Narratives of the Night: The Use of Audio Diaries in Researching Sleep

by Jenny Hislop, Sara Arber, Rob Meadows and Sue Venn
Centre for Research on Ageing and Gender (CRAG), Department of Sociology, University of Surrey

Sociological Research Online, Volume 10, Issue 4,
<http://www.socresonline.org.uk/10/4/hislop.html>.

Received: 7 Oct 2005     Accepted: 30 Nov 2005    Published: 31 Dec 2005

Abstract

This article draws on data from two major empirical studies of sleep to examine the use of audio diaries as an approach to researching sleep. Sleep has only recently emerged as a topic of interest to the sociologist, providing a valuable resource through which to examine the roles and relationships and gender inequalities which underpin everyday life. Yet accessing individual experiences of sleep is problematic. Considered a non-conscious activity, sleep takes place in most cases at night within the private domain of the home and is thus generally inaccessible to the social researcher and outside the conscious reality of the sleeper. In exploring the social aspects of sleep, we rely primarily on respondents' interpretations of the sleep period given retrospectively in focus groups and in-depth interviews, distanced from the temporal, spatial and relational dimensions of the sleep event. This article also focuses on the use of audio diaries as a method designed to help bridge the gap between events in real time and retrospective accounts. We examine the narrative structure of audio diaries, discuss the principles and practice of using audio diaries in sleep research, illustrate the contribution of audio diary narratives to an understanding of the social context of sleep, and assess the use of audio diaries in social research. We conclude that, used in conjunction with other methods, audio diaries are an effective method of data collection, particularly for understanding experiences of intimate aspects of everyday life.

Keywords: Qualitative Methods, Audio Diaries, Narratives, Sleep, Women, Couples, Social Context

Introduction

1.1 As the 'chief nourisher of life's feast', sleep plays a vital role in health, well-being and quality of life. Yet despite its prominent place in literature and in bio-medical research, sleep has in general been considered of little interest or relevance to sociologists. As a result, while sociologists such as Aubert and White (1959a, 1959b), Schwartz (1970), and Taylor (1993), proposed a social interpretation of sleep, their work has had relatively little impact on the development of sleep as a recognisable theme in sociological research. Recent studies by Williams (for example: 2002, 2004, 2005, Williams and Bendelow, 1998,Williams and Boden, 2004); Hislop and Arber (2003a, 2003b, 2003c, 2004); and Meadows (2005), however, have built on these early works, contributing to the emergence of a sociology of sleep, in which the study of sleep has become both an important topic in its own right, and a resource through which to gain insights into gendered relationships and social inequalities.

1.2 Rather than focusing exclusively on the physiology of sleep, sociologists contend that sleep takes place within a social context which influences the patterning of sleep throughout the life course. Moreover, as a mirror of everyday life, sleep reflects the roles and responsibilities, gender divisions, health patterns and transitions which characterise people's lives. Yet in studying sleep, the social researcher is faced with a dilemma. As an intrinsic part of everyday life and an innate part of people's experiences, sleep, like breathing, remains essentially a non-conscious activity, inaccessible to both the participant and the researcher. Bio-medical scientists can measure sleep objectively in laboratories and sleep clinics through the use of polysomnography[1] and actigraphy[2] while people are asleep, whereas sociologists must base their understandings of sleep on interpretations of conscious, subjective accounts given retrospectively through a framework of everyday experience. Indeed, it could be argued that, from a sociological viewpoint, recording people's experiences of non-conscious activities such as sleep, is a contradiction in terms. We acknowledge that so-called 'experiences' of sleep may in reality be experiences of wakefulness before, during, and after the sleep period; conscious reflections on sources of disruption and their impact on the quality of our sleep and well-being. Despite its seeming contradictions, however, researching sleep sheds light on an otherwise hidden dimension of our lives. Methods which capture subjective experiences of the sleep period may be equally applicable to researching other areas of everyday life and intimate behaviour.

1.3 This article examines the use of audio diaries as a qualitative method for collecting personal interpretations of the sleep period over consecutive nights. Following a brief overview of the use of paper diaries in social research, the article discusses the use of audio diaries as part of a multi-method approach adopted in the first empirical study of sleep in women aged 40 and over in the UK[3]. It then draws on data from a study of sleep in couples aged 20-59[4] to examine the narrative structure of audio diaries, and to illustrate how data from audio diaries acts as a window onto couples' lives. Data from both projects are used to assess the contribution of this method to empirical sociological research. The article suggests that the use of audio diaries could be extended to other research areas, otherwise inaccessible to the researcher.

The use of paper diaries in social research

2.1 Self-completion paper diaries, or logs, in which participants are asked to record simple quantitative data about their activities, are frequently used in scientific studies and in clinical and biomedical settings, and, less often, in social research. Primarily associated with the production of quantitative data, paper diaries are used where researchers are 'specifically interested in precise estimates of different kinds of behaviour' (Bryman, 2001: 136), for example, the UK 2000 Time Use Survey (National Statistics, 2003). Used on their own, however, these diaries give few insights into the embeddedness of behaviour within its social context.

2.2 Bryman outlines three main ways in which paper diaries have been used in social research: (a) as a researcher-driven method of data collection; (b) as an historical, autobiographical document written spontaneously by the diarist (used mainly by social historians); and (c) as a log of researcher activities similar to field notes written by ethnographers. It is on how diaries have been used in data collection that we are interested here. While a detailed discussion of the use of paper diaries in social research is given elsewhere (see for example, Plummer (1983), Elliott (1997), Bryman (2001), Bytheway and Johnson (2002), Roberts (2002)) and is outside the scope of this article, the following examples illustrate the potential of paper diaries, when used in conjunction with interviews, to inform social research.

