipt of the following financial sup- print, first published 19/9/2018, https://doi.org/10.1177/ port for the research, authorship, and/or publication of this article: This work was supported by the Leverhulme Trust Fiore-Gartland B and Neff G (2015) Communication, medi- under Grant RPG-2015-348. ation, and the expectations of data: Data valences across health and wellness communities. International Journal of Communication 9: 1466–1484. Notes Fox S and Duggan M (2013) Tracking for Health. 1. For a fuller account of these literatures, see Ruckenstein Washington, DC: Pew Research Centre. Gatineau M, Hancock C, Holman N, et al. (2014) Adult Obesity and Schu¨ ll (2017). and Type 2 Diabetes. Oxford: Public Health England. 2. Such as learning and routinising techniques, or making Gitelman L and Jackson V (2013) Introduction. In: Gitelman sense of and assessing the accuracy of the data. 3. Intermittent measurement aligns with Gorm and L (ed) ‘Raw Data’ Is an Oxymoron. Cambridge: MIT Shklovski’s idea of ‘episodic use’, although the timeframes Press, pp.1–14. differ in the studies. Where ‘episodic use’ denotes on and Gorm N and Shklovski I (2019) Episodic use: Practices of care off use across days in the week, intermittent measurement in self-tracking. New Media & Society 21, 11–12, 2505–2521. Hammersley M and Atkinson P (1995) Ethnography: in our study denotes periods of tracking and not tracking Principles in Practice. London: Routledge. across months or years. Harper D (2002) Talking about pictures: A case for photo 4. In further analysis we intend to consider whether we can see filtration work (Nielsen, 2015) in relation to the kinds elicitation. Visual Studies 17: 13–26. of conversations interviewees were hoping to pursue with Henwood F and Marent B (2019) Understanding digital their clinicians and the degree to which these data flow. health: Productive tensions at the intersection of sociology of health and science and technology studies. Sociology of We do not have space here to do justice to such analysis. Health & Illness 41: 1–15. Hooper L, Anderson A, Forster A, et al. (2016) Public ORCID iD Knowledge of the Link Between Obesity and Cancer. Kate Weiner https://orcid.org/0000-0002-5115-8866 London: Cancer Research UK. Kennedy H (2018) Living with data: Aligning data studies References and data activism through a focus on everyday experiences Ajana B (2018) Communal self-tracking: Data philanthropy, of datafication. Krisis: Journal for Contemporary solidarity and privacy. In: Ajana B (eds) Self-Tracking. Philosophy 1: 18–30. Kent R (2018) Social media and self-tracking: Representing Cham: Palgrave Macmillan, pp. 125–141. the ‘health self’. In: Ajana B (ed) Self-Tracking. London: Akrich M (1992) The de-scription of technical objects. In: Palgrave Macmillan, pp.61–76. Bijker W and Law J (eds) Shaping Technology/Building Lomborg S and Frandsen K (2016) Self-tracking as commu- Society: Studies in Sociotechnical Change. Cambridge: MIT Press, pp.205–224. nication. Information, Communication & Society 19: Ancker J, Witteman H, Hafeez B, et al. (2015) “You get 1015–1027. reminded you’re a sick person”: Personal data tracking Lomborg S, Thylstrup NB and Schwartz J (2018) The tem- poral flows of self-tracking: Checking in, moving on, stay- and patients with multiple chronic conditions. Journal of Medical Internet Research 17(8): e202. ing hooked. New Media & Society 20: 4590–4607. Andrejevic M (2005) The work of watching one another: Lupton D (2012) M-health and health promotion: The digital Lateral surveillance, risk and governance. Surveillance & cyborg and surveillance society. Social Theory & Health Society 2: 479–497. 10(3): 229–244. Davis JL (2017) Curation: A theoretical treatment. Lupton D (2013) Fat. London: Routledge. Information, Communication & Society 20: 770–783. Lupton D (2016) The Quantified Self. Cambridge: Polity. Didziokaite_ G, Saukko P and Greiffenhagen C (2018) Doing Lupton D (2017) How does health feel? Towards research on the affective atmospheres of digital health. Digital Health 3: calories: The practices of dieting using calorie counting app MyFitnessPal. In: Ajana B (ed) Metric Culture: 1–11, doi: 10.1177/2055207617701276. Weiner et al. 15 Lupton D (2018a) How do data come to matter? Living and digital living. Big Data & Society 4 1, 1–12, https://doi. becoming with personal data. Big Data & Society 5, 2, org/10.1177/2053951717700924. Piras E and Miele F (2017) Clinical self-tracking and moni- 1–11, https://doi.org/10.1177/2053951718786314. toring technologies: Negotiations in the ICT-mediated Lupton D (2018b) ‘I just want it to be done, done, done!’ Food tracking apps, affects, and agential capacities. patient–provider relationship. Health Sociology Review Multimodal Technologies and Interaction 2: 29. 26(1): 38–53. Lupton D (2019) ‘It’s made me a lot more aware’: A new Rich E and Miah A (2017) Mobile, wearable and ingestible materialist analysis of health self-tracking. Media health technologies: Towards a critical research agenda. International Australia 171(1): 66–79. Health Sociology Review 26: 84–97. Lupton D, Pink S, Heyes LaBond C, et al. (2018) Personal Ruckenstein M (2014) Visualized and interacted life: Personal data contexts, data sense and self-tracking cycling. analytics and engagements with data doubles. Societies 4: 68. International Journal of Communication 12: 19. Ruckenstein M and Schu¨ ll ND (2017) The datafication of Lupton D and Smith GJD (2018) ‘A much better person’: health. Annual Review of Anthropology 46: 261–278. The agential capacities of self-tracking practices. In: Schwennesen N (2019) Algorithmic assemblages of care: Imaginaries, epistemologies and repair work. Sociology Ajana B (ed) Metric Culture: Ontologies of Self-Tracking Practices. London: Emerald, pp.57–75. of Health & Illness 41: 176–192. Martens L, Halkier B and Pink S (2014) Researching habits: Sharon T and Zandbergen D (2017) From data fetishism to Advances in linguistic and embodied research practice. quantifying selves: Self-tracking practices and the other values of data. New Media & Society 19: 1695–1709. International Journal of Social Research Methodology 17: 1–9. Star SL and Strauss A (1999) Layers of silence, arenas of voice: Martens L and Scott S (2004) ESRC report – Domestic kitch- The ecology of visible and invisible work. Computer en practices: Routines, risks and reflexivity ESRC end of award report. Keele, Keele University. Supported Cooperative Work (CSCW) 8(1–2): 9–30. Mayer-Schonberger € V and Cukier K (2013) Big Data: A Urban M (2017) ‘This really takes it out of you!’ The senses Revolution That Will Transform How We Live, Work, and emotions in digital health practices of the elderly. Digital Health 3: 1–16. and Think. Boston, MA: Houghton Mifflin Harcourt. Van Dijck J and Poell T (2016) Understanding the promises Nafus D and Sherman J (2014) This one does not go up to 11: and premises ofonline healthplatforms. Big Data & Society The quantified self movement as an alternative big data practice. International Journal of Communication 8: 3, 1, 1–11, https://doi.org/10.1177/2053951716654173. Weiner K, Henwood F, Will C, et al. (2017) Self-monitoring 1784–1794. for health: Questions for an emerging field. Available at: National Institute of Health and Care Excellence (NICE) https://ssrn.com/abstract=3051201 or http://dx.doi.org/ (2011) Hypertension in Adults, Diagnosis and 10.2139/ssrn.3051201 (accessed 1 March 2020). Management. Clinical Guideline 127. London: NICE. Weiner K and Will C (2018) Thinking with care infrastruc- Neff G and Nafus D (2016) Self-Tracking. Cambridge: MIT Press. tures: People, devices and the home in home blood pressure Nielsen KD (2015) Involving patients with E-health – The monitoring. Sociology of Health & Illness 40(2): 270–282. Whitson J (2013) Gaming the quantified self. Surveillance & dialogic dynamics of information filtration work. Science Society 11: 163–176. & Technology Studies 28: 29–52. Will C, Weiner K, Henwood F, et al. (2019) Relating with O’Riordan K (2017) Unreal Objects: Digital Materialities, data: Stories of self-monitoring with care. In: 4S 2019, Technoscientific Projects and Political Realities. London: New Orleans, USA, 7 September 2019. Pluto Press. Williams R, Weiner K, Henwood F, et al. (2020) Constituting Oudshoorn N (2012) How places matter: Telecare technolo- practices, shaping markets: Remaking healthy living gies and the changing spatial dimensions of healthcare. through commercial promotion of blood pressure moni- Social Studies of Science 42: 121–142. tors and scales. Critical Public Health 30(1): 28–40. Oudshoorn N and Pinch T (eds) (2003) How Users Matter: Williams R, Will C, Weiner K, et al. (forthcoming) The Co-construction of Users and Technology. Cambridge: Navigating standards, encouraging interconnections: MIT Press. Infrastructuring digital health platforms. Information, Pantzar M and Ruckenstein M (2015) The heart of everyday Communication and Society. DOI 10.1080/ analytics: Emotional, material and practical extensions in self-tracking market. Consumption Markets & Culture 18: 1369118X.2019.1702709 (accessed 6 December 2019) Woodward S (2015) The hidden lives of domestic things: 92–109. Pink S and Fors V (2017) Self-tracking and mobile media: Accumulations in cupboards, lofts and shelves. In: Casey New digital materialities. Mobile Media & Communication E and Taylor Y (eds) Intimacies, Critical Consumption and Diverse Economies. Basingstoke: Palgrave Macmillan, 5(3): 219–238. Pink S, Ruckenstein M, Willim R, et al. (2018) Broken data: pp.216–232. Conceptualising data in an emerging world. Big Data & Woodward S (2016) Object interviews, material imaginings Society 5, 1, 1–13, https://doi.org/10.1177/2053951717753228. and ‘unsettling’ methods: Interdisciplinary approaches to Pink S, Sumartojo S, Lupton D, et al. (2017) Mundane data: understanding materials and material culture. Qualitative The routines, contingencies and accomplishments of Research 16(4): 359–374. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Big Data & Society SAGE

Everyday curation? Attending to data, records and record keeping in the practices of self-monitoring:

Download PDF
15 pages

Loading next page...
