Keywords

Introduction

How might we describe and make sense of an individual’s life? Is it best understood with reference to their accomplishments, family life, voluntary work and career—elements that might be narrated in a eulogy at their funeral? Or would this account miss the texture of their daily experience, the habits and routines that form the constant backdrop to these events? We now have substantial data resources from longitudinal studies that have tracked large samples of individuals over many decades. We also have myriad and increasing opportunities for tracking and recording our own daily lives and the lives of others. How might we extract and combine this information to understand, and potentially improve, individual lives?

This chapter has two parts. The first briefly explores the ways in which individuals have figured within longitudinal research in the social sciences and highlights an emerging set of methods focused on reconstructing individual cases within quantitative longitudinal research. The second is partly inspired by recent literature that emphasises the importance of attending to the mundane, the routine and the everyday (Highmore 2004, 2011; Pink 2012; Back 2015; Neal and Murji 2015). Specifically, I raise questions about the implications of the digital revolution (and in particular the self-tracking movement or ‘personal informatics’), for future research practices within longitudinal studies. It is now possible for detailed information to be collected in real time on individuals’ habits, behaviours and vital signs (Lupton 2016; Neff and Nafus 2016). This potentially provides researchers, and individuals themselves, with material that can be used to develop a different type of understanding of a life—one that focuses more on routine, lived experience and the practices and habits of daily life.

The thread that binds these two halves is the suggestion that it is instructive to think through what can be considered as ‘figure’ and ‘ground’ in our research, and in our representations of individuals’ lives. The metaphor is apt, partly because of the long sociological tradition of trying to understand the individual (or figure) in social context (the ground) without unduly privileging either the agency of the individual actor or societal structures and constraints (Mills 1959). It may also have renewed utility as we try to ‘figure out’ what it means to be an individual, and to lead a worthwhile and fulfilling life in today’s digital society. In particular, we could more easily conceive of figure and ground in temporal terms. Perhaps the ‘ground’ are the routines of daily life which, by definition, pass by almost unnoticed, while the figures are the events and experiences that loom large in our memories and our narratives about ourselves. Indeed, could a better understanding of the ‘ground’ of the habits and rhythms of our quotidian existence provide the key to understanding how to lead ‘better’, more fulfilling, lives?

In order to explore these questions I draw primarily on exemplars from Britain’s portfolio of national longitudinal studies of individuals’ lives. These include the long-running household panel study, known as “Understanding Society” (Buck and McFall 2011), together with the set of world-renowned cohort studies that have followed thousands of individuals from their birth (in the spring of 1946, 1958 and 1970 respectively) through childhood, adulthood and middle age (Power and Elliott 2006; Welshman 2012; Pearson 2016). As will be discussed below, these studies provide an instructive case study because they have been used by a wide range of researchers from different disciplines. This includes novel uses of the data to reconstruct or “refigure” individuals (Sharland et al. 2017, Tinkler et al. 2021; Carpentieri et al. 2022; Waller et al. 2020).

Section 1: Longitudinal Studies and Quantitative Representations of Individuals’ Lives

Some of the earliest longitudinal studies were carried out in the United States in the early twentieth century and focused on understanding children’s development (Phelps and Colby 2002). However, Britain is unique in the world in having a portfolio of four national birth cohort studies that have followed individuals, born in a specific year, through childhood, and into adolescence (Pearson 2016),Footnote 1 and adult life (Wadsworth et al. 2006; Power and Elliott 2006; Elliott and Shepherd 2006). A key feature of longitudinal research is that by maintaining contact with a large sample of individuals, and re-surveying them, typically every five to ten years, throughout their lives, it is possible to build up a rich and detailed record about the experiences of each member of the study. This is a type of quantitative life story, addressing many different aspects of each cohort member’s life. These include their education, childhood experiences, employment, housing, relationships, fertility, social participation and physical and mental health (Ferri et al. 2003; Wadsworth et al. 2006; Power and Elliott 2006; Elliott and Shepherd 2006).