2.3 As Johnson and Bytheway (2001: 183) note, research that focuses on 'everyday life' has to be oriented to questions of 'what happens in reality'. It is on capturing this reality, that the paper diary/interview method has been used successfully in social research. In their study of medication in later life, Bytheway and Johnson (2002) designed a two-week diary in the form of a double page checklist grid on which participants could record their use of medication across a 24-hour time scale in relation to six specified aspects of daily living. Space was provided for brief notes and comments, designed to facilitate recall of events during the follow-up interview. In association with in-depth interviews, these diaries offered insights into the management of medication in the broader context of everyday life. In evaluating the use of the diary-interview method as a means of investigating aspects of everyday life, Bytheway and Johnson (2002: 171) conclude that the use of the diary/interview method can 'help to distinguish what people actually do from what they say they do; and can offer insights into many aspects of daily living in later life'.

2.4 In her study of lay decision-making about health and illness, Elliott (1997: 4) used a diary-interview approach to observe the health behaviour of eight diarists, selected from a sub sample of 80 informants from a study designed to 'ascertain prevalence of musculo-skeletal symptoms and patterns of help seeking for these'. Although guidelines were given, diaries were otherwise unstructured, with respondents asked to record how they felt, and any health problems and action they had taken each day over a three week period. Elliott (1997: 9) found that the use of diaries 'served to bring to the foreground issues about illness which are commonly at the back of people's minds'. Diaries were kept as an intrinsic part of a wider research process which included interviews with participants, before, during and after the diary-keeping period. By reflecting on their illness in their diaries and using this as a basis for subsequent interviews, participants became active collaborators in the research process. As Elliott (1997: 5) notes, however, not all respondents warmed to the task of written diary keeping, with the quality of diary entries reflecting individual predisposition to the role of 'writer'.

2.5 Like Elliott, Coxon (1994) used a free text approach in his research on the response of gay men to the risks of HIV/AIDS. By providing guidelines for diarists on the kinds of information to be recorded, including the context in which sexual acts took place, details of 'who did what and with what and to whom' (Coxon, 1988: 356), Coxon was able to access data on gay men's sexual behaviour which may otherwise have been withheld, providing insights into aspects of the social world and lived experience of being gay in the era of HIV/AIDS. Coxon (1994: 129) argues that 'the diary makes it possible to obtain information in far greater detail than other methods, since it is designed to minimize recall and memory errors and cognitive strain (and) specifically adapted to gathering reliable information on the time-sequence of events'. Unlike the semantic and thematic analysis used with conventional diaries, however, Coxon's sexual diaries were subject to an elaborate system of coding and analysis using specially designed software, making the method 'undoubtedly demanding' on both respondents and analysts (Coxon, 1994: 144).

2.6 In contrast, the audio diary approach discussed below enables the researcher access to the intimate daily lives of individuals, while adopting a more familiar and conventional qualitative analytic approach. Audio diaries are not intended to replace paper diaries or for use as a stand alone research method. Rather, we argue, that, used within qualitative research in conjunction with other research methods, they represent an approach to research which can overcome problems of completion and motivation associated with paper diaries, and offer the social researcher a means of collecting and analysing qualitative data on topics which may otherwise remain inaccessible.

Using audio diaries in sleep research

3.1 Audio diaries were initially trialled in our EU-funded Sleep in Ageing Women project. This project aimed to examine women's sleep from three main perspectives: firstly, to increase understanding of the changing patterns of women's sleep across the life course; secondly, as a means of highlighting the interaction between institutional, relational and life course factors in the structuring of women's sleep; and thirdly, as a resource through which to gain insights into the ways in which studying sleep in its social context can highlight the underlying gendered roles and relationships and inherent constraints which characterise women's everyday lives. To accomplish this, the study adopted a multi-method approach, incorporating the use of focus groups (n=15), in-depth interviews (N=35), and a national postal survey (N=1445), as well as using the audio diary as a means of recording individual assessments of the sleep period.

3.2 Sleep takes place in a socially contextualised environment regardless of whether sleep occurs at night, or, by virtue of shift work or naps, during the day. While data from focus groups, face-to-face interviews and the survey provided understandings of sleep in its social context, women's accounts were distanced by time from the temporal, spatial and relational realities of sleep and as such, potentially subject to recall bias. While acknowledging the inescapable influence of retrospectivity in accounts of sleep, as in many areas of social research, the challenge for the researcher was to find a method which provided insights into sleep experiences on a nightly basis as close to the event as possible and independent of researcher involvement, thus ensuring relatively untainted individual interpretations of the experience of sleep. As Elliott (1997: 11) notes, the value of diaries is their 'potential to record events, over time, as close as possible to when they occur'. Thus rather than using written sleep diaries, it was felt that the use of an audio diary would offer a time effective way of recording an assessment of the sleep period close to the event, thus enabling a more detailed and richer micro-level account of the meanings, understandings and processes of sleep in everyday life.

3.3 Simple in its design and implementation, the audio diary method required women to record an assessment of their sleep within about 20 minutes of waking each morning for seven consecutive days using a hand-held Dictaphone. Respondents were given verbal instructions on the use of the Dictaphone, and brief written guidelines (see Appendix 1) on the types of information they might record in their audio diary, including the day/date; time going to bed and waking up; quality of sleep; strategies used to improve sleep; factors influencing sleep; and the influence of partner, if any, on sleep. In contrast to the structured nature of written sleep logs, these guidelines emphasised that respondents should feel free to record anything about their sleep, or other factors, now or in the past, which may have influenced their current sleep experience. Respondents were also asked to rate each night's sleep on a scale of 1-5, with 1 being 'very poor sleep', and 5 being 'very good sleep', to enable a comparison across nights of the week.

