'Is this thing working?'--the challenge of digital audio for qualitative research.
Abstract: The archiving of qualitative research data has developed greatly in recent years yet it is still the subject of much debate. Drawing on the experience of advising researchers and negotiating data deposits for the UK Data Archive, this article explores some of the common concerns about archiving. Particular emphasis is given to audio data as it represents a common stage in the creation of qualitative research data but is also the least likely to be archived. Some key issues relating to the status of audio data and its relationship with interviewing and transcription will be outlined. Concerns about the disclosive nature of digital sound recordings will be used to explore some of the wider issues of research practice in relation to preservation and sharing. The role of confidentiality agreements in allowing a more effective preservation and use of digital audio will also be explored.

Keywords: Audio, Archiving, Recording
Article Type: Report
Subject: Social science research (Technology application)
Sound recordings (Usage)
Qualitative research (Technology application)
Author: Southall, John
Pub Date: 03/22/2009
Publication: Name: Australian Journal of Social Issues Publisher: Australian Council of Social Service Audience: Academic Format: Magazine/Journal Subject: Sociology and social work Copyright: COPYRIGHT 2009 Australian Council of Social Service ISSN: 0157-6321
Issue: Date: Spring, 2009 Source Volume: 44 Source Issue: 3
Topic: Computer Subject: Technology application
Product: SIC Code: 3652 Prerecorded records and tapes
Geographic: Geographic Scope: Australia; United Kingdom Geographic Code: 8AUST Australia; 4EUUK United Kingdom
Accession Number: 222315601
Full Text: Introduction

This paper examines some of the common assumptions about the status, creation and use of audio recordings in qualitative social science research. In particular it will look at its representation within archived data collections. It will be argued that within research practice audio is often overlooked and relegated once text based material is produced. Except for a few notable exceptions it is routinely undervalued. Possible reasons for this view of audio material will be discussed. Some elements of the relationship of sound recordings to text will be examined as well as its relationship with methodology, consent and confidentiality. The challenges and prospects for digital audio as an archivable and re-usable resource will then be used to illustrate wider issues in re-using and archiving qualitative data.

The number of national data archives continues to grow, as shown by the website of support organisations such as the Council of European Social Science Data Archives (CESSDA 2009). However few have committed to the archiving and sharing of qualitative research data. Exceptions to this can be found in the United States, Finland and Australia. In the United Kingdom the Economic and Social Data Service, and in particular ESDS Qualidata promotes the archiving of such research as well as providing general guidelines on the creation and management of qualitative research. Many of the ideas discussed in this paper have been developed through work for ESDS Qualidata--the qualitative service of the UK Data Archive. The ability to effectively archive qualitative research has been based on developing rigorous systems for controlling access to collections by the UK Data Archive. This is outlined at length on the UK Data Archive website and elsewhere (Bishop 2005). Whenever data sharing or archiving is discussed in this paper, it is therefore presumed it is in the context of such systems of access control and user registration.

Despite the domination of data archiving by quantitative archives the corpus of qualitative research data available to researchers is steadily growing. A review of what is available in the case of the UK Data Archive shows that, like its international counterparts, it is predominantly text based consisting mainly of interview, focus group and field observation transcriptions. Audio recordings are rarely deposited. However, there is traditionally a longstanding relationship between recording technology and qualitative methods (Lee 2004) so this absence is puzzling. Reading qualitative interview transcripts made available through archives suggests that it is common for the first exchange between researcher and respondent to be an oblique reference to the recording equipment; 'I'll just check this is running ...' or sometimes the less confident '... is this thing working?' But once it is confirmed everything is operating properly the captured audio, the initial raw research material, is seldom mentioned again. The same attitude persists when multi-media research is offered for archiving and textual material becomes the focus of attention. Audio recordings created in the course of the project are routinely withheld, unavailable or destroyed. When considering research practice, from the position of archiving and sharing research data, perhaps the same question 'is thing working' could be posed.

There is evidence, as Lee (2004: 881) observes that:

Of course there are some notable exceptions to this pattern. For example, research in disciplines such as conversational analysis, oral history and socio-linguistics, place a greater value on audio material and use it extensively throughout their research practice. Nevertheless even in these cases doubts about the suitability of archiving audio data persist and are exposed by the limited amount of audio being archived. As is the case in other disciplines little is made available after its use within the primary project. The reasons for such assumptions are varied and will now be explored.