 
/lp/sage/everyday-curation-attending-to-data-records-and-record-keeping-in-the-v0N03Q80kW

References (63)

Publisher
SAGE
Copyright
Copyright © 2022 by SAGE Publications Ltd, unless otherwise noted. Manuscript content on this site is licensed under Creative Commons Licenses.
ISSN
2053-9517
eISSN
2053-9517
DOI
10.1177/2053951720918275
Publisher site
See Article on Publisher Site

Abstract

This paper is concerned with everyday data practices, considering how people record data produced through self-monitoring. The analysis unpacks the relationships between taking a measure, and making and reviewing records. The paper is based on an interview study with people who monitor their blood pressure and/or body mass index/weight. Animated by discussions of ‘data power’ which are, in part, predicated on the flow and aggregation of data, we aim to extend important work concerning the everyday constitution of digital data. In the paper, we adopt and develop the idea of curation as a theory of attention. We introduce the idea of discerning work to characterise the skilful judgements people make about which readings they record, how readings are presented, and about the records they retain and those they discard. We suggest self-monitoring produces partial data, both in the sense that it embodies these judgements, and also because monitoring might be conducted intermittently. We also extend previous analyses by exploring the broad set of materials, digital and analogue, networked and not networked, involved in record keeping to consider the different ways these contributed to regulating attention to self-monitoring. By paying attention to which data is recorded and the occasions when data is not recorded, as well as the ways data is recorded, the research provides specificity to the different ways in which self-monitoring data may or may not flow or contribute to big data sets. We argue that ultimately our analysis contributes to nuancing our understanding of ‘data power’. Keywords Self-tracking, curation, users, data power, partial data, material methods Introduction: ‘data power’ and the turn to repurposed. Scholars, for example, have drawn atten- everyday monitoring tion to the commodification of these data (e.g., Ajana, 2018; Van Dijck and Poell, 2016) and their potential The growth in apps, wearables and networked technol- contribution to surveillance, allowing, for example, ogies that measure or keep track of a plethora of bodily health professionals access to individuals’ conduct states, actions and experiences, has been referenced in a (Lupton, 2012). The terms dataveillance and lateral number of key discussions within social sciences. Self- surveillance are also used in this context, signalling monitoring has been characterised as disciplining and the more diffuse network of actors among whom normalising, creating particular kinds of neoliberal, self-regulating subjects and reinforcing obligations for self-care (e.g., Lupton, 2016). For some, it is seen as part of the broader ‘datafication of health’ (Mayer- Department of Sociological Studies, University of Sheffield, Sheffield, UK Schonberger and Cukier, 2013; Ruckenstein and School of Law, Politics and Sociology, University of Sussex, Brighton, UK Schu¨ ll, 2017; Van Dijck and Poell, 2016), in which, School of Applied Social Science, University of Brighton, Brighton, UK increasingly, aspects of bodily experience are trans- Corresponding author: formed into quantified data. Self-tracking data may Kate Weiner, Department of Sociological Studies, University of Sheffield, be seen as ‘lively’ (Lupton, 2016, 2018a) as they are Elmfield, Northumberland Road, Sheffield S10 2TU, UK. aggregated, analysed, circulated and potentially Email: k.weiner@sheffield.ac.uk Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (https:// creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Big Data & Society data may be circulated, including individuals who may to mindfulness and awareness of one’s own body and willingly share their data with their own social net- experience. Here, the act of monitoring or recording works (Andrejevic, 2005; Rich and Miah, 2017). may be as, if not more important than, reviewing The foregoing scholarship has been characterised as aggregated data (Nafus and Sherman, 2014; Sharon being centrally concerned with ‘data power’ (Kennedy, and Zandbergen, 2017). Scholars have also drawn 2018). Offering some critique of this, Ruckenstein and attention to the situated and embodied way people Schu¨ ll (2017) call for more attention to everyday make sense of, or assess the value of, tracking data in engagements with data in practice: ‘Scholars who relation to other ways of knowing, as well as the way attend to the power dynamics of datafication have emotions are intertwined with these valuations been faulted for their heavy focus on the oppressive, (Lupton, 2018a; Lupton et al., 2018; Nafus and normalizing, and exploitative forces of datafication and Sherman, 2014). their lack of attention to cases of noncompliance, A third theme relates to the hidden or invisible work appropriation and existential possibility’ (256). (Star and Strauss, 1999) of making data and allowing it Kennedy (2018) similarly argues that discussions of to travel. Pink et al. (2018), for example, are interested datafication tend to leave ‘little scope for agentic in the often obscured or hidden work of mundane engagements with data’ (20). One response has been repair. Introducing the idea of ‘broken data’ and to turn to more ethnographically informed research. ‘repair work’, they argue for ethnographic attention Gaining an understanding of everyday or mundane to the constitution of digital data, describing a process engagements with self-monitoring and the data that of improvisation or repair to fill in the inevitable gaps emerges, it is suggested, is important to inform both in people’s self-tracking data. For example, people use scholarship on, and policy and commercial expecta- multiple devices or use devices in unexpected ways such tions about, the role of data in society (Gorm and as using a step counter to record cycling. In this way, Shklovski, 2019; Kennedy, 2018; Pink et al., 2017; they suggest a focus on making sure data is coherent Weiner et al., 2017). for oneself with no responsibility to provide accurate There is now a blossoming scholarship on everyday data to each device or app (Pink et al., 2018). In her or mundane self-monitoring, often addressing fitness, work on a digitised, algorithmic physical rehabilitation exercise or food tracking, but also other areas including system, Schwennesen (2019) also enrols ‘repair work’ to self-monitoring of chronic health conditions. A number describe the way patients tinker with the system to of related themes are emerging in this scholarship and make it work in practice. Other scholars draw attention here we draw attention to three in particular. The first to the broader work of engaging in self-monitoring, takes seriously people’s emotional engagements with beyond generating data, that remains invisible to its self-monitoring data (Lupton, 2017; Pantzar and proponents (Ancker et al., 2015; Lupton, 2018b, 2019). Ruckenstein, 2015; Ruckenstein, 2014), countering This discussion of emotional engagements with and images of those who self-monitor as impartial, rational different values of data, and the work of making data, actors pursuing health aims (see Lupton, 2016, 2017; go some way to restoring a degree of agency to those Pantzar and Ruckenstein, 2015). This has included dis- who self-monitor. It helps to complicate narratives cussion of the enjoyment or pleasure derived from self- about the disciplining and normalising power of tracking, associated with for example seeing personal self-monitoring practices and about the flows of self- successes or supporting a self-identity as a fit or healthy monitoring data related to the potential for surveillance person, as well as disappointments, worry or frustra- and/or commodification. The ethnographically tion when these are not achieved (e.g., Ancker et al., informed work, in the tradition of user studies 2015; Gorm and Shklovski, 2019; Lomborg and (Oudshoorn and Pinch, 2003), therefore provides Frandsen, 2016; Lomborg et al., 2018; Lupton, empirical research ‘from below’ that helps to nuance 2018a, 2018b, 2019; Pink et al., 2017; Urban, 2017; the ‘data power’ argument. At the same time, some Whitson, 2013). of this work also considers the agency of things/devi- A second theme concerns the different values attrib- ces, which we discuss in more detail below. uted to data derived from self-tracking. In some instan- In this paper, we aim to extend important work ces, value is seen to derive from the (normalised) published in this journal concerning everyday data knowledge claims it allows, for example in the ability practices and, specifically, the everyday constitution to detect patterns, or lend credibility to facts (Fiore- of digital data (Lupton, 2018a; Pink et al., 2018, Gartland and Neff, 2015). However, data has also been 2017). Taking inspiration from and building on the shown to have communicative value, as a way to con- concepts of ‘broken data’ and ‘repair work’ (Pink nect with others, or share intimate stories (Fiore- Gartland and Neff, 2015; Sharon and Zandbergen, et al., 2018), we adopt and develop the idea of curation 2017). In other instances, self-tracking may be linked in relation to self-monitoring, using material from our Weiner et al. 3 study of the everyday practices of monitoring blood emotions associated with