3.4 A total of 35 women, aged 40 and over, completed audio diaries, which were subsequently transcribed and analysed in conjunction with other qualitative and quantitative data to develop a more comprehensive understanding of women's sleep in mid and later life. By tape recording entries each morning after waking up, the diarists documented sleep episodes as close as possible to the event thus maximising recall and overcoming possible effects of interviewer intervention. The resulting data provides a picture of sleep as a fluid process, responding constantly to the changing realities of individual circumstances. Moreover, in alluding to the tensions inherent in balancing multiple roles as wives, mothers, and workers, the diaries highlight the gendered inequalities which underlie women's lives. These insights are discussed in detail elsewhere (see for example, Hislop and Arber, 2003a, 2003b, 2003c, Hislop, 2004).

3.5 Building on the success of audio diaries in generating data on women's sleep, the method has been adopted as part of a multi-method, cross-disciplinary approach in the ESRC-funded project Negotiating Sleep: Gender, Age and Social Relationships amongst Couples. Using audio diaries, in-depth interviews and other methods, this project investigates the interrelationship between the sleep patterns of partners aged 20 to 59 to identify how couples negotiate sleep both in terms of their 'self' and in terms of 'others' with whom they share the sleeping environment; and examines the extent to which sleep is a site of gender inequalities. Forty couples were interviewed together, with each partner then completing seven consecutive days of audio diary recordings (a total of 560 diary entries), using the same procedure as in the women's sleep project. Follow-up interviews with each partner a month after the initial interview provided an opportunity for researchers to clarify and expand on points of interest identified from the audio diary transcripts, thus gaining a deeper appreciation of the social context of the couple's sleep from the viewpoint of both male and female partners. Couples were sampled equally from four groups: under 40 with and without children, and 40 and over with and without children living in the home. The data from this project provide a resource for examining the narrative structure of audio diaries and for illustrating the types of insights which emerge from the use of this qualitative research method.

Narrative structure of audio sleep diaries

4.1 The audio diary is a form of narrative text in which the individual speaks in a monologue form to record their subjective impressions of the sleep period. It recounts the story of an individual's experiences of sleep each night over a seven-day period. Like paper diaries, audio diaries range in length from relatively short, concise accounts, to highly complex narratives of the night. In the couples' project, for example, completed 7-day diaries ranged in length from 282 to 3808 words. Women's diaries were on average 15 percent longer than men's, with the average length of completed female diaries 930 words compared to 809 words for men. Where both partners completed seven days of diary entries (N=33 couples), the length of women's diaries exceeded those of their partners in the majority of cases; with men recording longer diaries in under one third of couples.

4.2 Though diaries differ in format, the overriding narrative structure found in most audio diaries includes:

4.3 Thus in adopting a free text narrative format, audio diaries move beyond the paper diary format of checklist grids used to measure aspects of everyday life. Through sleep narratives, respondents offer the researcher insights into their intimate lives, played out against a backdrop of social constraints and circumstances. In the next section, we examine how sleep narratives, as constructed within audio diaries, shed light on the social context of peoples' lives.

Audio diaries as a window onto couples' lives

5.1 How individuals sleep can be conceptualised in terms of the interplay between physiological, environmental and social factors. At the physiological level, sleep responds to an individual's circadian rhythms (or the 'body clock'), and to their needs during the night such as going to the toilet, hunger, thirst, temperature and light. Our night-time behaviour, including whether we are restless, snoring, dreaming or sleep walking; hormonal factors especially during the menopause; increased health problems associated with ageing; and environmental factors such as seasonal variations, light, heat and noise may also impact on sleep:
I did suffer from arthritis last night. I took some medication at about 10.30 and I woke maybe about 3.30 suffering from pain in my neck and shoulders. (F > 40, no children)

Woke up about 3.30, mainly due to a 57 year old bladder. (M > 40 with children)

Rudely awoken at 3am by a lorry going along the top road here. Struggled to get back to sleep after that. (M > 40 with children)

5.2 Yet while sleep is influenced by these physiological and environmental processes, from a sociological perspective, it is the social context which, we argue, has more impact on how we sleep, particularly when the sleeping environment is shared. Influenced by our roles and relationships with significant others within the home; with people, institutions and schedules outside the home; and by biographical events and transitions such as marriage, parenthood, divorce, retirement and widowhood, our sleep patterns change across the life course. The audio diary acts as a mirror not only onto sleep but onto everyday life, with excerpts from diaries providing unique insights into sleep as a socially contextualised embodied process. We now focus on how data from audio diaries reflects the social realities of individual's and couples' lives.

The social environment of sleep

5.3 For individuals, a good night's sleep may be predicated on an expectation of personal space; 7-8 hours of uninterrupted sleep to recuperate from the day's activities; freedom from noise or interference from others, except in an emergency; and exemption from normal role obligations. For couples in our study, however, sleeping takes place in a shared environment in which the quality of one person's sleep is vulnerable to disturbance by significant others within the sleeping environment. Once in another's immediate presence, individuals will necessarily be faced with personal-territory contingencies (Goffman, 1959), potentially creating a tension between the individual self with its own physiological needs, and the shared self defined by relationships with others and by intrusions from the physical environment. While the implications of 'shared sleep' may only be realised in the conscious reality of pre- and post-sleep time and when sleep is interrupted, sleep quality is mediated by partner behaviour; lack of control over the sleeping environment; disruptions from others, be it snoring partners, or children; and by the extension of daytime roles into the sleep period through caring commitments and worries about work. This tension is illustrated by diary entries which reveal the impact of relational and institutional factors on sleep, including 'doing partner', 'doing parenting', and 'doing paid work'.