Assumptions Regarding Interview and Transcription

At its most fundamental level, qualitative research is associated with the use of interviewing. It remains the most accepted method of providing .access to the meanings people attribute to their experiences and social worlds (Miller and Glassner 2004: 134). Moreover, as Mason (2002: 39) has noted:

Whilst interviews may vary greatly in terms of formality, structure and content, the first objective of all researchers is generally the same. This is to develop a rapport with the person being interviewed and to create a relationship allowing the free flow of ideas. This means that the researcher/informant relationship is one of mutual trust and some intimacy (Gibbs 2007: 101) with strong associations of confidentiality. In most cases this interaction is captured--either through written field notes or audio recording--and eventually analysed and preserved in the form of a text transcription. Traditionally the use of transcription as the basic form of raw data has been accepted within qualitative research. Nevertheless it has been noted by Oliver et al. (2005: 1273):

The general lack of reflection on the use of transcription and indeed on the various influences and processes that lie behind their production is a key point. The spirit of enquiry and will to enlightenment that underlies qualitative research is--in itself--not enough to gloss over these concerns. It may be argued that there is a need to consider how transcriptions are created, their general validity, and the way each is moulded by the particular research interests of a project. Indeed there is a body of criticism which seeks to examine the use and the 'taken-forograntedness' of transcription work (Tilley 2003: 753). It is one that argues the transcript itself is a product (Barbour 2008: 193). If it is to be viewed in such as way then it could be said to be a product that represents the final stages of the conversion of audio recordings into textual material. It may be rewarding to add to this discussion by also considering the creation and use of audio as representative of earlier stages of audio-to-text translation. It is a common stage in all qualitative research, and yet, like making field notes and transcriptions, it is hidden and part of the invisible oral tradition of craft knowledge (Hammersley & Atkinson 1995: 177).

In the majority of qualitative research projects audio recorders are seen as a valuable aid to the process. Their use can range from augmenting field notes to replacing the need to produce any notes whatsoever. Even so, there can still be concern about the impact of such equipment and recordings on the general interview relationship. Such concerns may vary. At its simplest, a case may be made that recording inhibits the interview and acts as a potential barrier to a free flow of conversation. Bryman (2001: 322) comments:

Frequently the appropriate use of audio recordings are, once again, described in terms of craft knowledge (Lapadat & Lindsay 1999: 69) or field craft skills (Kvale & Brinkmann 2008) which implies they should remain in the background of most research projects. It must be conceded, there are genuine concerns about the use and sharing of audio created as part of qualitative research (Corti 2000: 10). The view that such material will always be problematic contrasts strongly with the relatively straightforward use of transcripts in most investigations. Concerns about the use of audio in research projects can be illustrated in a number of ways. First of all it can be viewed from the perspective of creating audio resources and using recording equipment. Secondly it can be viewed from the perspective of the data management of a research project. Thirdly it can be viewed from the perspective of how audio is used within a research project after it has been created.

Creating Audio Recordings Within A Research Project

An early step in the practice of creating audio is the choice of recording equipment. It may be argued that the majority of researchers view such decisions as a purely technical exercise (Lee 2004: 870). However, a lack of interest in audio technology by some researchers overlooks the potential benefits to research practice of the replacement of analogue technology by digital. These developments have significant implications for the creation of audio as a resource and as a type of usable and archivable data within qualitative research. At the most basic level digital sound recorders offer the opportunity for much better quality recordings than analogue cassette tape when working in the field. The digital files that are produced can be stored on a variety of devices and content can be accessed quickly--much more quickly than analogue. In addition, when digital audio needs to be copied this is virtually instantaneous with no loss in quality. These characteristics of digital audio offer improved prospects for a much wider sharing of audio. It also offers the opportunity for audio to be used in different ways by different researchers. For example the objective need not be simply to reduce it to text before analysis (Brown 2002: 10) as it may have other potential for communicating findings (Maloney & Paolisso 2001) or for future analysis.

An example of how digital audio can be used to widen access and use of qualitative resources can be seen in the Pioneers of Qualitative Research website. This is a web based resource recently launched by ESDS Qualidata that makes available online a series of in depth interviews with leading qualitative researchers. Each interview covers early years, intellectual development, fieldwork, key influences and discussions of major works. The project has interviewed researchers such as Peter Townsend, Ray Pahl, Leonore Davidoff and Raymond Firth. Only a few are available at the moment but more are being added. In each case thematic selections from the interview are identified and then made available in the form of matching text and audio extracts. The full interview is also offered as a downloadable file. The interview project that lies behind this website has been running for almost twenty years and most of the original materials were analogue; paper transcriptions and cassette tape. Access to this was limited as there were only a few physical copies. Therefore this resource only became possible once the materials were digitised and made available as digital text and audio files.