The ability to follow the development of individuals throughout their lives has an inherently appealing, narrative quality (Elliott 2008). Indeed, parallels can be drawn between the 1958 cohort study and Michael Apted’s popular long-running documentary ‘Seven Up!’. This has followed a much smaller sample of just 14 individuals from when they were 7 years old in 1964 (Burawoy 2009; Thorne 2009).Footnote 2 The original premise for the series of documentaries was the Jesuit adage: ‘give me a child until he is seven years old and I will show you the man’. Apted deliberately chose children from contrasting social class backgrounds in order to see how material circumstances impact on individuals’ aspirations and life chances (Willis 2009). The British Birth cohort studies have also focused on inequality and on understanding the extent to which poverty and deprivation prevent individuals from realising their potential (Wedge and Prosser 1973; Wedge and Essen 1982).

However, in contrast to Michael Apted’s documentary approach, the majority of information collected in the British Longitudinal Studies is highly structured or quantitative. Therefore, its analysis typically involves the estimation of multivariate and longitudinal models. These focus on associations between different variables, and identifying which factors have the greatest impact on an outcome of interest later in life. The models produced, typically populated by columns of coefficients and standard errors, can seem a far cry from stories, or narratives, about real human beings (Elliott 2005).

There is therefore a sense in which the cohort members themselves are obscured in the quantitative analyses that characterise the majority of work carried out using data from these studies (Elliott 2005, 2008). As Armstrong (2019) has argued, “Ironically just as these data points could claim to reveal a new numerical description of the individual, their combination and comparison … involved choreography of data points quite separate from the individual” (p. 110). In other words, we risk losing touch with the uniqueness and complexity of individual lives as these are represented as sets of summary variables that can be manipulated by the epidemiologists, economists, psychologists and sociologists who use the datasets. And we lose touch too with the ability of individuals to reflect on their own lives and, perhaps, to compare them with those of others. It is rare for cohort members to be given a voice and enabled to reflect on their own experiences.

One exception to this is a qualitative study conducted with a sub-sample of 220 members of the 1958 cohort between 2009 and 2010 (Elliott et al. 2010). Individuals were asked about their communities, social participation and weekly routines as well as being given an opportunity to tell their own life story. At the end of the interviews, cohort members were also asked about their experiences of being in the study throughout their lives (Parsons 2010). Many had positive memories of how it had made them feel ‘special’ in early life to be part of a study that would be useful to wider society. However, there was also a desire among some study members to receive more feedback about the study in the form of case studies and stories about other cohort members. As one cohort member said:

I think most of the feedback that comes back is very, very generic which--, I tend to get bored halfway through reading so I don’t bother…maybe some examples, some, I don’t know, common case studies, stories, that sort of stuff would make it more interesting and I’d read it then. [Interview 239] (From Parsons 2010, p. 15)

It is perhaps too strong to claim that individual cohort members actually disappear in the multivariate statistical analysis of their data. Rather they provide an essential background, contributing to the mass of data points from which statistical models are estimated. Whether we are researchers or readers of research findings, we know that the individuals are there. It is the representative nature and large size of the sample that ensures the statistical models are credible representations of underlying processes in society (Hawkes and Plewis 2006; Mostafa et al. 2020). Even so, the intense focus on variables in multivariate analyses means that the agency and reflexivity of individuals are likely to be obscured (Abbott 1992). In the quantitative, multivariate, longitudinal models that capitalise on the detailed prospective information in the cohort studies, it is the coefficients that populate the models which figure, while the cohort members themselves provide the ‘ground’.

Set against this, the relative invisibility of individual study members has the advantage of protecting the anonymity of those who have contributed a great deal of very personal, and sometimes sensitive, data throughout their lives. In contrast, in Apted’s ‘Seven Up!’ series the individuals are the key figures in the documentaries. Indeed these individuals have taken on an almost celebrity status.Footnote 3 However, this level of visibility has led some participants to opt-out. Five of the fourteen participants have declined to participate in at least some of the updates over the years. For example, Charles, recruited for the documentary from an elite public school, dropped out after 21 Up and has never returned; whereas Peter dropped out of the series after 28 Up, following a campaign against him in the tabloid press due to his criticism of the Conservative government during his TV interview. He returned to the series for 56 Up in order to publicise his band.