Doing partner

5.4 The 'doing' of partner is a culturally defined role which, in the UK, generally involves the sharing of a double bed and a commitment to caring for partner's well-being during the night. It involves balancing one's own needs and preferences with those of one's partner. Partners must negotiate, for example, bedtimes and wake up times, when to set the alarm, which side of the bed to sleep on, whether to have the windows open or closed, whether to have the heating on, what type of mattress and pillows to have, whether to read in bed or not, whether to sleep cuddled up or apart, and what to do about snoring and/or other behaviours that potentially disturb the other partner. The potential for disturbance between partners when sharing a bed is high:

She (partner) woke me up last night talking about something, and because she stole all the covers and it was freezing cold. I had to actually prise it off her body because she can be quite a quilt snatcher. (M <40 with children)

We've been going to bed a little bit later than I want, a little bit earlier than he (partner) wants and I think it's making him wake up early in the morning — instead of him coming to bed two hours later than me, he's waking up three hours earlier. (F <40 no children)

(Partner) did his lovely 'not snoring', but holding his breath … . I woke up a couple of times in the night to check he was still alive. (F <40 with children)

5.5 Data from audio diaries may also challenge stereotypes of partner behaviour, particularly in relation to snoring. While snoring is more common among men than women and is cited by women in audio diaries as a key factor in their sleep disruption, the myth that women don't snore, is contradicted in the following excerpts:

(Partner) was going through her full repertoire of snorts, grunts and snores but it probably wasn't as loud as I thought. But the fact that I was awake and she was asleep, added to the problem for me. (M > 40 with children)

(Partner) was snoring extremely loudly, which she did for most of the night. In fact, probably all of the night, so I hardly got any sleep at all. (M > 40 with children)

5.6 The opportunity to sleep alone when a partner is away can have both positive and negative effects on sleep. For the following women, having the bed to themselves provided a welcome respite from sleep that was usually disrupted by their partner:

Had a good night's sleep actually, 8 hours which is lovely. (Partner) wasn't here, so I had the bed all to myself. (F < 40 with children)

(Partner) fell asleep on the sofa, so I watched a film and then went to bed. Had a really, really good night's sleep. Had the whole bed to myself and I went straight off into a real deep sleep. (F < 40 with children)

However, others prefer the intimacy and security of sleeping together to sleeping apart:
(Partner) was back after being away on the Saturday night. It felt strange when she wasn't there. Felt more relaxed with her back. (M < 40 no children)
For half the couples in our study, parenthood added a further dimension to the sleep experience.

Doing parenting

5.7 Having children in the home can have a major impact on parent's sleep, with the potential for disturbance extending from birth to young adulthood. As the following excerpts show, men and women may both be disturbed by their children during the night, with younger children demanding attention, sleeping in their parent's bed, and waking up early; and teenagers coming in late and playing loud music:

Couple 1

Male
Good night's sleep, only again we were having problems with (6 year old son). I put him in his own bed, then he joined us at about 2am. (M > 40 with children)

Female
(Son) was causing us a few problems last night and wouldn't settle. (Partner) put him back in his bed but he joined us again at 2am and woke me up just to explain why he'd arrived. (F > 40 with children)

Couple 2

Male
Disturbed at 2.30 by (teenage son) coming in from the neighbour's. Away with the fairies — was playing music in the adjoining bedroom. Went and told him to turn it down, but the noise persisted, and I then had to get up and tell him again. (M > 40 with children)

Female
(Teenage son) was up a lot last night and he was very noisy when he came in. (F > 40 with children)

5.8 The impact of children on parent's sleep is highlighted in diary comments on perceived improvements in sleep when children are away:

Woke up this morning at 8, which was lovely, because the children aren't here, they've gone to stay with their dad — it was lovely just to lie in this morning without any interruptions. (F < 40 with children)

5.9 Yet while men and women may be equally disturbed by their teenage children's noise, audio diaries suggest that it is the female who remains primarily responsible for children's well-being during the night, carrying out the physical and emotional labour associated with caring. This commitment to care extends beyond childhood, with mothers continuing to feel a sense of responsibility for their children's night-time well-being into adulthood:

(7 year old daughter) was ill around 5 this morning, so I was up until about 6, just clearing up and sorting her out and getting her back off to sleep. (F < 40 with children)

Did wake up a few times in the night with my daughter (9 months old) crying. First at 12 o'clock. Went and rubbed her head, she went back to sleep, tucked her back in, then I went straight back to sleep. And then again later in the night, not sure what time, but I didn't have to get up. (F < 40 with children)

I was awake at half past five this morning and I think that was more about knowing my son had to get up to get to the airport. I was very aware that he needed to be getting up at that time. (F > 40 with children)

5.10 The role of women as primary carers is further exemplified when audio diary entries of each partner in couples with children are compared. In the following diary excerpts from a couple aged under 40 with children, experiences of the night differ considerably for each partner despite sharing the same sleeping environment. It is the female partner who takes on the role of caring while her partner gets a good night's sleep:

Judy's experience
Sunday morning. Took ages to go to sleep, because (son) woke up and the cat changed its mind and needed to go out, and everybody seemed to need me just as I was dropping off, so I was up and down a bit for a while and got to sleep … … . got woken up with a start at quarter past six with (son) shouting "Mummy", because it was time for him to get up I think, got up and put him back in his bed, settled him down again, went back to bed for what felt like eternity, but actually was about ten minutes before he started shouting again, just enough for me to wake up gently maybe and that's it really. I just feel like I have been hit by a bus today, my limbs ache, perhaps it's time for a new mattress (laugh).

Tom's experience
Sunday morning, Had a good night's sleep, woke up a couple of times, no particular reason. (Partner) said she got up quite a few times, didn't really notice. I notice she got up this morning quite early, and then I finally woke up and got out of bed, must be about half seven, quarter to eight I suppose and that, good night's sleep. That's it really I think.