Data Management and Qualitative Audio

Many of the decisions relating to the use of audio technology to create audio recordings will form a part of a projects overall data management plan. This is especially important as collecting digital audio will increase the number and volume of files being administered. Indeed, it is a commonly acknowledged (McLellan-Lemal 2008: 165) that there is a tendency in qualitative research to produce enormous amounts of data. In response to this, data management planning within projects has become increasingly important (McLellan-Lemal 2008: Chapter 8). A rigorous data management plan helps ensure that data are well managed throughout the duration of a research project. (UK Data Archive 2009: 4) allowing efficiency in retrieval and analysis (Padgett 1995: 75). Data management may include such issues as naming and structuring of documents, coding, file tracking, version control and procedures on data access within a research team (Miles & Huberman 1994: 7).

Creating digital audio files as part of a research project also adds to the problem of tracking and managing digital material. For example as audio files are significantly larger in size than text files, planned storage capacity for a project has to account for this. In addition, authors of a data management plan have to recognise that working with digital audio places additional demands on a researcher. In particular a familiarity with the technology is required so that potential dangers of digital data loss may be anticipated. Digital audio, like all other digital files, are vulnerable to the problem of changes in hardware and software potentially rendering them unreadable. Furthermore the physical media that hold the data--such as hard drives, memory cards or recordable DVDs--can also fail. This can be potentially very serious for a project and often there is little warning of such failure. An effective data management plan should be designed to cope with such occurrences, as well as the more usual demands of data organisation, retention and use. As part of this it should aim to produce documentation regarding the audio files being created and information on how they were created.

All decisions regarding audio capture, from choice of microphone through to type of audio file, should be taken in relation to the wider goals of the research project. In this way data management interacts with the methodology of the study and has a strong influence on how data is used. However the control implied by data management may not always be completely positive. It may be argued it has also played a role in downgrading the place of audio recordings in research. Discussion of handling and analysing qualitative research data often stresses problems of reducing and coping with the volume of material being produced (Bryman & Hardy 2008: 4), (Richards 2005: 57). Whilst this forms an important stage in analysis, it also often presents audio as 'raw material'. Only after data management has co-ordinated its conversion into text does it become manageable data. This tendency for data management to narrow the focus of what is defined as data has also been noted in the context of its use with Research Ethics Committees (Boden et al. 2009: 738). Nevertheless the overall role of data management is positive and it does not have to routinely downgrade non-textual data. Indeed failure to appreciate the importance of such data management can be serious. As has been pointed out (McLellan et al. 2003: 69)

The Use Of Audio, Transcription And Disclosure

As has been argued already, it is common for audio recordings to be reduced in value as part of data management practices and put aside once text transcriptions have been prepared. It may further be argued, that this reflects some of the governing assumptions many researchers have about the use of audio after it has been created. These assumptions are complex and varied. Some have their origins in working practices established when using analogue audio. The time consuming work of accessing, copying and sharing analogue recordings in a fast moving research context, branded audio as cumbersome and encouraged a quick shift to using text transcription as the 'real' data. Yet many of these issues have disappeared with the development of digital audio. Sound files can be quickly accessed, searched, copied and easily shared. Other reasons are more fundamental and reflect the nature of academic discourse which is itself predominantly text based. Many researchers promote the methodological potential of digital multi-media either as computer assisted analysis (Fielding & Lee 1998) or hypermedia (Dicks et al. 2005) but these are currently still in the minority. The dominant forms of academic discourse and legitimation--reports, publications, journal articles, and so on--remain text based. Once again it may be this that contributes to a continued pre-occupation with transcription as the primary form of data and the relegation of audio material.