tracking and the value of the pressure and/or body mass index (BMI). data. In our discussion above, we have illustrated how people may gain pleasure or satisfaction and are able to communicate particular stories about themselves Adding a curatorial lens through the (hidden) work of curating their records. Curation is multivalent. Davis (2017) offers a theoret- In this context, curation helps to bring together the ical treatment of digital curation, describing curation as three emerging themes we identified relating to self- a theory of attention, concerned with how people allo- monitoring, linking the hidden work of making data cate and control attention. Drawing on examples relat- with the emotional aspects and the value of the data. ing to social media, she suggests that curation How does curation relate to the notions of repair ‘broadly .. . refers to the discriminate selection of mate- work (Pink et al., 2018) and filtration work (Nielsen, rials for display [online]’, where ‘productive curation’ 2015)? All these concepts help to bring to light the involves deciding what to ‘document, make, share, and hidden work of making data. While curation signals with whom’ and is integral to performances of self for the possibility of selectivity, repair work is suggestive oneself and for others (Davis, 2017: 771, 772). of an ultimate hope of completeness. Yet it does While there has not been a thoroughgoing applica- involve putting materials together for one’s own satis- tion of the notion of curation to self-monitoring, a faction. Where curation may be broadly communica- sense of this selectivity is present in some emerging tive, part of identity construction for oneself and studies of everyday self-monitoring practices. Kent conveying this to others, filtration work in Nielsen’s (2018: 67), in a study of how self-trackers represent (2015) account is solely orientated to others. It is con- ‘health’ through social media, discusses the way her cerned with opening up or closing down particular con- participants construct an appropriate self-tracking per- versations with specific actors. In this paper, we would sona ‘through careful inclusion and exclusion of certain like to propose curation as an overarching concept, health information’. Studies of calorie and of fitness where repair work and filtration work offer particular tracking have documented the way participants may examples of this concept. Curation helps to illuminate manipulate data input, for example not recording every- the hidden or underarticulated work of producing and thing consumed on days of excess (Didziokaitee _ t al., sharing self-monitoring records. It, thus, helps to bring 2018; Lomborg et al., 2018), or not saving ‘unflattering’ the agency of those who monitor into view. Yet the runs (Esmonde, 2020: 84). They might also engage in concept of curation does not only illuminate the work ‘episodic use’ (Gorm and Shklovski, 2019) of tracking of human actors, but may also acknowledge the work technologies, recording calories or wearing fitness track- of materials. ers only on days when they anticipate good, interesting or useful numbers (Didziokaite_ et al., 2018; Esmonde, Curation as socio-material practice 2020; Gorm and Shklovski, 2019; Lupton et al., 2018). In this way, participants are selective in the records they In her discussion of curation, Davis (2017: 775) attends create either imagining an external audience, or support- to the agency of materials through making a distinction ing their motivation and protecting themselves from dis- between ‘human’ and ‘machine’ curation, discussing appointing outcomes. how the design of platforms and algorithms shapes In Nielsen’s (2015) work, the external audience is and constrains the way users produce and consume particularly important for patients, who she suggests online content. Pink et al. (2017: 3) make a similar undertake ‘filtration work’ when making entries to a move in relation to self-monitoring data, wanting to new e-health system. This involves being selective in rela- take ‘users’’ perspectives seriously but also to ‘decentre tion to what information to provide and has a particular, the human’, suggesting that ‘personal data’ is ‘consti- dialogic, orientation; patients imagined the receiver, and tuted and experienced between human and digital/algo- shaped their entries in line with conversations they hoped rithmic devices and processes’. The breakages in data to pursue or avoid. Work on the development of a clin- they describe, when devices are not charged or lack connection, when software updates make existing devi- ical self-monitoring system for diabetes similarly showed how patients might decline to share data or respond to ces redundant, or devices track some activities but not clinicians’ messages (Piras and Miele, 2017). All of the others, draw attention to the way devices and platforms studies discussed so far illustrate selectivity in records shape the production of self-monitoring records. As we made or shared, suggesting there is value in the concept have discussed, their work also documents the way of curation in relation to self-monitoring. users may attempt to get around material constraints In considering the value of curation as a conceptual through their repair work (Pink et al., 2018). In previ- lens in this context, we need to acknowledge that cura- ous research we have drawn attention to the multi-user torial work is suffused with and inseparable from the functionality of some devices for measuring blood 4 Big Data & Society pressure and weight, to highlight the way these shape, who generated them. Ultimately we propose curation or script (Akrich, 1992) particular ways of recording can therefore be helpful in interrogating concerns with and sharing data (Williams et al., 2020). These sorts data power. of socio-material analyses illuminate the way platforms and devices shape the production and management of Methods self-monitoring records without resorting to technological The paper is based on a UK study involving interviews determinism. They allow space for both users and tech- with people who self-identified as monitoring their blood nologies (and their developers) to have agency (Henwood pressure or BMI/weight. Our engagement with self- and Marent, 2019; Lupton, 2018a; Oudshoorn and monitoring stemmed from our broader interest in every- Pinch, 2003). day health practices, the use of health technologies in Yet, in considering the material dimensions of cura- domestic settings and the way these might redistribute tion we would like to draw attention to the kinds of health work between the home and the clinic (see self-monitoring so far discussed in critical scholarship. Henwood and Marent, 2019; Weiner et al., 2017; This has, with notable exceptions (e.g., Lupton and Weiner and Will, 2018; Williams et al., 2020, forthcom- Smith, 2018; O’Riordan, 2017), tended to focus on dig- ing). Home blood pressure monitoring and BMI moni- ital and networked types of self-tracking involving, toring offer particularly interesting cases in the way they especially, fitness and diet apps and wearables. Yet, blur the boundary between the clinic and the home. as Neff and Nafus (2016: 98) note: ‘self-tracking tools In the UK there are well established consumer mar- do not have to be fancy’ and might include low tech kets for both blood pressure and BMI monitoring. A materials such as pen and paper. Indeed, Fox and range of devices are available to purchase in supermar- Duggan’s (2013) oft-cited research reported that the kets, pharmacies and online retailers, such as digital majority of Americans who track a ‘health indicator’ blood pressure monitors, digital and analogue weighing did this with pen and paper or ‘in the head’. Rather scales and digital body analysis scales. These products than equating self-tracking with digital and networked include both stand-alone and networked devices and self-tracking, we think it is important to consider the may be accompanied by proprietary apps, but also wider materials and technologies and their place in paper booklets or diaries for recording readings (see the practices of self-monitoring. What, for example, Williams et al., 2020, forthcoming, for further analysis are the implications of different materials for data of this market). There are also other apps to calculate/ flows? Further, since curation is concerned with allo- track BMI or track blood pressure, such as cating and controlling attention (Davis, 2017), are MyFitnessPal and Apple Health, where data may be there material dimensions to paying attention, avoiding entered manually or pushed through from networked noticing or being inattentive to self-monitoring? devices, as well as websites providing online BMI calcu- In sum, we propose that a curatorial lens facilitates lators. Both forms of monitoring have clear links to the exploration of the way self-monitoring data is con- clinical interests. Monitoring blood pressure is well stituted in practice, illuminating the work of both established in clinical practice and self-monitoring is humans and materials. Further, the idea of curation increasingly sanctioned as one response to white coat helps to link the work of making data with the emo- hypertension (doctor-induced high blood pressure) tional aspects of self-monitoring and the value of the (National Institute of Health and Care Excellence data. In our analysis, we adopt this lens to develop a (NICE), 2011). Clinical concern with BMI and weight socio-materialist account (Henwood and Marent, 2019; relate to obesity, and to risks