5.11 Yet while the above excerpts reinforce the view that gendered inequalities associated with caring are perpetuated during the sleep period, references in audio diaries suggest that some men are also taking an active role in parenting particularly during the teenage years. Like their female partners, they express concerns about the welfare of teenage children, often waiting until they return home safely before they can get to sleep, or having their sleep disturbed by teenagers when they get home:

Not the best of night's sleep. I was aware that (teenage son) hadn't come home — half awake a couple of times during the night. In fact I think he was in about 2.30, but he was very quiet and I didn't hear him. (Male > 40 with children)
For the majority of respondents, paid work plays a central role in structuring their everyday lives, and, as a consequence, their sleep.

Doing paid work

5.12 Paid work roles often have a significant impact on an individual's (and their partner's) sleep. The intrusion of institutional factors such as work and school schedules into natural sleep rhythms is often symbolised in sleep accounts by references to the alarm clock and by use of language suggesting obligation rather than choice:

When it's a work day morning, which for me is six days a week, the alarm goes off at 5.40 in the morning. We don't get up till 6.30, but the alarm buzzer goes off every ten minutes. (F > 40 no children)

Had to get up at quarter to six, because I'm off to work today. (F > 40 with children)

I woke up annoyingly about 15 minutes before the alarm went off and knowing it was a Monday morning was horrible, especially as the winter is now here and it's so dark. (M < 40 no children)

5.13 In a shared sleeping environment, the incompatibility of each partner's schedules may also affect sleep quality. Individuals may be awoken by their partner regardless of whether or not they need to get up at that time. This can be particularly problematic when partners start work at different times:

The alarm went off at 5.15 for my husband to get up. The second alarm went off at 5.25. After he left the house I heard the car start and him go, and then I dozed and didn't properly go back to sleep, and the alarm went off for me at ten to seven, at which time I got up to get ready to go to work. (F > 40 with children)

Got woken up about 6.15 this morning when (partner) got up to get herself ready for work. (M < 40 no children)

5.14 Audio diaries also provide opportunities for individuals to reflect and comment on how their work lives may influence their sleep. As Jokinen (2004: 340) states, diaries can be sites of 'individual confession and self-reflection'. In the following excerpts, individuals use the audio diary as a passive listener through which to reveal the incompatibility of shift work, family life and sleep; and the frustrations inherent in working at night and sleeping during the day:

Kids were a bit noisy today, kept waking me up, so can't have been in all that deep a sleep … . (Partner) kept coming upstairs and trying to point things out to me in a magazine, but (I) wasn't really interested. I just wanted to sleep. (M < 40 with children)

Silly little things (annoy me), like having breakfast with my wife (when I get off shift in the morning), which'll take half an hour, but really it'll take like an hour, because of the conversations that surround it, and all of a sudden I've wasted half the morning and I'm f … knackered when I go to work. It's no way to live my life … . Anyway I just wanted to share that with you without anybody listening. (M < 40 with children)

5.15 The key role which work plays in structuring our sleep is highlighted by changes in sleep patterns during weekends and holiday periods. These changes not only relate to the diarist's own life but to those of the people with whom they share their sleeping space:

It's Saturday today, so it's my only lie in of the week. (F < 40 no children)
It's half-term, so I won't be waking up to take (daughter) to school, so I've been sleeping in. (F < 40 with children)
Experiences of work and study referred to in audio diaries suggest an awareness of the relationship between work and sleep:
Looks like I've got a normal day at work today, no real pressures, so I expect I'll probably have a normal pattern evening and then go to bed again around 10.30. (M < 40 no children)
I did some college work in the evening as well, so by the time I went to bed I was very tired but also very wound up. Consequently it took me longer than usual to get to sleep. (F < 40 no children)
Yet while for the purposes of discussion we have isolated the influence of parenting, partnerships and paid work on sleep, in real life these factors interact, contributing to a complex sleeping environment.

Interaction of physiological, environmental and social factors on sleep

5.16 Teasing out the causes of sleep disruption is rarely as simple as isolating a single physiological, environmental or social factor which prevents a good night's sleep. Understanding sleep involves putting pieces of a complex jigsaw together. As the following excerpt shows, an individual's sleep duration and quality is subject to multiple influences. These influences may be caused by factors internal or external to the sleeping environment.

Sunday the 15th, uh went to bed last night about half past ten, uh, (partner) got up at six to make a cup of tea, so from half six I was wide awake, uh, but didn't get up till quarter past seven, uhm. I didn't hear my daughter come in, but I heard my son come in, I think it was about two-ish and I had a conversation with him, and I got up and went to the loo and had a drink of water about three fifteen, I think it was, uhm. And following on from that, (partner) was snoring quite a lot which kept me awake on and off, uhm, and then there was a strange noise, a bit like an alarm or something outside, that we heard on and off until he woke up at six, uhm, and once again I had rather a few strange dreams. Uhm, think that's it. (F > 40 with children)

5.17 Audio diaries thus highlight how the nature and quality of sleep respond to individual physiological factors (for example, needing to go to the toilet), environmental factors (for example, noises outside) and social factors (for example, disturbances from children and partner). Used in conjunction with other methods, they play a key role in uncovering the complexity of the social context of sleep. As individual assessments of sleep recorded within 20 minutes of waking for a period of seven days, the diaries allow the researcher access to experiences which would otherwise be unobservable. Moreover, if recorded later in the day, it is likely that the detail of some of these night-time occurrences and actions, and the individual's reaction to them, would have been forgotten. In this sense, Elliott's (1997: 2) observation in relation to the contribution of written diaries to sociological research is equally valid for audio diaries:

'(the) proximity to the present, the closeness between the experience and the record of experience means that there is the perception at least that diaries are less subject to the vagaries of memory, to retrospective censorship or reframing than other autobiographical accounts.'
We argue, however, that while paper diaries can be an important adjunct in research, particularly in measuring activities across a 24 hour period, audio diaries provide richer and more comprehensive insights into the social context of people's lives. Drawing on data from both the Sleep in Ageing Women project and the Couples' project, the next section assesses the use of audio diaries in social research by highlighting some of the advantages and disadvantages of the method.