A significant assumption about audio in qualitative research is that it is inherently disclosive. This is the overwhelming reason given by researchers who object to archiving audio files with the UK Data Archive. This is crucial in a field which stresses the sensitivity of the data it collects and frames discussions of its use and suitability for archiving in the context of this. Qualitative researchers are acutely aware of the importance of the relationship they develop with research participants and of the complicated issues of confidentiality they confront (Padgett 1995: 38; Shaw 2008: 400). This relationship is most vividly defined in consent and confidentiality agreements. Such agreements clearly indicate how data is to be collected and used by the researchers. It can be argued that the assumption audio is unusable after transcription is clearly illustrated by these agreements. An in-depth review of collections offered to the UK Data Archive, suggests that in the majority of cases consent agreements define data almost exclusively in terms of text transcription. The option of using or even preserving audio as a potential data resource is seldom detailed. Instead the focus is on draft transcriptions being approved by participants, access to transcripts, or assurances about protecting the confidentiality of transcripts. Whilst these may be good initial data management strategies, it is clear that even at this stage of the research process text has become dominant and, as Finnegan (1992:182) suggests:

This absence of sound recordings from the deliberations of consent agreements re-enforces the view that their only function are to allow transcriptions to be created and that they have no further value once sound has been translated to text. This assumption is further confirmed by a review of agreements received by ESDS Qualidata as part of deposited collections. In the majority of cases no mention of audio is made. When these consent agreements do mention audio, it is most often to assure participants it will be destroyed following transcription. This view is in line with many guides to qualitative research practice which also recommend destruction of recordings (McLellan et al. 2003: 81)

The question then is why is audio assumed to be so unsuitable as research data that it has to be withdrawn, hidden or destroyed? Audio data may perhaps be compared to anthropological field note data which remain hidden whilst informing research. Hammersley and Atkinson (1995: 177) have commented such material,

It may be argued that the prevailing view of audio data is that it is similarly private. It is regarded as a raw and uncontrolled account of the research interaction. An account that begins to be shaped and smoothed out once the process of transcription begins. As Samuel (1998: 389) wrote:

In this sense a process of ownership, re-shaping and validation for the researcher can be said to be part of the translation of data from audio to text. Transcription then does not represent a 'correct super-existent text' but rather a tamed and more controlled version of audio data. This is a form of data that, following the text editing of 'anonymisation', has had any potentially disclosive details removed. Once again the view that collected information only becomes data once it is more 'manageable' is implicit. This text transcription is suitable for research and archiving as, unlike audio, it is now 'once removed' from the original research interaction. A comparison may also be made with the common view of qualitative researchers that quantative data is readily archivable (Broom et al. 2009) precisely because it is removed and more manageable. In these cases the taming is achieved because the data is depersonalised and abstracted after a similar process of editing.

It is often assumed that editing audio in the same way as text is difficult further supporting the view that it cannot be shared. This may partly be the result of past difficulties in working with analogue media or a simple reluctance to move away from text based media. Nevertheless in some specific cases anonymising audio files can be problematic. Whilst editing digital audio is far easier than with analogue audio, it is still a laborious activity requiring a lot of time and planning. In addition it could be argued that removing sections of speech is much more distracting than when similar editing is carried out on text. However audio editing is not an impossibility and it should be examined as an choice within research projects. It can be a practical option if it is done selectively and in a limited way. A recent release from the UK Data Archive contains transcripts and audio from interviews and focus group discussions where selective sound editing was applied. The study 'Discourse of the School Dinners Debate, 2004-2008' (Study Number 6228) addresses recent political controversy in the United Kingdom about nutrition in school meals. The inclusion of audio recordings as part of the archived collection was made possible by the removal of names from the content. The interviews and focus groups were structured so that the occurrence of names were limited and infrequent. It was technically possible to remove each name relatively seamlessly but, in order to maintain data integrity, names were instead replaced with a low tone sound. This made clear where a section had been removed and how much. Overall the editing process was neither time consuming or detrimental to the quality of the recordings.

It might be argued that the problem of having too much material to remove from an audio recording could itself be eliminated by more rigorous planning regarding content and consent. For example, a project may include interview recordings and interview transcripts where each participant begins by confirming their name and address. But if it is agreed that these identifiers will be kept confidential later work will then be needed to remove them from the record. The majority of qualitative collections deposited at the UK Data Archive suggest this situation is common. The solution for most researchers is to edit such details out of transcripts and simply destroy the audio data. A more effective approach would perhaps be to clarify the use of personal identifiers when consent for participation is sought. If it is agreed that such identifiers cannot be used then the next step is to question whether the information needs to be recorded at all. For example, one strategy could be to establish a system of false names or numerical identifiers before interviews take place. These could then be used within the interview removing concerns about archiving transcriptions or audio due to the presence of particular content.