of diabetes and cancer and Weiner and Will, 2018; Williams et al., 2020, forthcom- forms part of public health messages (Gatineau et al., ing) of everyday data practices relating to self- 2014; Hooper et al., 2016). In sum, both have clear clin- monitoring, exploring what records people keep, what ical relevance and established self-monitoring markets. materials are involved and whether and how records While our study involves voluntaristic self-monitoring, are shared. We suggest that this curatorial approach we acknowledge the non-innocence of self-monitoring helps to clarify the relationship between self- monitoring and the accrual and flow of data. By technologies and their links with broader socio-political paying attention to which data is or is not recorded, contexts. Notwithstanding the contested history of BMI, as well as the ways data is recorded, the research pro- the measure links with weight management which is asso- vides specificity to the ways in which self-monitoring ciated with strong narratives of personal responsibility, may or may not contribute to big data sets in different guilt and shame (Lupton, 2013). This and other forms ways. It allows reflection on the ‘liveliness’ (Lupton, of tracking intended to work on the body relate to gen- 2018a) of self-monitoring data, in terms of their poten- dered norms of beauty and fitness as well as to health tial to be circulated, reconfigured and monetised, and (Esmonde, 2020). Relatedly, there are clinical/psycholog- do so in ways that might act back on the individuals ical concerns about the possible links between Weiner et al. 5 food-tracking apps, such as MyFitnessPal, and eating In this paper, we focus on self-monitoring data prac- disorders (Lupton, 2018b). Discourses relating to tracking tices and the materials this involves. It is not our inten- are also infused with assumptions about people’s capacity tion to provide a definitive definition of self-monitoring to incorporate tracking which do not chime with gen- and we do not see an obvious difference between this dered, classed or marginalised experiences of daily life and self-tracking. Resonating with other research (e.g., or work routines (Ancker et al., 2015; Esmonde and Lupton, 2019) we followed an emic approach, keeping Jette, 2018; Lupton, 2018b). At the same time, there are our recruitment material broad and allowing people to concerns that fitness tracking may be pushed or imposed identify themselves as engaging in self-tracking in order (Lupton, 2016) by healthcare insurers or employers to study what this involves for them. Lupton (2016) (Ajana, 2018; Esmonde and Jetter, 2018; Lupton, 2016). proposes that self-tracking entails ‘practices in which We note, however, the relevance of these concerns is lim- people knowingly and purposively collect information ited in the UK context, where healthcare is largely about themselves which they review and consider apply- accessed through a universal, national, government- ing to the conduct of their lives’ (2). In our analysis, we funded system. Even so, self-tracking is likely to be explore the distinctions and relationships between these linked with uneven and differentiated experiences and different potential aspects of self-monitoring focussing effects. on three main themes: the relationship between taking In our study, we made efforts to recruit a diverse and recording measures, how and where records are sample. Following institutional ethics approval, we made, and storing and reviewing records. advertised on email lists at three UK universities and noticeboards across campuses, at older people’s groups and at community centres in less-advantaged areas. Findings The advert sought people who identified themselves as ‘measuring and keeping track’ of either their blood 1. The relationship between taking and recording pressure or BMI. In this paper, we draw on 67 inter- measures views conducted with 81 people, including 14 inter- a. No records views with couples. Participants varied in terms of We start our analysis by considering the approxi- age, sexuality, ethnicity, socio-economic background and health. All had acquired monitoring devices for mately one-quarter of our participants who took meas- themselves and no one reported acquiring these from ures but did not record these. Understanding curation employers or clinicians. While we were alive to issues of to be concerned with attention, we consider what diversity, we did not find these significant in the current people are attending to in these cases. In other words, analysis, although they are more central to other if they are not recording their data, what are they doing themes (see Will et al., 2019). when they self-monitor? Sometimes participants did In interviews, we asked people how they came to this for reassurance, just wanting to know if their monitor or acquire a device, what they do or do not blood pressure or BMI was in the normal range and do with it and who else might use it, how this may have they were able to recall this without needing to remem- changed over time and with whom data is shared. The ber the precise number or to keep records. People limitations of ‘conventional’ social science methods talked of monitoring ‘to keep an eye on something’ such as interviews for researching everyday life are or ‘for peace of mind’, illustrating the emotional reso- well rehearsed (Martens et al., 2014: 3). People may nance of the practice. For example, Gary explains he find it difficult or are unable to talk about certain ele- has anxiety issues and uses his blood pressure monitor ments of their everyday practices, in particular embod- for reassurance: ‘I need to know if there’s something ied, tacit and affective aspects (Martens et al., 2014; wrong you know ... So if I think I’ve got a bit of a Martens and Scott, 2004). The use of material objects headache or I get some palpitations I’ll check it’ or photos in interviews can provide an aid to memory (Gary, 45, school administrative officer, white British). and reflexivity that interviews alone cannot elicit (e.g., Gary does not record his readings, and cannot recall Harper, 2002; Woodward, 2016). In our interviews, we the precise numbers from the last time he used his monitor invited participants to demonstrate their monitors and (four weeks before the interview), but knows they were talk through any records they kept and where these ‘under the 140 and 90’ which he called ‘the bench mark’. were stored. This helped both to prompt reflection For other participants, monitoring was concerned and tie practices to particular time periods and with managing day-to-day conduct. Linda, for exam- events. We analysed the interviews thematically ple, does not keep a record of her weight and uses BMI (Hammersley and Atkinson, 1995), collaboratively as trigger to take action when she sees herself ‘creeping’ developing a coding frame, which synthesises our the- oretical grounding with emergent themes. near to the boundary between normal and overweight: 6 Big Data & Society If I see myself creeping .. . I haven’t actually got to the I don’t write down the good numbers, I only write point of going into the next category .. . so it’s sort of down the bad numbers. So when it’s fine I don’t time to take some action in the sense of, you know, just bother, but when it’s bad I think I probably should, cutting back on what I’m eating, being more careful because I’ve got a rubbish memory I think I probably about portion sizes, that sort of thing. (Linda, 67, need to keep a record of that. (Annie, 45, university retired Further Education teacher, mixed heritage) administrative officer, white British) Occasionally, blood pressure measures also resulted in So in contrast to Ayo, who only records her measures immediate action such as drinking some green tea or when they go in the right direction, Annie only records trying to relax. ‘bad’ numbers. Participants measuring blood pressure In these cases, people were not seeking to under- also discussed processes of selecting or averaging mul- stand patterns in their data, but to attend to their tiple readings for recording, e.g. best of three. We also immediate bodily status for reassurance or potential encountered occasional stories of participants curating actions. Self-monitoring helps to address questions charts to make them more meaningful or pleasing, for such as: how am I today? Am I stressed? Do I need example by removing outlying data points or choosing to go to the doctors? Should I eat less today? This helps the span of time and the right axis. Gareth, for exam- to explain why some participants cannot recall precise- ple, showed a graph of his weight on his Google Fit ly or do not record or review their data. app, illustrating how the axis changes when he selects different years, and how the falling line pleases him: b. Discerning work and partial data ‘That’s me overall graph. I’m quite pleased with how that’s steadily falling. Especially when I put that year in Yet among those who did record data, participants it puts a different axis on it and it’s whoosh’ (Gareth, described selective recording, including not recording 58, property maintenance engineer, white British). particular readings. Ayo weighs herself on stand- The foregoing accounts illustrate how participants alone digital scales, and records this into her are selective in the way they compile data into records. Samsung Health app, which calculates her BMI. We propose that they are undertaking discerning work She told us she only records her weight when it had (rather than ‘repair