Reconstructing the Individual Within Longitudinal Cohort Studies

Despite the tendency of longitudinal studies to obscure the individuals who take part in them, there are a few examples of research which do take a more individual case-based approach. These studies recognise that the detailed and temporal nature of the studies, and the location of cohort members in a specific historical context, mean that the studies have considerable narrative potential (Elliott 2005; Elliott et al. 2010; Waller et al. 2020).

Indeed, a number of researchers have adopted imaginative methods which in some senses reconstruct the individuals who have been fragmented into a set of variables so that those who were in the background come to figure. For example, Singer et al. (1998) use the Wisconsin longitudinal study to understand more about the factors that can lead to depression for some women. They use different waves of the Wisconsin study (many years apart) to piece together individual life stories for a small sub-sample of individuals. Singer et al. argue that “new insights are obtained as detailed information about real people are brought into focus” (Singer et al. 1998). These insights can then be used to generate hypotheses, which can in turn be tested using statistical models.

A recent paper, drawing on this approach, has used data from the long-running British Household Panel Study to construct case studies of families who have been supported by social workers (Sharland et al. 2017). A key aim was to explore whether this more narrative methodology, focused on the lives of individual families, would provide insights into a counter-intuitive finding emerging from statistical analysis. Namely that families who have contact with social services have poorer outcomes than families in similar circumstances without support. In the authors’ words “In the absence of complementary qualitative material, quantitative life histories seemed worth trying, to catch a glimpse of the stories beneath the aggregates” (Sharland et al. 2017 p. 670). Sharland et al. are understandably tentative in their advocacy for this method, based on its limited use to date. However, they conclude by arguing that given the impressive array of quantitative longitudinal studies in the UK, the USA and Europe, researchers might make better use of the “largely untapped narrative potential that may enrich our understanding of how lives unfold. The quantitative life history narrative method offers a chance to realise this potential” (Sharland et al. 2017).

Very recently, in the UK, two separate historical studies have adopted similar techniques and risen to this challenge. Peter Mandler’s study on the history of secondary education since 1945 includes the creation of 150 pen portraits of cohort members from the 1946, and the 1958 cohort studies in order to understand more about the family backgrounds, educational and occupational trajectories of two separate generations (Carpentieri et al., 2022). The Girlhood and Later Life project led by Penny Tinkler focuses on girls growing up in Britain in the 1950s to 1970s. The team uses materials from the 1946 British Birth Cohort study (known as the National Survey of Health and Development (NSHD)) to reconstruct biographies of women from different education and class backgrounds, in order to understand more about their opportunities and life courses. As they write: “we can do more than generate statistics from birth cohort studies such as the NSHD; we can also recompose persons. The crux is how we understand data and persons. Recomposition entails scavenging for various (including unrecognised) data, and combining them to generate biographical collages” (Tinkler et al. 2021).

These studies provide examples of ways in which individuals, who are usually expected to fade into the background within large-scale studies can be re-configured or refigured by researchers who have an interest in documenting the experiences of individual cases. Indeed, what makes the cohort studies a compelling resource for this kind of work is that the large sample size makes it possible to select very specific cases for analysis and to understand those individual lives in context of the much broader sample.

It is noteworthy that historians are prominent in the cadre of researchers who have started to use the cohort studies in this new way. Case studies of individual cohort members provide insights into the past, and their prospective nature means that, in contrast to the use of oral histories, there are fragments of detailed information collected contemporaneously. In these uses, while it is individuals who figure they are primarily of interest for the insights they provide into the broader historical picture, the experiences of going to a Grammar School or a Secondary Modern School in Post-war Britain, or the different opportunities perceived as available for boys and girls. As Tinkler et al. reflect, “Recomposition is … interested in the singularity of individuals, it attends too to the historical and relational embeddedness of personhood” (Tinkler et al. 2021). The particular appeal of these case studies is perhaps that we can gain some sense of the ‘big stories’ of individual lives. We can look for continuity and change in circumstances over many decades, and we can gain insights into the childhoods and young lives experienced half a century ago.