Assessing the use of audio diaries in social research

Comparing audio diaries with paper diaries

6.1 Data from the Sleep in Ageing Women project illustrated the benefits of using audio diaries rather than paper diaries. While the majority of women felt comfortable using Dictaphones, a few older women preferred to write their diaries. Women who chose to write a diary record for each night's sleep for a week were given exactly the same written and verbal guidelines as those recording audio diaries. As the following excerpts show, the data generated by the audio sleep diaries were much more detailed and were presented in the form of a narrative, recounting women's sleep experiences in relation to their roles and relationships. Written accounts, in contrast, listed facts in log form about the respondent's sleep, but provided little affective commentary and few links to everyday life.
Audio account:
Tina, age: 44, living with partner, daughter aged 14, works part-time

Wednesday
Last night I slept very heavily, I had a very long night's sleep. Today's my day off, so I tend to lie in on my day off. I went to bed about midnight, in fact I was so tired I fell asleep on the bed before I finished getting myself ready for bed, so I had to wake up again and get into bed. I think I'd give my night's sleep a rating of about 4 (on a five-point scale) because I woke up with a bit of a headache this morning, probably too long asleep. I first woke, my partner woke me up about 7.50, he was anxious that my daughter was going to miss her bus, so he woke me just to tell me that, and then I went back to sleep again. A calm night — my partner has currently broken his ankle so he has tended to be rather restless, but I didn't notice last night at all. I wasn't aware of dreams except that when he did wake me up at 7.50, I was dreaming about Paris for some reason. No strategies for getting ready for bed, I didn't need them last night, just very tired, I've had a very busy fortnight including a house move and obviously extra work involved with looking after my partner, because he can't do all sorts of things around the house, I think it's made me extra tired. Oh, something I forgot to add — I went to sleep, my partner comes to bed later than me, and I go to sleep with a small lamp on low, down on the ground with a cover on it, so I went to sleep with that dim light on and he came to bed after an hour or so, I heard him coming, he's on crutches so I heard the crutches coming down the hallway, and I was aware of him getting into bed, but I just sort of came to and went straight back to sleep.

Written account:
Gloria, age: 71, married, adult children

Monday
Scale: 4
Hours (time went to bed, time got up): 2300-0635
Quality (of sleep): sound, but aware of back aching near dawn
Strategies (what she did to help her sleep): short walk before going to bed
Factors (influencing the way she slept): restlessness late evening
Partner influence: none
Partner comment: none
Waking (how she felt when she woke up): naturally, rested
Other (comments about her sleep): none

6.2 Whereas the first (audio) excerpt provides fascinating insights into Tina's life and her resultant sleep patterns, the second (written) excerpt relies heavily for structure on the guidelines given to respondents with very few embellishments. The audio excerpt provides not only information about Tina's sleep, but about social factors which impact on the quality of her sleep, including her roles as wife and mother, her paid work, and the added responsibilities associated with her partner's injury and moving house. While Tina had no problems sleeping, her exhaustion from these multiple roles and responsibilities leads to a heavy sleep pattern which leaves her feeling groggy the following morning. It is this detail which is invaluable for the social scientist in exploring the relationship between sleep and social context.

6.3 The inadequacy of paper diaries in the Sleep in Ageing Women project was addressed in the Couples' project where participants were specifically asked to complete audio diaries, without the option of a paper alternative. Respondents were all under age 60 and were more confident with the use of electronic media. They thus accepted tape recording, rather than writing, an assessment of their sleep. In addition, the opportunity to follow-up each audio diary when interviewing individuals a month after completion enabled clarification of details and further probing of diary entries where necessary. Moreover, in both projects, audio diary data were not analysed in isolation, but formed part of a multi-method approach to understanding the nature of sleep in its social context.

Completion rates

6.4 Completion rates for audio diaries in both the Sleep in Ageing Women project and the Couples' project were very high, with the vast majority of respondents recording sleep assessments every night over the seven day period. In the Couples' project, one couple failed to complete all audio diary entries after forgetting to take the Dictaphones on a weekend away break. On return, while the female partner resumed her diary, the male felt there was little point in continuing given the missing nights. Another respondent, when asked why she had completed entries for only four nights, replied that she had in fact recorded entries every morning, implying either a possible fault with the tape or in her operation of the equipment. A small number of male and female diarists missed one or two nights of diary entries for which no reason was given.

6.5 While in the studies discussed in this article, faults with the equipment were minimal, researchers should be aware of the need to instruct participants carefully in use of the Dictaphone to ensure quality of data and to overcome possible frustration during the recording process. As in these studies, participants should be given a contact number to report mechanical breakdowns or other problems with the equipment if they arise.

6.6 Motivation to complete audio diaries was possibly enhanced in the Couples' project by the involvement of both partners in the study, and by the payment of an incentive. Partners participating in the study were able to remind each other of the need to complete diaries, thus reinforcing the habit of recording audio diary entries each morning. We acknowledge that the payment of an incentive to respondents for participating in the study may have had an effect on completion rates, however, it was felt that the need to compensate respondents for their time and inconvenience outweighed any possible impact on study outcomes. The completion rates in both projects suggest that the method is effective as a means of recording assessments of sleep, and may be preferable to completing written paper-based diaries soon after waking.

Cost

6.7 Audio diaries were a cost-effective means of collecting data in both sleep projects. The main outlays included the cost of Dictaphones and travel costs to pick up completed diaries. In the Couples' project, for example, eight basic model Dictaphones were purchased at a cost of around £25 each. These were rotated between couples as interviews were completed, and collected by the researchers a week after the initial couple interview. It would also be possible to ask participants to return the Dictaphones containing completed audio diaries by registered post, in cases where travel to pick up diaries was not feasible.