There is also the assumption that audio data is implicitly disclosive simply because interviewee's voices are audible. There may even be a temptation to attempt to mask or disguise a voice on a recording. However such a technique is intrusive and arguably detrimental to the usefulness of audio data. It is also unnecessary. It should not be assumed that sharing audio data is inherently disclosive. First of all, there is no reason to assume that where a respondent has agreed to be recorded they will object to its use without anonymisation (Silverman 2010: 167) or indeed outside of the primary research project. Individual participation may in fact be motivated by a wish to make a visible and public contribution to a subject (Giordano et al. 2007: 264).

Audio Data and Problems of Internal Confidentiality

Nevertheless there may be instances where some level of confidentiality is indeed required. Even in these cases it should not be assumed that sharing audio will in itself breach assurances of confidentiality. Hearing a recording of a voice does not disclose identity except in specific cases. Recognition only occurs when a listener knows, or has prior knowledge of, the person who is speaking. So whilst there is some potential for disclosure from audio data it is in quite particular cases and circumstances, such as studies involving family members. This is not exclusively an issue of working with audio data as such concerns are common to many qualitative research projects. It has led to the distinction between external and internal confidentiality (Tolich 2004). External confidentiality agreements offer general protection against identification. Internal confidentiality agreements offer protection against identification from those who have a closer relationship with participants. This is an example of how careful consideration of consent and confidentiality agreements could possibly offer safeguards of confidentiality that do not rely on the destruction of audio material.

Other projects have attempted to archive text and audio data where internal confidentiality is a major concern. One example is the Timescapes project of the University of Leeds, which was established with advice from the UK Data Archive. This is a longitudinal qualitative project that focuses on how personal and family relationships develop and change over time. It is based on seven related projects that span the life course: two on young lives (siblings and friends, the changing lives of teenagers), three on mid lives (motherhood, fatherhood and work life balance) and two on older lives (grandparenthood and the oldest generation). Potential problems of internal confidentiality are clearly present. Nevertheless, confidentiality agreements have been created for the project which encompasses in depth interviews, oral narratives, photographic data and audio recordings. It is important to note that equal stress is placed upon all the types of data being collected. Whilst there are some issues still to be resolved about the most effective dissemination of some of this material, it is still being gathered on the assumption that all will be preserved and then made available in due course. No data is being collected or created that can only be used within the lifetime of the project or that has to be withheld or destroyed.

A further example of sharing of audio data where internal confidentiality is an issue, can be seen in a recent qualitative study undertaken into the impact of Foot and Mouth Disease. The study 'Health and Social Consequences of the Foot and Mouth Disease Epidemic in North Cumbria, 2001-2003' has been successfully archived at the UK Data Archive (Study Number 5407) and is another illustration of how it is possible to frame consent agreements that preserve both text and audio data for sharing in even the most sensitive situations. Whilst official government inquiries focused on economic and agricultural policy issues, this research was designed to produce evidence about the human health and social consequences of the epidemic. The research was based in the Cumbria area, where economic, social and political life was greatly affected by the outbreak. A standing panel of 54 respondents was recruited as the core of the study which initially used free-text diaries documenting the effects of the disaster and the process of recovery. The panel members produced 3,200 weekly diaries over an 18-month period. The data were supplemented by in depth interviews with each respondent, and focus group discussions. In addition 16 other interviews with stakeholders were conducted. From the beginning of the project, there was a commitment to the production of good quality audio data as well as text-based material. Care was taken in the selection of recording equipment to ensure all audio would be of listenable quality. Within the research team it was felt that a concentration on text would diminish the value of the audio material being created. This realisation informed the methodology of the project and allowed some innovation in the approaches used for communicating analysis and finding. Whilst the bulk of the findings were presented in traditional form--written reports, academic papers and journalistic briefings--a number of special exhibitions were also organised. The intended audience were the local communities themselves and they made use not just of the written findings and extracts from interview transcriptions but also the audio data. Audio montages were used in tandem with text to more fully communicate what had been said and expressed during the project.