work’), making judgements about gone down: which readings are useful, or worth remembering or drawing attention to, and how to process or clean read- Ayo: When I weigh and it’s more and I’ve put on ings to make best sense of them. Further, rather than weight I don’t enter into the app to update my the metaphor of ‘broken data’, we propose the idea of BMI .. . I only do it when I lose weight partial data may be more apposite in these instances. Partial here has a double meaning, understood in the Interviewer: So how come you don’t put it in? sense that only some of the data get recorded, but also in the sense of interested or partisan, in contrast to Ayo: Because it makes me sad ... The fact that I’ve put impartial or neutral. The readings written down or on weight, which is not what I want. I want to lose entered into apps may only be a subset of the readings weight .. . So that’s sad for me so I don’t bother putting taken and may be selected for very particular reasons. it onto .. . each time I found out I’ve lost weight then I add my weight here ... I want to see that I’m losing c. Intermittent measures and partial data weight on my app’. (Ayo, 33, university researcher, black African) In other interviews, participants told us of intermit- tent measurement which lead to intermittent records. In Ayo’s account again underscores the affective reso- interviews relating to blood pressure, participants some- nance of self-monitoring records, which have the times compiled records for time-limited periods specifi- capacity to make her ‘sad’ if they go in the wrong direc- cally to take to clinical consultations. For example, Fred tion. This chimes with studies of calorie and fitness records his blood pressure for one month prior to his tracking (Didziokaite et al., 2018; Esmonde, 2020; appointment compiling a spreadsheet which he prints to Gorm and Shklovski, 2019). take to his doctor’s appointment. Such intermittent Other participants reported, in a more pragmatic records resulting from intermittent measuring might be vein, that there was no point noting down weight or understood as a second form of partial data, in the sense blood pressure when it was stable, only noting when that it is compiled from time to time. there is change. Annie records her blood pressure on Fred’s account is useful in reinforcing an important scraps of paper and in a booklet. She told us she only point of our analysis by demonstrating the consider- notes this down when her reading is high: able work of making records. Fred told us that he Weiner et al. 7 records his blood pressure on pieces of paper by hand we have described, or in the sense that they measure, (Figure 1), transcribing these data into a spreadsheet and therefore record, intermittently. From a user per- which he compiles specifically for his appointment spective, records may not be intended to be comprehen- (Figure 2). He labels the spreadsheet with the name of sive or continuous, so here there is no ‘broken data’ and his GP practice and the date of his appointment. This therefore no need for ‘repair work’. This means that, helps to illustrate the curatorial work involved in beyond the data breakages identified by Pink et al. making data ready to share. In moving from hand- (2018), there are other reasons why data may not flow written slips of paper to a neatly presented spreadsheet, seamlessly from measurement to an individual’s records Fred demonstrates the skilfulness and probity of his self- to be aggregated by third parties. All of this should act monitoring. The materiality of the spreadsheet commu- as a gentle corrective to expectations about the exploi- nicates ‘I am a responsible patient, you can take my tation of such data in existing literature (see Ruckenstein readings seriously’. and Schu¨ ll, 2017; Van Dijck and Poell, 2016). So far we have shown that the practices of taking a reading do not always lead to keeping a record of those data. For some of our participants, self-monitoring did 2. How and where to record not involve making any records. Where people do keep records, the readings entered into apps or recorded The analysis so far has mostly concerned which measures people record, but has also touched on the importance of elsewhere may only be a subset of the readings taken. We have suggested the term discerning work to describe how records are made and presented, particularly in our the skilful judgements people make about which data discussion of Fred. We now turn to this theme in more to record, which to omit, and how to process and pre- detail, to expand on issues around how and where people sent records that differentially draw attention to suc- record and the materials involved. We will consider the cesses, warning signs or the credibility of the person socio-material arrangements in this curatorial work that making the record. We have also suggested that partic- draw attention to or deflect attention from different ipants may create partial data both in the sense that aspects of self-monitoring. they may choose only to record some of the readings they take, selecting these through the discerning work a. Visibility and being reminded Figure 1. Fred’s handwritten notes, retrieved from the waste paper bin during the interview. 8 Big Data & Society Figure 2. Fred’s spreadsheet compiled for his doctor’s appointment, practice name blanked out. While half of the participants in the study had expe- with her sister and sister’s wife. They met on Saturday rience of using an app to track BMI, the visibility of mornings to record their weight and kept a joint record paper records emerged as an important theme. For on a sheet of paper. The record had been set up on a example, Becky told us she was losing weight together spreadsheet, but this was printed off and weights Weiner et al. 9 written on by hand (Figure 3). When asked why they resonates with some studies of digital self-tracking did not simply enter the data onto the spreadsheet, (Lomborg et al., 2018; Pink and Fors, 2017, cf. Kent, Becky responded: ‘I think because it was going to be 2018), placing into question expectations of widespread a group thing that we could all jot it down while we lateral surveillance (Rich and Miah, 2017). Here we were together. So I think that’s why I have a physical extend the existing analysis by considering not only sort of ... ’ (Becky, 36, charity researcher, white with whom participants share, but also the materiality of sharing. British). The visible emplacement of monitoring devices in In contrast to the digital co-presence discussed by Pink and Fors (2017), where people who are physically particular domestic spaces, for example close at hand separate share data and are present together online, on a table next to a favourite armchair, may encourage self-monitoring in our study involved physical co- people to monitor (Weiner and Will, 2018). In the same presence where different materials come to the fore. way, the emplacement of self-monitoring records, such A paper record in this account appears to allow these as a chart pinned to a wall or a record on a mobile three women to participate together and to attend to phone that is always to hand, might act as a reminder their data collectively. in different ways. Participants told us that leaving It was striking that records for sharing with part- paper records and charts somewhere visible within ners, relatives and friends tended to be paper, charts the home helped to remind them to monitor or and/or DIY forms of digitally networked communica- helped to keep commitments in mind. Becky, for exam- tion such as texting or setting up WhatsApp groups. ple, told us that the shared record she made with her We encountered very little discussion of sharing sister and sister-in-law was pinned to her notice board through broader social media or proprietary self- in her home office. She explained that this was visible monitoring apps that would facilitate sharing with enough to remind herself she was trying to lose weight, wider social networks or publics, even where partici- but not so public that visitors to the house would read- pants had devices with the capacity to do so. The mate- ily see it (compared with, for example, pinning it to the rials our participants discussed appeared to allow them fridge in the kitchen). It is placed to hold her attention to do things together with limited, selected others while avoiding drawing the attention of visitors. This (friends, family), and offer each other encouragement, concern with the emplacement of the chart suggests while precluding broader attention. This finding that even paper records have the capacity to act as Figure 3. Becky’s ‘group thing’, names blanked out. 10 Big Data & Society ‘indiscreet technologies’ (see Oudshoorn, 2012), she continued to record her calorie intake in the app, she making public aspects of identity or practice that had stopped recording her weight in this because she people would rather keep to themselves. In Becky’s found it contributed to her becoming ‘obsessive’ about case, the materiality and emplacement of the record it. At that point she told us she preferred to record her lend themselves both to monitoring in a group and to weight on a weekly basis in a notebook: keeping a project in mind, while allowing the record to remain relatively private. So although I weigh myself at the minute, I’m not put- Others told us that they recorded on the phone ting into MyFitnessPal because I found I was getting because it is always with them, unlikely to be forgotten maybe a bit too – I started weighing