Archiving qualitative data in Australia: an introduction.
Article Type: Report
Subject: Data libraries (Research)
Qualitative research (Analysis)
Authors: Cheshire, Lynda
Broom, Alex
Emmison, Michael
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: Event Code: 310 Science & research
Geographic: Geographic Scope: Australia Geographic Code: 8AUST Australia
Accession Number: 222315596
Full Text: Introduction

The prospect of the digital archiving of qualitative data for re-use and analysis by other researchers is now a distinct reality in Australia following the development of similar facilities in the UK, the US and Finland. While these archives are now well-established and have become reasonably well-accepted among the scholarly community, their development has not been uncontroversial and has stimulated much debate about the desirability and feasibility of archiving qualitative data on ethical, epistemological and ideological grounds (Hammersley, 1997; Mauthner, Parry and Backett-Milburne, 1998; Moore, 2007; Parry and Mauthner, 2004; 2005). Such debates have done much to improve the practice of data archiving by making it more sensitive to the distinct characteristics of qualitative data and to the particular needs and concerns of qualitative researchers. At the same time, the very real prospect of data archiving has induced researchers to take stock of their own practices and to re-examine deeply held views about the nature of qualitative research, the way it is conducted, and the claims to truth and knowledge that can be generated from it (see Broom, Cheshire and Emmison, 2009). In this sense, the development of an archive--regardless of its controversiality--adds much to the debate, and therefore the advancement, of the practice of qualitative research.

Here in Australia, similar debates are beginning to emerge following the expansion of the Australian Social Science Data Archive (ASSDA) at the ANU in order to provide a storage and retrieval facility for a broader range of research data. ASSDA has been archiving quantitative survey and demographic data since the early 1980s and currently holds over 2000 datasets which are available for researchers to use, subject to a range of different access conditions. With no equivalent archiving facility currently in operation for qualitative data, one of ASSDA's key priorities has been to establish the Australian Qualitative Archive (AQuA), which will be housed at The University of Queensland as part of a distributed ASSDA network. The development of AQuA has been accompanied by a process of consultation with the qualitative research community, involving a series of focus groups in 2007 to 2009, various workshops at meetings of disciplinary associations and, more recently, a discussion paper outlining the key issues under consideration (Cheshire, 2009). The feedback we have elicited so far indicates that the Australian qualitative community is equally reticent about data archiving as its overseas counterparts have been, and that the concerns over depositing and sharing data are very similar to those already articulated elsewhere.

The papers in this special issue contribute to the debates about qualitative data archiving in two key ways. First, they represent one of the first opportunities among Australian researchers to comprehensively engage with the issues arising from qualitative data archiving and to begin this dialogue in the Australian context. In addition, however, they also contribute to the international debate on data archiving rather than simply rehearsing what has already said before. The inclusion of papers from British scholars in this special issue--Libby Bishop, Natasha Mauthner and Odette Parry, and John Southall--all of whom have worked or written extensively on qualitative data archiving means that this collection also represents the next stage of thinking on these issues more broadly. As the papers indicate, this thinking is becoming increasingly sophisticated and has moved well beyond the polarising debates between qualitative archive believers and skeptics to consider the philosophical and scientific terms on which we engage with these issues (Mauthner and Parry, this issue). It is important to note that virtually all the papers are underscored--explicitly or implicitly--with a model of qualitative research practice that emphasises its interpretive character and the centrality of the researcher-researched relationship. Central to this model is the use of interviewing--either face to face or in focus groups--as the key method of obtaining data, largely in the form of respondents's personal experiences and narratives. Although widely accepted as the 'gold-standard' for qualitative research, the use of this method must be seen as only one of the approaches that are available for qualitative inquiry. An increasing number of researchers are coming to question, as Silverman has put it 'the fashionable identification of qualitative method with an analysis of how people "see things", preferring to focus instead on how people "do things"' (Silverman, 1998: 3). Whereas once this would have entailed methods such as participant observation, data for these inquiries are now increasingly taking the form of audio and video recordings of naturally occurring behaviour. Here the analytic focus is less on the discovery of the 'meaning' of the behaviour than with the explication of the methods and socially organized practices which constitute particular settings (e.g. Buscher, 2005).

In the remainder of this introduction we provide a background to qualitative data archiving in Australia while considering international and methodological precedents that provide models to follow. We also outline some of the arguments put forward in favour of data archiving before outlining, briefly, the main points of contention about both the feasibility and the philosophy of data sharing. These issues have already been debated extensively and can be categorised into four key areas of concern: first, practical considerations, particularly those relating to research ethics; second, epistemological misgivings over whether data archiving is consistent with the interpretivist framework upon which qualitative research generally rests; third the culture of qualitative research which, it has been suggested, has not been conducive to data sharing; and finally ideological opposition arising from a concern that archiving forms part of a bureaucratic agenda to erode academic freedom. In the final section of this introduction, we briefly outline the key points raised in each of the five papers.

Qualitative archiving: international and historical precedents.