Big Stories and Small Stories

The focus in both conventional multivariate analysis and the relatively recent work on the re-composition of individuals within longitudinal studies leaves us with another question or conundrum—namely what are the best ways of documenting and understanding individuals’ more quotidian experiences? This question highlights an interest in understanding figure and ground in a more temporal sense. When we recount our own life stories or compile a CV we focus on key events, experiences or transitions—the dates of birth of children, when we changed job, or moved house. Indeed these are also the key pieces of information documented in many longitudinal studies about peoples’ lives. It is these events that ‘figure’ in our lives against a backdrop or ‘ground’ of quotidian routine. “Almost by definition, the quotidian can be overlooked, not actually noticed for much other than for its sameness and its continuities” (Neal and Murji 2015, p. 812). In the second half of this chapter I want to focus on how, and why, we might rehabilitate these daily experiences and place them centre stage, to make them figure. Habit and routine are central features of our everyday lives, and yet the every day has been largely ignored by the cohort studies.Footnote 4 The metaphor of figure and ground can therefore be applied not just to the contrast between the individual case study and the large sample, but also to our temporal focus. In the analyses of the cohort studies it is life events and key transitions that figure against the taken for granted ground of everyday experience.

There are a few examples of the cohort studies trying to collect some of this mundane and everyday information in the past. For example, journalist David Ward reports that the 1946 Birth Cohort Study recorded that he had ‘meat (unspecified), peas and potatoes (and blancmange for pud) for dinner on 15 June 1950’ (when he was aged 4) (Ward and Payne-Humphries 2013). Indeed, there have been a few isolated and relatively unsuccessful attempts in the cohort studies to collect and analyse a few days of dietary diaries and activity diaries (Crawley and While 1996). However, the burden that this puts on respondents, and the difficulty of collecting data in a consistent manner, has led the studies to focus on recording more major life events such as house moves, job changes and births, marriages and deaths. Where there is interest in more regular activities such as exercise, and other forms of leisure or social participation, the cohort studies have typically relied on standard self-report retrospective survey techniques (Sacker and Cable 2006).

There are some parallels here with the distinction made between big, medium and small stories in work on different levels of narrative in the social sciences (Phoenix and Sparkes 2009; Griffin and Phoenix 2016; Back 2015). The big and medium stories are the accounts that individuals give about aspects of the long durée of their lives, often in response to interview elicitation, whereas the small stories are only likely to occur in conversation and correspond to reflections on the everyday and routine aspects of life.

Section 2: Opportunities and Challenges for Longitudinal Research Provided by Self-tracking

The emergence of new technologies for monitoring and recording daily life at an individual level provides both opportunities and threats to well-established longitudinal studies. Wearable devices such as Fitbits, and an increase in techniques and tools for ‘self-tracking’ or ‘personal informatics’, now make it more possible to understand, or at least to record, life as it is lived at the quotidian level. Digital self-tracking “has become a mass phenomenon through omnipresent smart phones” (Heyen 2020 p. 124). Self-tracking technologies are marketed as providing insights for the individual user, but they could also be adopted for use in large-scale studies. Digital wearables and associated apps could provide new methods for collecting and recording data that would correspond to some of the small stories of daily life. As will be discussed below, these methods would need to be acceptable to participants to avoid jeopardising continued involvement in longitudinal research. Before exploring the potential use of new technologies for collecting data in the major longitudinal studies, it is worth briefly discussing the growing literature on self-tracking and the ‘quantified self’.