The importance of archiving their data and making it available for use by other researchers after the project was completed was also appreciated by the research team. Interview and focus group transcriptions were quickly prepared for archiving but initially there were reservations about archiving text diaries and the bulk of the audio material. The planned exhibitions were to make use of selected sound extracts not the full recordings. In addition a commitment to withholding the identity of participants had been made and this needed to be preserved after data had been archived. Although great care was taken in the creation of the digital audio it was assumed from the start by the research team that it would be problematic and unsuitable for re-use or archiving. Nevertheless, after further discussion and negotiation, it was recognised that the access controls of the UK Data Archive would allow even this data to be archived and made available without breaching confidentiality agreements. The audio extracts that were used for sound montages are currently available as part of the main archived data collection. In addition full and complete copies of the audio recordings are available by special permission of the depositor. This has been supplied to other researchers on a number of occasions; sometimes along with text data and sometimes in isolation as a data resource in itself.

Conclusion

The creation and use of digital audio data within qualitative research can be seen as a useful illustration of some of the challenges of preparing such material for archiving and re-use. The key issue is to recognise the value of digital audio material and to save it from destruction--even after transcription. Even then, when looked at in isolation, such material can still seem to be unsuitable for sharing in any way outside of the primary research project. The problems of confidentiality and disclosure in particular seem insurmountable. However when viewed in relation to wider working practices and options open to researchers, it becomes clear that audio data can very often be preserved and shared. Sometimes this may be a full collection of audio alongside text based material or it might be a specially selected sample. Nevertheless such success shows how digital audio should be valued alongside other forms of data and made to work fully as another facet of qualitative enquiry.

References

Barbour, R. (2008) Introducing Qualitative Research: A Student's Guide to the Craft of Doing Qualitative Research, London, Sage.

Boden, R. Epstein, D. & Latimer, J. (2009) 'Accounting for ethos or programmes of conduct? The brave new world of research ethics committees', The Sociological Review, 57(4), 727-750.

Bishop, L. (2005) 'Protecting respondents and enabling data sharing: Reply to Parry and Mauthner', Sociology, 39, 333-336.

Broom, A. Cheshire, L. & Emmison, M. (2009) 'Qualitative researchers' understandings of their practice and the implications for data archiving and sharing', Sociology 43, 1163-1180.

Brown, D. (2002) 'Using technology in the qualitative research process' Forum: Qualitative Social Research, 3(2), 1-17.

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

Bryman, A. & Hardy, M.A. (2005) The Handbook of Data Analysis, London, Sage.

CESSDA (2009) 'Member Organisations', CESSDA, www.cessda.org/about/ members, 2009.

Cook, G. (2009) 'Discourse of the School Dinners Debate, 2004-2008, SN 6228', UK Data Archive, www.dataarchive.ac.uk/findingData/ snDescription.asp?sn=6228, 2009.

Corti, L. (2000) 'Progress and Problems of Preserving and Providing Access to Qualitative Data for Social Research--The International Picture of an Emerging Culture', Forum: Qualitative Social Research, 1(3), 1-17.

Dicks, B. Mason, B. Coffey, A. and Atkinson, P. (2005) Qualitative Research and Hypermedia: Ethnography for the Digital Age, London, Sage.

ESDS Qualidata (2009) 'Pioneers of Qualitative Research', Economic and Social Data Service, www.esds.ac.uk/qualidata/pioneers/, 2009

Fielding, N. G. & Lee, R. M. (1998) Computer Analysis and Qualitative Research, London, Sage.

Finnegan, R. (1992) Oral Traditions and the Verbal Arts: A Guide to Research Practices, London, Routledge.

Gibbs, G. (2007) Analysing Qualitative Data, London, Sage.

Giordano, J. O'Reilly, M. Taylor, H. & Dogra, N. (2007) 'Confidentiality and autonomy: The challenge(s) of offering research participants a choice of disclosing their identity', Qualitative Health Research, 17, 264-275.

Hammersley, M. & Atkinson, P. (1995) Ethnography: Principles in Practice (2nd edition), London, Routledge.

Kvale, S. & Brinkmann S. (2008) Interviews: Learning the Craft of Qualitative Research Interviewing, London, Sage.

Lapadat, J.C. and Lindsay, A.C. (1999) 'Transcription in research and practice: From standardization of technique to interpretive positionings', Qualitative Inquiry, 5, 64-86.

Lee, R. M. (2004) 'Recording technologies and the interview in sociology, 1920-2000', Sociology, 38, 869-889.

McLellan, E. MacQueen, K.M. & Neidig, J. L. (2003) 'Beyond the qualitative interview: Data preparation and transcription', Field Methods, 15, 63-84.

McLellan-Lemal, E. (2008) 'Chapter Eight: Qualitative Data Management' In Guest, G. and MacQueen, K. M. (eds.) Handbook for Team Based Qualitative Research, Plymouth, Altamira Press.