myself every day or because it formed a convenient mode for transport- and I may have got a little bit too obsessive about ing records. Bella, for example, related that it was not it .. . I felt I’ve gradually started being calmer about until she got her smartphone and found a free blood it .. . So I thought I’m going to gradually start doing pressure app that she started to record her readings. all the things that I used to do again which is weigh Before that she had not made a record: myself once a week. (Andrea, 27, university adminis- trator, white British) because it was a pain in the neck .. . Because I had to keep writing it down and then remember to write it While MyFitnessPal is designed to allow and encour- down and find the piece of paper, like that. So I was ages daily recording of weight, Andrea’s ‘weekly weigh really happy when I found the app. (Bella, 57, charity in’ notebook is quite literally scripted for weekly administrator, white British) recording (Figure 4). In Andrea’s case, she recounts being overly concerned with, or over-attentive to, her In contrast to a ‘piece of paper’, Bella’s phone is always weight and in changing to a different way of recording nearby. Yet recording on phones did not always tries to regulate this over-attentiveness. involve using proprietary tracking apps, as participants In this section, we have paid particular attention to also told us they used note apps or Google Sheets to the materiality of records, looking at the different ways record self-monitoring data. Samuel, for example, these contribute to paying and regulating attention to talked about recording his blood pressure readings in self-monitoring. Like others (Lomborg et al., 2018; a note app, to take to his doctor’s appointment: Pink and Fors, 2017), we found little sharing through proprietary self-monitoring apps, but the employment Interviewer: you said you’d put the records on your of other materials – paper, spreadsheets, WhatsApp phone for a time. Was there any reason for that? groups – which work to limit attention to small, select- ed groups of people. The visibility and emplacement of Samuel: Only for ease of transport, I knew I’d have my paper records may facilitate collective practice within phone with me. I haven’t kept them as a record, it was the home and help to remind people to monitor or keep just a way of transporting information to the surgery commitments in mind. The nearness of phones facili- with me. tates the recording of readings. Yet in one case, the immediacy of a mobile app was associated with over- Interviewer: Okay, so was it an app for the blood attentiveness, and a paper record helped to remedy pressure? this. More baldly, the diverse materiality of self- monitoring records, including, but not limited to pro- Samuel: No, it was just a note. (Samuel, 62, university prietary self-monitoring apps, suggests that self- counsellor, white British) monitoring data may not always be readily compiled or harvested by third parties, placing brakes on the The constant presence of phones facilitated the record- liveliness (Lupton, 2019) of this data. ing of measures and made sure records were always at hand. 3. Storing and reviewing records a. Broken data and repair work b. Not being reminded Curation involves both what people make and what they keep and display. This relates to what they would While a few participants told us they had moved to like to remember or be reminded of. We found occasion- apps from other ways of recording, in one instance a al stories of people going to efforts to retrieve data from participant had moved away from an app, precisely different sources in order to retain a complete record. because of the attention it demanded. Here again the emotional resonance of tracking comes into view. These may be understood as examples of the repair Andrea told us she used MyFitnessPal and that, while work and broken data proposed by Pink et al. (2018). Weiner et al. 11 Figure 4. Andrea’s ‘weekly weigh in’ note book. John (55, IT support, white British) had recorded his b. Dormant records weight and BMI on a weekly basis for the last decade using a number of different platforms. Initially he used a In considering self-monitoring through the lens of weight loss website called Weightloss Resource, because curation, we have so far discussed record keeping prac- tices in fairly deliberate terms. We have portrayed self- his wife was already subscribed to this. He ended this monitoring records as being created and shaped subscription in 2014, downloading his data to a Google Sheets spreadsheet, and moved to MyFitnessPal, which through a combination of the discerning work of he used for 10 months before getting a Fitbit. He told us humans and the materiality of the devices and broader that to export his 10 months of data from MyFitnessPal technologies involved. Our final brief section provides a would incur a fee, which he was not prepared to pay, caveat to this view, suggesting that sometimes the although he lamented the ‘gap’ in his data. human and material elements combined in such a We asked him at different points in the interview way that participants found it difficult to keep track why he had downloaded his data and if this was impor- of their records. For example, Tony keeps records of tant to him. His responses suggested an emotional con- monitoring his blood pressure in a rather ‘haphazard’ nection to graphs as ‘comforting’. They also posed a fashion on various slips of paper and backs of enve- link between records and biography – ‘a reminder of lopes. He told us he stored most of his records in a bag where you’d been and where you’d come to’. John but that he threw a lot of these away: related this to one particularly significant time in his life, during which one of his daughters was diagnosed Interviewer: Do you normally keep the readings in with and treated for a serious illness. For John, retain- that? ing and looking over his records appeared to be both a way to celebrate his successes in weight control and to Tony: Yeah in a rather haphazard fashion. On bits of notes .. . I just used to write them down on bits of paper remember how he and his family had come through his daughter’s illness. This underscores the emotional and and shove them into this bag. And then I was packing communicative aspect of these records. this all away one day and suddenly thinking gosh 12 Big Data & Society there’s an awful lot of ancient results here, I’m never through the concept of curation, we illuminate both the going to do anything with these and I remember I human work involved in making and retaining records, chucked a load away. (Tony, 54, electronic engineering while, at the same time, taking seriously the role of lecturer, white British) materials. Understanding curation as a theory of atten- tion, we have analysed the different ways both humans The emplacement of records ‘shoved’ into a bag means and materials are implicated in drawing attention to, or that they do not seem to hold Tony’s attention. Like detracting attention from, the practices of self- Tony, participants were often not trying to discern pat- monitoring and the data these create. terns in their measures in any sustained way, nor did In thinking through the work of curation, we have they look over them to derive comfort or pleasure. proposed the concepts discerning work and partial data People talked of losing records and in some cases in relation to self-monitoring. In suggesting these we they rediscovered records during the course of the have been influenced by Pink et al.’s (2018) concern interview that they did not remember making or keep- with the way (digital) data is constituted in everyday ing. While these stories were most notable in relation to situations. We find that their ‘concept metaphor’ of paper records, participants also talked of difficulties broken data and focus on the ‘work of repair’ do locating and retrieving digital records. Terry (83, useful analytic work, although they describe only a retired credit controller, white British), for example, small amount of the curatorial work we encountered recounted that he plugs his digital blood pressure mon- in our study. The ideas seem to imply an aspiration for itor into his computer every six months or so to look at completeness which we find often absent. We think that the data, but that when he did this recently, prompted discerning work in the context of self-monitoring pro- by receiving an invitation to participate in the study, he vides a broader term for describing the work that was unable to locate previous readings, telling us ‘I people do to create self-monitoring records. We have must have saved it somewhere, and I can’t find it any- shown how people do not necessarily record all the where’. He attributes this to having acquired a new readings they take, but make decisions about which computer. to record. In this way, records may be selective where One way to interpret these accounts of lost or inac- people record only the data they are happy with, or cessible records is through the lens of broken data, that they feel they need to be reminded of. Here, data characterised by ruptures in people’s records. Yet, in may be partial, but not necessarily broken, in the sense these cases, these ruptures were not accompanied by of representing an incomplete set of the data created efforts to repair the records except perhaps for the pur- and capturing the selectivity or interestedness of the poses of our interviews. Pieces of paper were stashed data recorded. We have also suggested