Self-tracking and the ‘Quantified Self’

The proportion of those using a smartphone in the UK has risen very rapidly from around 17% to 87% between 2008 and 2020. And it is those in the youngest age groups who are most likely to use a smartphone (99% of those aged 16–24) (Statista 2021). Using data collected in 2016 it was estimated that around a third of internet-connected people worldwide track their health and fitness via an online or mobile app or a wearable device (Herder 2016). In 2017 there were reported to be as many as 325,000 health apps (Research2Guidance 2017). Now that technology to facilitate constant monitoring of all sorts of different types of behaviour is so available to individuals—what is the potential for longitudinal research to incorporate this type of information?

The pace of change makes it difficult to know with any accuracy how many people are engaged with some form of purposeful or ‘active’ self-tracking. There will of course also be a spectrum of engagement. While some individuals may occasionally use a form of self-tracking (e.g. a steps counter), others are much more deeply engaged in projects to observe, analyse and change daily habits and behaviour. One manifestation of this is the ‘quantified-self movement’, started in California in 2008 by Gary Wolf, which now includes conferences and meet-ups around the world. The quantified-self website has the strap line ‘self-knowledge through numbers’, and provides numerous resources designed to help individuals understand themselves better, and make changes to their habits and routines.

A clear theme of the quantified-self movement is that by observing, recording and then analysing their data over time, an individual can gain greater insights, greater control over their life and the ability to improve outcomes. As Heyen (2020) has discussed, using examples from his ethnographic work on self-tracking, “self-related insights are taken into account by the self-tracker in his daily routines … and they contribute, according to his own perception, to his improved well-being” (Heyen 2020, p. 129). Arguably, the individuals who engage in self-tracking are also seeking to distinguish between figure and ground. The process of collecting and recording data using wearables and apps helps to discern the aspects of daily life which are most salient for influencing an outcome of interest. Frequently, the emphasis is on being able to visualise patterns in the data so that the important figures emerge from the background “noise” of irrelevant measurements (Ruckenstein 2014; Kristensen and Ruckenstein 2018). This individual approach to gaining insights typically does not make use of the same principles of statistical inference used in large-scale longitudinal studies. Here the sample size is a single individual (i.e. an n of one), and the logic is that by collecting multiple data points over time and varying different factors (usually individual behaviour) clear patterns will emerge from the data. However, while both self-tracking practices and longitudinal studies both rely on time, this is framed in very different ways. While self-tracking practices rely on a cyclical and repetitive conception of time in order to observe, record and modify behaviour on a daily basis, longitudinal studies in the social sciences rely much more on a linear conception of time. Time, therefore, figures in rather different ways in these two approaches.

Within the growing body of literature on the practicalities, advantages and experiences of self-tracking, questions have been raised about the type of self that is promoted and constituted by these practices. For some, there is potential for these digital practices to constitute a new kind of surveillance, building in normative expectations about appropriate behaviours, sleep patterns, body size, etc. (Lupton 2012; Ruckenstein 2014). There is also concern that ‘self-knowledge through numbers’ as supported by the QS community promotes the model of the ideal neoliberal citizen, that is, the self-monitoring and self-optimising individual who voluntarily aims to control and discipline their everyday behaviour (Lupton 2012; Depper and Howe 2017; Sanders 2017).

In a more optimistic vein, Kristensen and Ruckenstein (2018) use longitudinal engagement with a group of Danish self-trackers to explore the concept of the ‘laboratory of the self’. They suggest that “Self-trackers use technologies to take the self apart, to highlight certain ‘authentic’ aspects of it or to intensify human agencies or senses. They try out applications and devices: starting off somewhere, learning about themselves and coming out of the experience in another place” (p. 3635). This leads to the argument that self-trackers are not necessarily dupes skilfully cajoled into digital consumption and constant utilitarian self-improvement. Rather Kristensen and Ruckenstein provide evidence of reflexive individuals whose engagement with personal informatics makes them more attuned to the emergent properties of the self and enables them to be more conscious of their “agentic aims and powers” (p. 3631).