Maloney, R. S. & Paolisso, M. (2001) 'What can digital audio data do for you?', Field Methods, 13, 88-96.

Mason, J. (2002) Qualitative Researching, London, Sage.

Miles, B. M. & Huberman, A. M. (1994) Qualitative Data Analysis: An Expanded Sourcebook (2nd Edition), London, Sage.

Miller, Jody and Glassner, B. (2004) 'The inside and the outside: Finding realities in interviews', In D. Silverman (ed.) Qualitative Research: Theory, Methods and Practice (2nd Edition), London, Sage.

Mort, M. (2006) Health and Social Consequences of the Foot and Mouth Disease Epidemic in North Cumbria, 2001-2003, SN 5407, UK Data Archive, www.data-archive.ac.uk/findingData/snDescription. asp?sn=5407&key=5407, (2009)

Oliver, D.G. Serovich, J.M. & Mason, T.L. (2005) 'Constraints and opportunities with interview transcription: Towards reflection in qualitative research' Social Forces, 84(2), 1273-1289.

Padgett, D. K. (1995) Qualitative Methods in Social Work Research: Challenges and Rewards, London, Sage.

Pawson, R. and Tilley, N. (1997) Realistic Evaluation, London, Sage.

Richards, L. (2005) Handling Qualitative Data: A Practical Guide, London, Sage.

Samuel, R. (1998) 'Perils of the transcript' In R. Perks and A. Thomson (ed.) The Oral History Reader, London, Routledge

Shaw, I. (2008) 'Ethics and the practice of qualitative research', Qualitative Social Work, 7, 400-414.

Silverman, D. (2001) Interpreting Qualitative Data: Methods for Analysing Talk, Text and Interaction (2nd edition), London, Sage.

Silverman, D. (2010) Doing Qualitative Research (3rd edition), London, Sage.

Southall, J. (2009) 'Audio Visual Equipment, Economic and Social Data Service, www.esds.ac.uk/qualidata/create/audiovisual.asp, 2009.

Speer, S.A. & Hutchby, I. (2003) 'From ethics to analytics: Aspects of participants' orientations to the presence and relevance of recording devices', Sociology, 37 (2), 315-337.

Tilley, S.A. (2003) 'Challenging research practices: Turning a critical lens on the work of transcription', Qualitative Inquiry, 9 (5), 750-773.

Timescapes (2009) 'Timescapes: An ESRC Qualitative Longitudinal Study', University of Leeds, www.timescapes.leeds.ac.uk/, (2009)

Tolich, M. (2004) 'Internal Confidentiality: When Confidentiality Assurances Fail Relational Informants', Qualitative Sociology, 27 (1), 101-106.

UK Data Archive (2009) 'Ordering Data Restrictions', UK Data Archive, www. data-archive.ac.uk/orderingData/restrictions.asp, 2009.

UK Data Archive, (2009) Managing and Sharing Data--a Best Practice Guide for Researchers, University of Essex.
Within the qualitative tradition there is a sense in which the tape
   recorder (sic) has come to be seen less as a device for recording
   sound and more as a means of producing text.


Interviews are one of the most commonly recognised forms of
   qualitative research method. Perhaps for this reason, it is not
   uncommon for a researcher to assume that their study will
   involve such interviews.


Despite its centrality in qualitative data collection, transcription
   practices remain superficially examined. It is not uncommon for
   transcription to be presented as a behind the scenes aspect of
   data management rather than as an object of study in its own
   right.


As with just about everything in conducting social research,
   there is a cost ... in that the use of a tape (sic) recorder may
   distract respondents, who become self-conscious or alarmed at
   the prospect of their words being preserved.


The inadequate documentation and monitoring of data
   activities may threaten data integrity. In addition, inadequate
   data tracking practices may hamper analysis and increase the
   likelihood of research pandemonium.


... there is the idea of a 'correct' super-existent text, somehow
   existing 'out-there' in its own right, of which the (original)
   spoken words are merely the secondary reflection--a model
   lying behind many transcriptions.


... have often been regarded as highly personal and private
   documents. Although field notes are the basis of public domain
   scholarship, the authors have rarely shared them with other
   scholars.


Questioning itself, however sympathetic, produces its own forced
   sequences, and the editing of a transcript is almost bound to
   reinforce this. The writer has his own purposes, and these may
   be only coincidentally those of his informant.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.