that data may away in bags or with devices, computers and phones be understood as partial when monitoring is undertak- were upgraded, and old ones were discarded or moved en intermittently, perhaps with specific purposes in to transitional spaces in the home such as the loft, just mind (e.g., for a doctor’s appointment) or in seemingly as old diaries were stored in the cellar. We find parallels less patterned ways. We recognise that all data is par- within the sociology of consumption in Sophie tial (Gitelman and Jackson, 2013), but think the notion of partial data, in contrast to broken data, helps to Woodward’s (2015) notion of ‘dormant things’. This keep hold of this sense of selectivity and intermittency. references the accumulation of things not currently A second contribution of this paper is our analysis of being used which may be stored deliberately, but may the material dimensions of curation in relation to self- also be forgotten. Drawing on Woodward, we propose monitoring. Rather than figuring self-monitoring as these accumulated records might be considered dor- exclusively digital or networked, we have documented mant. While an important focus of our analysis has the variety of materials associated with records and been to highlight people’s agential engagements with pointed to the way different materials help to hold or the constitution of data, the notion of dormant records regulate participants’ attention. The visibility of paper helps to acknowledge disengagements and lack of records may facilitate people to monitor together when intentionality. Records may have been created and they are physically co-present. Notebooks or charts stored deliberately but become dormant when they no prominently emplaced might also help participants to longer hold participants’ attention. remember to monitor or keep a commitment in mind. Self-tracking and other apps such as Google Sheets, and Discussion phone memos or notes helped to retain attention This paper introduces and develops an analysis of self- through their emplacement, always present and unlikely monitoring through the lens of curation. In doing so, it to be forgotten. Yet, as exemplified by one participant, builds on and extends a now growing scholarship on the permanent presence of smartphones might risk everyday self-monitoring. By analysing data practices people becoming over-involved in self-monitoring and Weiner et al. 13 the relatively static emplacement of paper records might into continuous monitoring and recording (see also enable monitoring to be kept at a distance. Further, Gorm and Shklovski, 2019), and therefore the data when people shared self-monitoring records, these were they produce may be limited, even if it is in a material mostly in the form of paper records and digital DIY form that can easily flow. The different materials networks. Compared with tracking apps, we suggest enrolled for making and sharing records might further these are perhaps more straightforwardly discreet dampen expectations about the potentials for data because sharing is more readily limited. flows. We propose that curation, as a theory of attention, To what degree does our analysis stem from cases helps brings together different aspects of self- we chose? Blood pressure monitoring and weight/BMI monitoring discussed in the more ethnographically currently involve measuring devices that may be, but informed scholarship. It links the work of making are often not, networked. Whether and how to keep records (e.g., Pink et al., 2018) with the emotional records is relatively open. However, devices are likely aspects of self-monitoring (e.g., Ancker et al., 2015; to become increasingly networked or even wearable, Gorm and Shklovski, 2019; Lomborg et al., 2018; suggesting a move from manual data input to system- Lupton, 2018a, 2018b, 2019; Pantzar and generated records and more continuous measurement. Ruckenstein, 2015; Pink et al., 2017; Ruckenstein, Yet the discerning work in the creation of records we 2014; Urban, 2017) and what scholars have discussed and others have observed suggests that people may still as the different values of self-tracking data (Fiore- exercise a degree of agency over their self-tracking Gartland and Neff, 2015; Lupton, 2018a; Lupton data. Further, as Lupton et al. (2018), Esmonde et al., 2018; Nafus and Sherman, 2014; Sharon and (2020) and Gorm and Shklovski (2019) amply illustrate Zandbergen, 2017). In undertaking curation, people in relation to activity tracking, people may still remove constitute records that are pleasing or communicate devices or delete unwanted data points, or only moni- aspects of their identity or biography (e.g., a trustwor- tor on days that they think are likely to show desirable thy patient, a successful dieter). Materials may help to results. Moreover, as we have discussed, the materiality distance self-tracking so as to reduce obsessiveness or of records is entwined with the practices of monitoring. anxiety, or may act as a reminder of a commitment. In People’s willingness to use specific technologies may this way, we have shown that curation complements depend on the levels of visibility and discretion they other research on how people make sense of or evaluate offer and the degrees to which they are suited to the tracking data (Lupton, 2018a), by underscoring the types of individual or collective practices of monitoring way these valuations may prefigure and shape the gen- we have described. This means that, even when they eration of data in the first place. could use them, people may sometimes eschew digi- While our analysis finds space for the agency of those tal/networked technologies and use analogue/non- who self-monitor in creating records, we have illustrated networked forms of monitoring and recording instead. the difficulties some participants had marshalling unruly What does this all mean for the generation of big materials, as they decided what to keep or tried to remem- data and our understandings of data power (Kennedy, ber if or where they had stored records. Following 2018)? Adopting a curatorial lens helps to unpack pre- Woodward (2015), we have suggested the term dormant cisely which data points are recorded and omitted from records to account for records that have been stored in self-monitoring records, and the ways in which these case of potential future use, as well as those that are data may or may not travel beyond the people who unattended and those that have been forgotten. generate them to be aggregated into big data sets Our analysis has pointed to the way people engage and/or used by other actors. It thus adds specificity and disengage with self-monitoring and the data that it to discussions about data that does not become ‘big’ produces. In this sense, it helps to put data and records and lends nuance to our understanding of the poten- in their place. Accounts of discerning work and partial tials for data flows in practices of self-monitoring. data return a degree of agency to users of self-tracking Acknowledging the importance of discerning work, technologies in the creation and circulation of data, the partiality of data, the varied materiality of self- while being attentive to the constraints imposed by monitoring and the dormancy of some records suggests the diverse materialities involved. Like others we should temper expectations about data flows, data (Didziokaite et al., 2018; Esmonde, 2020; Gorm and power and claims about surveillance and exploitation Shklovski, 2019; Lomborg et al., 2018), we have linked to these. shown that, even where people do record their data in ways that might be compiled by third parties (i.e. Acknowledgements through apps), they do not necessarily give up all We would like to thank all the people who volunteered to their data, and may be selective in what they record. participate in the research for their time and interest. We are They may also not be ‘hooked’ (Lomborg et al., 2018) grateful to the editors of Big Data & Society, three 14 Big Data & Society anonymous reviewers and colleagues for their helpful com- Ontologies of Self-Tracking Practices. London: Emerald, ments, questions and suggestions on earlier versions and pre- pp.137–155. sentations of this article. Esmonde K (2020) ‘There’s only so much data you can handle in your life’: Accommodating and resisting self- surveillance in women’s running and fitness tracking prac- Declaration of conflicting interests tices. Qualitative Research in Sport, Exercise and Health 12 The author(s) declared no potential conflicts of interest with (1): 76–90. respect to the research, authorship, and/or publication of this Esmonde K and Jette S (2018) Assembling the ‘fitbit subject’: article. A Foucauldian-sociomaterialist examination of social class, gender and self-surveillance on fitbit community message Funding boards. Health: An Interdisciplinary Journal for the Social Study of Health, Illness and Medicine 1–16.Epub ahead of The author(s) disclosed receipt of the following financial sup- print, first published 19/9/2018, https://doi.org/10.1177/ port for the research, authorship, and/or publication of this article: This work was supported by the Leverhulme Trust Fiore-Gartland B and Neff G (2015) Communication, medi- under Grant RPG-2015-348. ation, and the expectations of data: Data valences across health and wellness communities. International Journal of Communication 9: 1466–1484. Notes Fox S and Duggan M (2013) Tracking for Health. 