This explicit examination of the nature of the self that is promoted via self-tracking is echoed in the works of Rapp and Tirassa (2017). Their focus is on how we might try to improve the technologies that enable personal informatics in order to go beyond the rather ‘utilitarian self’ of the quantified-self movement. Contrary to Kristensen and Ruckenstein, Rapp and Tirassa argue that the self currently implicit here is the self of behaviourist psychology: a self that is ultimately unknowable and therefore under-theorised. This can result in a self that appears to consist only of a set of observable behaviours reduced to data points (Armstrong 2019). This perspective on self-tracking suggests that what actually changes is not the self but the behaviour or indeed the visible (or measure-able) body.

By invoking the phenomenological, subjective self as a far more interesting object for study, Rapp and Tirassa prompt an exploration of how personal informatics could be developed to allow individuals to engage much more fully with this subjective, experiencing self. Using the framework of four aspects of the phenomenological self (the past, present, future and interrelated self), they proceed to sketch a research agenda and set of guidelines. Key to their argument is that technologies should be developed in a way that transcends the focus on behaviour change and allows for a more thoroughgoing reflection on the self, one that foregrounds the importance of both context and environment. Their work, therefore, resists the pressure for us to become neoliberal subjects who “are constantly encouraged to change their habits – rather than society and institutions – in order to become happier more productive people” (Chun 2016). It is also noteworthy that their four aspects of the phenomenological self move the focus from the cyclical time of habit and routine and place the individual more clearly in linear time.

Rapp and Tirassa (2017) and Kristensen and Ruckenstein (2018) clearly start from divergent perspectives on the conceptualisations of the self promoted by personal informatics practices. However, both sets of scholars provide us with the promise of re-figuring the self from being constituted only by behaviour to being fully subjective, reflexive and focused on interaction with the environment. This suggests that personal informatics could have a “grander ambition” and develop “the capability of revealing something of the individual's self" (Rapp and Tirassa 2017 p. 340).

The Potential Use of Self-tracking in Longitudinal Studies (Figure and Ground)

For any long-term longitudinal study, there is likely to be a tension between making maximum use of innovative data collection techniques and maintaining consistency to ensure that the longitudinal design of the studies can be exploited to the full using appropriate statistical analyses.

Arguably, if there were enough resources, then it would be possible both to preserve and to add in new data collection strategies that make use of emerging digital technology. However, those running the studies also need to be cognisant of the burden that data collection places on cohort members. The longitudinal studies have been able to maintain extremely high response rates due to the loyalty of cohort members, who have participated since childhood (Mostafa et al. 2020). There is understandably a fear that introducing new forms of data collection may alienate long-term respondents and compromise the quality of the studies for future researchers.

Part of the problem here is the relative lack of research to date on the ways that individuals routinely use digital devices in their daily lives. There are the beginnings of a body of research on individuals’ self-tracking (Nafus 2014; Ruckenstein 2014; Ajana 2020; Heyen 2020; Lupton 2020). However, the more active and engaged individuals who constitute the Quantified-Self movement are still only a tiny percentage of the population.Footnote 5 This means that it is difficult to assess the potential for using digital recording and tracking methods in a representative sample of British cohort members.

Recently, to address this issue, the directors of the cohort studies commissioned qualitative research to assess the acceptability to cohort members of using innovative methods to collect new types of data (Ipsos 2019). During 2019, interviews were conducted with samples of 28 individuals from each of 4 cohort studies (i.e. a total of 112 interviews), complemented by a focus group discussion from each cohort. Key questions included how cohort members would feel about providing access to their social media activity, their travel (as automatically recorded via travel cards) and their financial transactions (using a specially designed app). Interviews and focus groups also covered the more general use of new technologies such as apps to actively or passively collect detailed data including screen time, GPS and activity tracking.