6.8 In both studies all Dictaphones were returned in good condition. However, this may not be the case when Dictaphones are used with children or teenagers. Venn, in pilot work investigating teenagers' sleep, reported that young people either failed to return the equipment or did not record their entries properly resulting in lost data. She found that while completed audio diaries gave good insights into participants' sleep, the use of Weblogs or email may be preferred media for recording sleep experiences amongst this age group.

6.9 A further cost to be taken into account when using audio diaries is transcription. In the studies discussed here, the length of diaries was such that transcriptions were carried out by the researchers, taking on average half to one hour to transcribe a week's audio diary. Transcriptions were completed in advance of the individual interviews and informed the nature of questioning at these interviews. Transcribed data were then input into QSR NVivo for analysis alongside data from the joint couple and individual interviews.

Ethical issues

6.10 Ethics are a matter of balancing the researcher's desire to uncover the hidden dimensions of social life while protecting the rights and integrity of research participants (Bulmer, 2001). In giving informed consent to participate in research, participants give the researcher access to their private world as a means of informing knowledge and understandings of sleep. Yet, while the maintenance of anonymity, privacy and confidentiality is ensured by the researcher, the research may inevitably impact on the lives of respondents.

6.11Comments in audio diaries provide the researcher with an indication of the effect of research methods on participants' sleep. In the Sleep in Ageing Women project, for example, while respondents' reactions to the use of audio diaries was in general positive, their narratives highlighted the possible intrusion of the method on the quality of their sleep:

Mary, age 64, married, three adult children, retired

Sunday
'...... I certainly haven't slept very well this week. I'm wondering whether it is because I am doing this sleep survey for you. You never know. ....... Anyway I do hope that this will have been of help to you and I have quite enjoyed doing it.' (our emphasis)

Tamara, age 71, married, three adult children, retired

Monday
(final comment) 'I think the monitoring (of my sleep through the sleep diary) had an effect (on my sleep) too. I was always aware of it.' (our emphasis)

6.12 Diary entries from the Couples' project also suggest that the methods used in researching sleep may encourage an unnatural focus on sleep; an event which would otherwise be taken for granted. This is particularly evident in entries for the first night of the audio diary, but may also be reported when individual's reflect on their week's sleep:

Really mish mash, uhm, odd nights sleep, probably due to the fact that the whole sleep interview and everything was playing on my mind. (M <40 no children)

Keep wondering whether these little waking up in the night, whether it's, this week's been bad or whether it's been always like that, but I don't really take much notice of it. (F < 40 no children)

6.13 These comments highlight the need for the researcher to be aware that any research method implies an intrusion, which may contaminate the social environment that is being studied. Through individual audio diary entries such as those above, the researcher is made more aware of the possible influence of method on the sleep experience. While some effect of the research method on people's lives is inevitable in any study, the use of follow-up interviews can provide a forum for discussion of any issues arising.

6.14 A further concern for researchers in using audio diaries is the extent to which participants should be given guidelines. Asking participants to give specified factual data about their sleep, such as time of going to bed and getting up, is in effect the spoken equivalent of a semi-structured paper diary. Thus it could be argued that there is little benefit recording this information on a Dictaphone rather than in a paper diary. However, the overall audio diary entry provides both facts about each person's sleep, enabling comparison of individual sleep patterns across the week's diary entries, and between participants, as well as more detailed accounts of aspects affecting their sleep as illustrated earlier. Requests for information about factors which disturbed sleep may also focus attention on the pre-determined interests of the researcher, rather than allowing the participant the freedom to record what is important to them. While we believe that semi-structured guidelines were essential in this project to uncover the social context of sleep among couples, in retrospect a more unstructured approach which simply asked participants to 'talk about their previous night's sleep' may have produced valuable insights which we had not considered. Future research could usefully compare the types of data which emerge from semi-structured and unstructured audio diary guidelines.

Conclusion

7.1 The use of audio diaries is a viable means of capturing and understanding experiences of sleep, enabling the researcher entry to an otherwise hidden private phenomenon. Through audio diary narratives, respondents not only provide factual information about their sleep patterns and the quality of their sleep, but 'paint' pictures of their lives. Seen through the frame of reference of the respondents, audio diaries provide insights into the roles and relationships, and life events and transitions which impact on everyday life. In the absence of 'real time' access to and observation of respondents' sleep, audio diaries provide an invaluable resource for the researcher through which to interpret the experience and meaning of sleep in its social context.

7.2 Audio diaries enable respondents to record their experiences of sleep soon after waking up, thus minimising the gap between the sleep event and data collection, and reducing problems of retrospectivity and recall associated with other methods. For the researcher, audio diaries are a cost-effective, non-intrusive means of collecting data on the intimate world of sleep. The resulting data are compact and relatively easy to transcribe and analyse. For the respondent, the convenience of using a Dictaphone, and the short time needed to record each diary entry, increases acceptability of the method in the field, and contributes to improved completion rates over written paper diary methods. Moreover, the fact that individuals work independently in recording diary entries, reduces the impact of researcher presence on data.

7.3 Audio diaries are not designed to replace the use of logs, grids and other paper diary formats, where the primary aim of the researcher is to measure activities and attitudes. Nor are they meant to take the place of written diaries when used as an adjunct to qualitative interviews, or as a resource for social historians. Rather, as a complement to other methods, audio diaries provide an effective means of data collection, particularly in studies where the researcher is called upon to investigate 'behind the scenes' aspects of everyday life. In this regard, the use of audio diaries can be readily extended beyond the parameters of sleep research to a wide range of research contexts. By narrowing the gap between lived reality and retrospective interpretation, audio diaries provide a simple and effective means of bringing to consciousness those everyday events and activities which we carry out routinely and largely without reflection.