1. For a fuller account of these literatures, see Ruckenstein Washington, DC: Pew Research Centre. Gatineau M, Hancock C, Holman N, et al. (2014) Adult Obesity and Schu¨ ll (2017). and Type 2 Diabetes. Oxford: Public Health England. 2. Such as learning and routinising techniques, or making Gitelman L and Jackson V (2013) Introduction. In: Gitelman sense of and assessing the accuracy of the data. 3. Intermittent measurement aligns with Gorm and L (ed) ‘Raw Data’ Is an Oxymoron. Cambridge: MIT Shklovski’s idea of ‘episodic use’, although the timeframes Press, pp.1–14. differ in the studies. Where ‘episodic use’ denotes on and Gorm N and Shklovski I (2019) Episodic use: Practices of care off use across days in the week, intermittent measurement in self-tracking. New Media & Society 21, 11–12, 2505–2521. Hammersley M and Atkinson P (1995) Ethnography: in our study denotes periods of tracking and not tracking Principles in Practice. London: Routledge. across months or years. Harper D (2002) Talking about pictures: A case for photo 4. In further analysis we intend to consider whether we can see filtration work (Nielsen, 2015) in relation to the kinds elicitation. Visual Studies 17: 13–26. of conversations interviewees were hoping to pursue with Henwood F and Marent B (2019) Understanding digital their clinicians and the degree to which these data flow. health: Productive tensions at the intersection of sociology of health and science and technology studies. Sociology of We do not have space here to do justice to such analysis. Health & Illness 41: 1–15. Hooper L, Anderson A, Forster A, et al. (2016) Public ORCID iD Knowledge of the Link Between Obesity and Cancer. Kate Weiner https://orcid.org/0000-0002-5115-8866 London: Cancer Research UK. Kennedy H (2018) Living with data: Aligning data studies References and data activism through a focus on everyday experiences Ajana B (2018) Communal self-tracking: Data philanthropy, of datafication. Krisis: Journal for Contemporary solidarity and privacy. In: Ajana B (eds) Self-Tracking. Philosophy 1: 18–30. Kent R (2018) Social media and self-tracking: Representing Cham: Palgrave Macmillan, pp. 125–141. the ‘health self’. In: Ajana B (ed) Self-Tracking. London: Akrich M (1992) The de-scription of technical objects. In: Palgrave Macmillan, pp.61–76. Bijker W and Law J (eds) Shaping Technology/Building Lomborg S and Frandsen K (2016) Self-tracking as commu- Society: Studies in Sociotechnical Change. Cambridge: MIT Press, pp.205–224. nication. Information, Communication & Society 19: Ancker J, Witteman H, Hafeez B, et al. (2015) “You get 1015–1027. reminded you’re a sick person”: Personal data tracking Lomborg S, Thylstrup NB and Schwartz J (2018) The tem- poral flows of self-tracking: Checking in, moving on, stay- and patients with multiple chronic conditions. Journal of Medical Internet Research 17(8): e202. ing hooked. New Media & Society 20: 4590–4607. Andrejevic M (2005) The work of watching one another: Lupton D (2012) M-health and health promotion: The digital Lateral surveillance, risk and governance. Surveillance & cyborg and surveillance society. Social Theory & Health Society 2: 479–497. 10(3): 229–244. Davis JL (2017) Curation: A theoretical treatment. Lupton D (2013) Fat. London: Routledge. Information, Communication & Society 20: 770–783. Lupton D (2016) The Quantified Self. Cambridge: Polity. Didziokaite_ G, Saukko P and Greiffenhagen C (2018) Doing Lupton D (2017) How does health feel? Towards research on the affective atmospheres of digital health. Digital Health 3: calories: The practices of dieting using calorie counting app MyFitnessPal. In: Ajana B (ed) Metric Culture: 1–11, doi: 10.1177/2055207617701276. Weiner et al. 15 Lupton D (2018a) How do data come to matter? Living and digital living. Big Data & Society 4 1, 1–12, https://doi. becoming with personal data. Big Data & Society 5, 2, org/10.1177/2053951717700924. Piras E and Miele F (2017) Clinical self-tracking and moni- 1–11, https://doi.org/10.1177/2053951718786314. toring technologies: Negotiations in the ICT-mediated Lupton D (2018b) ‘I just want it to be done, done, done!’ Food tracking apps, affects, and agential capacities. patient–provider relationship. Health Sociology Review Multimodal Technologies and Interaction 2: 29. 26(1): 38–53. Lupton D (2019) ‘It’s made me a lot more aware’: A new Rich E and Miah A (2017) Mobile, wearable and ingestible materialist analysis of health self-tracking. Media health technologies: Towards a critical research agenda. International Australia 171(1): 66–79. Health Sociology Review 26: 84–97. Lupton D, Pink S, Heyes LaBond C, et al. (2018) Personal Ruckenstein M (2014) Visualized and interacted life: Personal data contexts, data sense and self-tracking cycling. analytics and engagements with data doubles. Societies 4: 68. International Journal of Communication 12: 19. Ruckenstein M and Schu¨ ll ND (2017) The datafication of Lupton D and Smith GJD (2018) ‘A much better person’: health. Annual Review of Anthropology 46: 261–278. The agential capacities of self-tracking practices. In: Schwennesen N (2019) Algorithmic assemblages of care: Imaginaries, epistemologies and repair work. Sociology Ajana B (ed) Metric Culture: Ontologies of Self-Tracking Practices. London: Emerald, pp.57–75. of Health & Illness 41: 176–192. Martens L, Halkier B and Pink S (2014) Researching habits: Sharon T and Zandbergen D (2017) From data fetishism to Advances in linguistic and embodied research practice. quantifying selves: Self-tracking practices and the other values of data. New Media & Society 19: 1695–1709. International Journal of Social Research Methodology 17: 1–9. Star SL and Strauss A (1999) Layers of silence, arenas of voice: Martens L and Scott S (2004) ESRC report – Domestic kitch- The ecology of visible and invisible work. Computer en practices: Routines, risks and reflexivity ESRC end of award report. Keele, Keele University. Supported Cooperative Work (CSCW) 8(1–2): 9–30. Mayer-Schonberger € V and Cukier K (2013) Big Data: A Urban M (2017) ‘This really takes it out of you!’ The senses Revolution That Will Transform How We Live, Work, and emotions in digital health practices of the elderly. Digital Health 3: 1–16. and Think. Boston, MA: Houghton Mifflin Harcourt. Van Dijck J and Poell T (2016) Understanding the promises Nafus D and Sherman J (2014) This one does not go up to 11: and premises ofonline healthplatforms. Big Data & Society The quantified self movement as an alternative big data practice. International Journal of Communication 8: 3, 1, 1–11, https://doi.org/10.1177/2053951716654173. Weiner K, Henwood F, Will C, et al. (2017) Self-monitoring 1784–1794. for health: Questions for an emerging field. Available at: National Institute of Health and Care Excellence (NICE) https://ssrn.com/abstract=3051201 or http://dx.doi.org/ (2011) Hypertension in Adults, Diagnosis and 10.2139/ssrn.3051201 (accessed 1 March 2020). Management. Clinical Guideline 127. London: NICE. Weiner K and Will C (2018) Thinking with care infrastruc- Neff G and Nafus D (2016) Self-Tracking. Cambridge: MIT Press. tures: People, devices and the home in home blood pressure Nielsen KD (2015) Involving patients with E-health – The monitoring. Sociology of Health & Illness 40(2): 270–282. Whitson J (2013) Gaming the quantified self. Surveillance & dialogic dynamics of information filtration work. Science Society 11: 163–176. & Technology Studies 28: 29–52. Will C, Weiner K, Henwood F, et al. (2019) Relating with O’Riordan K (2017) Unreal Objects: Digital Materialities, data: Stories of self-monitoring with care. In: 4S 2019, Technoscientific Projects and Political Realities. London: New Orleans, USA, 7 September 2019. Pluto Press. Williams R, Weiner K, Henwood F, et al. (2020) Constituting Oudshoorn N (2012) How places matter: Telecare technolo- practices, shaping markets: Remaking healthy living gies and the changing spatial dimensions of healthcare. through commercial promotion of blood pressure moni- Social Studies of Science 42: 121–142. tors and scales. Critical Public Health 30(1): 28–40. Oudshoorn N and Pinch T (eds) (2003) How Users Matter: Williams R, Will C, Weiner K, et al. (forthcoming) The Co-construction of Users and Technology. Cambridge: Navigating standards, encouraging interconnections: MIT Press. Infrastructuring digital health platforms. Information, Pantzar M and Ruckenstein M (2015) The heart of everyday Communication and Society. DOI 10.1080/ analytics: Emotional, material and practical extensions in self-tracking market. Consumption Markets & Culture 18: 1369118X.2019.1702709 (accessed 6 December 2019) Woodward S (2015) The hidden lives of domestic things: 92–109. Pink S and Fors V (2017) Self-tracking and mobile media: Accumulations in cupboards, lofts and shelves. In: Casey New digital materialities. Mobile Media & Communication E and Taylor Y (eds) Intimacies, Critical Consumption and Diverse Economies. Basingstoke: Palgrave Macmillan, 5(3): 219–238. Pink S, Ruckenstein M, Willim R, et al. (2018) Broken data: pp.216–232. Conceptualising data in an emerging world. Big Data & Woodward S (2016) Object interviews, material imaginings Society 5, 1, 1–13, https://doi.org/10.1177/2053951717753228. and ‘unsettling’ methods: Interdisciplinary approaches to Pink S, Sumartojo S, Lupton D, et al. (2017) Mundane data: understanding materials and material culture. Qualitative The routines, contingencies and accomplishments of Research 16(4): 359–374.

Journal

Big Data & Society – SAGE

Published: Apr 22, 2020

Keywords: Self-tracking; curation; users; data power; partial data; material methods

There are no references for this article.