Despite the strong loyalty of cohort members to the longitudinal studies, it was striking that across the interviews and focus groups, study members consistently reported that novel data collection felt like a form of surveillance and therefore regarded it with unease. Some of the comments included:

The more that the study moves towards big brother tracking, I would struggle with it and may withdraw from the study. (BCS70, telephone interview, 83, did not take part at age 46) (p. 87 Ipsos Mori, 2019)

I wouldn’t like to do any of it it’s too personal, too private that feels like big brother is watching me. (BCS70, telephone interview, 103, took part at age 46) (p. 85, Ipsos Mori 2019)

No, I wouldn’t agree to any of that… I know they always say: ‘big brother knows where you are’ and I’m sure somebody does but I don’t want to have all these apps and things to make it even more. I’m not interested in any of that, no. (NCDS, telephone interview, 10, took part at age 55) (p. 85, Ipsos Mori 2019)

There was a more positive reaction to the idea of collecting exercise data using a Fitbit, or similar wearable device, as this was seen by cohort members to be directly linked to health research and therefore an acceptable part of the study.

I feel that is a difference as it can show the study how many steps I have taken and how many calories I have burned then yes as it was just health focus which is important rather than how long I have spent checking the weather on my phone. (BCS70, telephone interview, 103, took part at age 46) (p. 86, Ipsos Mori 2019)

It is interesting that in the first three of these quotations the cohort members each invoke the fictional ‘Big Brother’, originally conceptualised within Orwell’s dystopian novel 1984 (and then popularised by the reality TV series). This is a ready shorthand for surveillance that covers the most private and seemingly inconsequential activities of life. Here then we see a contrast between individuals being uncomfortable with tracking of everyday habits and experiences that seem to have no readily understandable benefit for research, while there is an acceptance that monitoring the body—calories input and expended—can have a value for understanding and improving health.

There are aspects of the major longitudinal studies that now capitalise on the use of the web and personal computers to simplify data collection (e.g. the age 62 sweep of the 1958 cohort is collecting a dietary diary using the web). However, no extensive use is being made as yet of wearable devices or the ability of smartphones to prompt the user to report on activities over the course of a day. This means that consistency is maintained. However, what remains missing from the detailed quantitative linear chronicles of longitudinal studies is a feel for the daily lives and everyday practices of cohort members—how much time they spend commuting, working, watching television, out with friends or asleep; how many steps they take; how their heart rate varies over the course of a day, whether they eat three meals or multiple snacks; etc. This description of what is missing is not to diminish the value of the rich data of the cohort studies but rather to serve as a reminder that they provide only a partial picture of individuals’ lives. They foreground linear time, and it is this which figures against assumed, but invisible, daily experiences. As Back has argued, “the everyday matters because it offers the ability to link the smallest story to the largest social transformation” (Back 2015 p. 834)

In an article for the NY Times magazine, Gary Wolf, a co-founder of the Quantified-Self movement, wrote that:

We track ourselves all the time, but something changes when we digitize this self-monitoring … when the familiar pen-and-paper methods of self-analysis are enhanced by sensors that monitor our behaviour automatically, the process of self-tracking becomes both more alluring and more meaningful. Automated sensors do more than give us facts; they also remind us that our ordinary behaviour contains obscure quantitative signals that can be used to inform our behaviour, once we learn to read them. (Wolf 2010)

This desire to adopt methods which allow patterns to surface from the background noise of data and to figure out what is meaningful, once again returns us to the metaphor of figure and ground. We want to believe that there is more to life than random noise and that meaningful patterns will emerge if we only have the tools and patience to be able to observe what is really there.

There are also some interesting parallels here between the promise or ‘allure’ for individuals that once we fully understand ourselves we will be able to improve our lives and our well-being, and the promise of the cohort studies whose overriding aim has always been to provide policy insights that will improve the lives in the aggregate, especially for disadvantaged groups within society. As Ferri et al. wrote in conclusion to their 2003 book on the cohort studies:

To gain a proper understanding of what (policies are) likely to be most effective, when and with whom, we need much more research on the mechanisms and processes of success and failure in an increasingly complex changing world. Investigation of the interactions of the effects of social change with the development of individual lives will continue to drive research using the cohort study data in the years to come. (Ferri et al. 2003: p. 312)