Notes and Acknowledgements

1 Polysomnography uses electroencephalograms (EEG) to measure brain wave patterns during sleep. This resulted in the identification in the early 1950s by Aserinsky and Kleitman (1953) of the basic pattern of sleep in humans, comprising 90 minute cycles of REM (rapid eye movement) and non-REM sleep

2 Actigraphy is the measurement of movement during sleep through a watch-like device worn on the wrist. It is used to distinguish periods of sleep and of wakefulness.

3The Sleep in Ageing Women project was funded by the European Union, grant no. QLK6-CT-2000-00499.

4The Negotiating Sleep: Gender, Age and Social Relationships' project was funded by the ESRC, grant no. RES-000-23-0268.

Appendix 1

Audio sleep diary guide (Sleep in Ageing Women)

Dear Participant

Thanks for agreeing to record an audio sleep diary as part of the Women and Sleep project. This will involve recording a brief summary of your sleep as soon as possible after you get up each morning for a period of 7 days. The following notes provide a guide to the sort of information you may include, but feel free to record any points about your experiences of sleep or other factors, now or in the past, which may influence your current sleep patterns.

At the beginning of the tape please record your name and age.

For each diary entry, please record, where applicable:

When you have completed your audio sleep diary, please phone (researcher) on (phone). Thank you for your participation. (signed)

References

ASERINSKY, E. and Kleitman, N. (1953) Regularly occurring periods of eye motility, and concomitant phenomena, during sleep, Science, 118: 273-74.

AUBERT, V. and White, H. (1959a) Sleep: a sociological interpretation 1, Acta Sociologica, 4(2): 46-54.

AUBERT, V. and White, H. (1959b) Sleep: a sociological interpretation II, Acta Sociologica, 4(3): 1-16.

BRYMAN, A. (2001) Social Research Methods, Oxford University Press: Oxford.

BULMER, M. (2001) The ethics of social research. In N. Gilbert (Eds.) Researching Social Life, London: Sage, pp. 45-57.

BYTHEWAY, B. and Johnson, J. (2002) Doing diary-based research. In A. Jamieson and C. Victor (Eds.) Researching Ageing and Later Life, Buckingham: Open University Press, pp. 156-174.

COXON, A. (1988) 'Something sensational.....' The sexual diary as a tool for mapping detailed sexual behaviour, The Sociological Review, 36(2): 353-67.

COXON, A. (1994) Diaries and sexual behaviour: The use of sexual diaries as method and substance in researching gay men's response to HIV/AIDS. In M. Boulton (Eds.) Challenge and Innovation: Methodological Advances in Social Research on HIV/AIDS, London: Taylor & Francis.

ELLIOTT, H. (1997) The use of diaries in sociological research on health experience, Sociological Research Online, 2(2): 1-13 <http://www.socresonline.org.uk/2/2/7.html>>.

GOFFMAN, I. (1959) The Presentation of Self in Everyday Life, Penguin: London.

HISLOP, J. (2004) The Social Context of Women's Sleep: Perceptions and Experiences of Women aged 40 and over, unpublished PhD Thesis, University of Surrey, Guildford.

HISLOP, J. and Arber, S. (2003a) Sleepers Wake! The gendered nature of sleep disruption among mid-life women, Sociology, 37(4): 695-711.

HISLOP, J. and Arber, S. (2003b) Understanding women's sleep management: beyond medicalization-healthicization?, Sociology of Health and Illness, 25(7): 815-837.

HISLOP, J. and Arber, S. (2003c) Sleep as a social act: A window on gender roles and relationships. In S. Arber, K. Davidson and J. Ginn (Eds.) Gender & Ageing: Changing Roles and Relationships, Maidenhead: McGraw Hill/Open University Press, pp. 186-205.

HISLOP, J. and Arber, S. (2004) Understanding women's sleep management: beyond medicalization-healthicization: A response to Simon Williams, Sociology of Health and Illness, 26(4): 460-463 <http://ejournals.ebsco.com/direct.asp?ArticleID=4B128E193C9272316FCD>.

JOHNSON, J. and Bytheway, B. (2001) An evaluation of the use of diaries in a study of medication in later life, International Journal of Social Research Methodology, 4(3): 183-204.

JOKINEN, E. (2004) The makings of mother in diary narratives, Qualitative Inquiry, 10(3): 339-359.

MEADOWS (2005) The Negotiated Night: An embodied conceptual framework for the sociological study of sleep, The Sociological Review, 53(2): 240-54.

NATIONAL STATISTICS (2003) UK 2000 Time Use Survey, <http://www.statistics.gov.uk/timeuse/default.asp>.

PLUMMER, K. (1983) Documents of Life, George Allen and Unwin: London.

ROBERTS, B. (2002) Biographical Research, Open University Press: Buckingham.

SCHWARTZ, B. (1970) Notes on the sociology of sleep, Sociological Quarterly, 11(Fall): 485-99.

TAYLOR, B. (1993) Unconsciousness and society: The sociology of sleep, International Journal of Politics, Culture and Society, 6(3): 463-471.

WILLIAMS, S. (2002) Sleep and health: Sociological reflections on the dormant society, Health, 6(2): 173-200.

WILLIAMS, S. (2004) Beyond medicalization-healthicization? A rejoinder to Hislop and Arber, Sociology of Health and Illness, 26(4): 453-459 <http://ejournals.ebsco.com/direct.asp?ArticleID=414EA0F306574EBECD08>.

WILLIAMS, S. (2005) Sleep and Society. Sociological Ventures into the (Un)known, Routledge: London.

WILLIAMS, S. and Bendelow, G. (1998) The Lived Body. Sociological themes, embodied issues, Routledge: London.

WILLIAMS, S. and Boden, S. (2004) Consumed with sleep? Dormant bodies in consumer culture, Sociological Research Online, 9(2): 1-10 <http://www.socresonline.org.uk/9/2/williams.html>.