Whereas the power of the cohort studies lies in the large sample size as well as the length of observation, for individuals using digital methods to track and record their behaviour the sample size is an n of one. Both approaches hope to be able to discern meaningful ‘patterns’ from among the background noise of a superfluity of data points. Both, therefore, use methods of analysis that will enable the figure to be distinguished from the ground. What is also shared here is the possibility of collecting data over time and observing how change in one domain impacts on outcomes in another. However, as discussed above, there is a sense in which the conception of time is subtly different in the two approaches. In large-scale longitudinal studies, the emphasis is on linear time with cyclical time assumed, but relatively obscured in the background. In contrast, the process of self-tracking has tended to interrogate habits embedded in cyclical time, “practices acquired through time that are seemingly forgotten as they move from the voluntary to the involuntary, the conscious to the automatic” (Chun 2016, p. 6).

Conclusions

This chapter has explored the ways in which individuals can be made to appear, or disappear, in longitudinal research, whether that is in large-scale cohort studies or in recent work on personal informatics. Invoking the metaphor of figure and ground raises the question of what counts as the ‘ground’ that is, what is the backdrop or context against which the subjects of research (i.e. the figures) can be made to appear or disappear, and to what extent does that context actually serve to constitute the figure itself. When focusing on large-scale, quantitative and structured cohort studies, the backdrop or ground can be understood to be both the large representative sample that frames and makes sense of each individual’s set of unique data points, and the historical and geographic context. Indeed this methodological approach to understanding individual lives is already well-rehearsed within the literature on the Life Course (Giele and Elder 1998). And this literature draws attention to the way in which historical events, such as the Great Depression, not only provide a backdrop to a life but actively constitute the experience and subjectivities of each individual.

What is key in large-scale quantitative research is that, paradoxically, in order to focus on understanding the factors that may influence individual outcomes the individual research subjects are effectively removed from sight. Although each individual contributes myriad data points, their data is deliberately anonymised. It is the researcher and not the research subject who crafts causal narratives. Using multivariate, and sometimes multilevel, statistical modelling techniques, variables and coefficients appear to have agency, that is, these are the figures of interest here. Even innovative case-study approaches that have sought to refocus attention on individuals rarely seek ultimately to foreground the individual but rather to develop deeper understandings of causal process or historical context and change.

The second half of this chapter shifted attention to the implications for longitudinal research of the increase in self-tracking practices and personal informatics. While these activities, with an n of 1, appear to put the individual centre stage it is still instructive to consider what constitutes figure and ground in this novel approach to ‘personal science’ (Heyen 2020). For an individual self-tracker looking for patterns in their data over time, the ground is perhaps those aspects of individual experience and behaviour found not to be relevant for achieving the outcome of interest; whether this is improved fitness, attention, sleep patterns, or wellbeing. The practice of self-tracking is motivated by a belief that with the right tools and techniques it will be possible to discern the meaningful patterns in the data, to figure out what matters and to adjust behaviour accordingly.

What the burgeoning literature on personal informatics often neglects however is a deeper or more explicit theory of what constitutes the self (Rapp and Tirassa 2017; Kristensen and Ruckenstein 2018). Arguably if the data points, collected by and on an individual, are no more than representations of behaviour, then the self becomes no more than the coordinator of that behaviour. Such a self would arguably be completely uninteresting and one dimensional if it were not for two narrative elements, the ability to infer causal links from the quotidian data observed, recorded and visualised in cyclical time, but also the possibility for change over linear time. In this context, narrative serves to vivify data points and constitute a self that is traceable over time and can change over time in a way that can be meaningfully understood. In seeking to figure out the individual in longitudinal research, we therefore need to attend to more than the contrast between (or mutual constitution of) figure and ground, but their mutual constitution in cyclical and linear time. Perhaps the greatest challenge for the future is how to make best use of new technologies for data collection while also considering how to place a thoroughgoing subjective, or phenomenological, self at the centre of our research narratives.