To some extent, the idea of archiving and re-using qualitative data is still a relatively new one, which partly explains why many researchers have responded to it with caution. This contrasts with the way data archiving has become a commonly accepted practice among the quantitative research community with researchers frequently interrogating large, national survey datasets to address new research questions. Institutionally, data archives have also had a preference for working with machine-readable statistical data and many international archives still lack any comprehensive facilities for dealing with qualitative material. The principal exception to this is the UK Data Archive's Qualidata facility which was established in the mid 1990s, initially as a recovery exercise to prevent the loss of data from classic sociological studies such as John Goldthorpe et al's The Affluent Worker (1967), Peter Townsend's studies on Poverty in the UK (1979) and Stan Cohen's work on Folk Devils and Moral Panics (1972). Located at The University of Essex, Qualidata also acquires data from current projects, which it typically obtains through an agreement with the Economic and Social Research Council that all grant holders are required to offer their data to the UK Data Archive. The archive also provides a range of other services to researchers depositing and accessing data, including support with data management plans, the provision of teaching resources, methods training, and support with research ethics. While other European countries have attempted to follow the Qualidata model, the only other dedicated national qualitative archive is located in Finland at Tampere University.

The slow take-up of qualitative archiving might be viewed as surprising when one considers that there are precedents for qualitative archiving in the social sciences, both internationally and within Australia. For example, the Mass Observation Project commenced in the UK in 1937 and continued through to the 1950s, documenting the everyday lives of ordinary British people. After moving to the University of Sussex, the project was revived in the early 1980s where it now collects and archives data from a team of observers and a panel of volunteer writers on topics of contemporary interest such as marriage, diet and election outcomes. The collection contains a variety of data forms, including diary entries, survey responses, written responses to open-ended questions, life-story documents and observational records of conversations and everyday interactions (see website at http//www.massobs.org.uk/index.htm). Similarly in the US, The Murray Research Centre, located at Harvard University, has been collecting a broad range of quantitative and qualitative social science data since 1976 on human development and social change (Corti and Thompson, 2004). Here in Australia, the Pacific and Regional Archive for Digital Sources in Endangered Cultures (PARADISEC) offers a facility for digital conservation and access of languages from the Pacific region that are likely to be lost in coming decades. The archive contains field recordings from the 1950s and 1960s, as well as audio-visual recordings of songs, languages, dance and cultural rituals. Other, smaller archives are also in operation, holding data on specific fields of research or for particular methodological approaches. These include a collection of narrative interviews on experiences of health and illness generated through the DIPEx project (see Seale and Charteris-Black, 2008), the oral history archive of 'German Memory' in Hagen (Corti and Thompson, 2004) and the Timescapes archive at the University of Leeds which will hold data collected through the ESRC-funded 'Timescapes' qualitative longitudinal study on personal and family relationships.

Why archive data?

Most of the archives described above have been initiated by individual or team researchers in response to concerns about the preservation and future use of research data. From this perspective, the rationale behind data archiving is fairly compelling, especially when it involves the provision of a safe and secure site for preserving research materials in the long term. A great deal of research data is typically stored on computers, in filing cabinets or even in the homes of researchers, which increases the risk of data loss. A national archive will store data indefinitely in a safe place, long after the original study has been completed. This is especially important for datasets that have national or historical significance, and which need to be preserved for the future to minimise loss. As Corti and Thompson (2004) report, this was the primary motivation for the establishment of Qualidata in the UK after a pilot study commissioned by the ESRC in the early 1990s found that 90 percent of all qualitative research data was either lost or at risk of loss, including data from Margaret Stacey's community study of Bambury (1974) and Rex and Moore's work on race and conflict in Sparkbrook (1967) which had been destroyed upon the retirement of the researchers. Australia has its own legacy of pioneering studies which also need to be preserved for posterity.

But the arguments in favour of data archiving go beyond mere preservation and include other benefits that arise from making data available for other researchers to access and re-use. In particular, data sharing will be of considerable benefit to researchers undertaking historical research in the future and will allow for comparative cross national or case study research. Further, archived material is an excellent resource for teaching qualitative research methods across a range of disciplines. Teaching staff will be able to access various qualitative datasets to provide examples of how different researchers approach their research; the sorts of questions they ask of the world around them; and the way they operationalise their research through their choice of methods. Students learn a great deal about qualitative research when they are given the opportunity to handle data and to practice writing it up. A national qualitative archive will provide such resources, complete with important related data, such as research questions, sampling frameworks and interview schedules. Deposited data could also be made available to Honours and/or undergraduate students for research projects, especially in cases when time, resources and ethical considerations prohibit students from generating their own material.

Another point is that qualitative data are expensive to collect and the principal investigator(s) often do not have the time or money to analyse all aspects of the data they have obtained. When data are made available through a data archive, maximum utilisation of the data is possible as researchers are able to access deposited datasets rather than continually generating new ones. As well as avoiding duplication of research efforts, this increases the value and accountability of publicly funded research: something funding agencies are becoming increasingly aware of. At the same time, data archiving enhances the value of investment made by research participants into the research process. It is widely accepted, for example, that researchers have an ethical obligation to treat participants' stories with due care and respect. While most researchers would certainly agree with this principle, the conventional approach in most forms of qualitative research is to de-identify data and destroy them after analysis. Such practices are often underpinned by a medical model of research ethics that continues to guide many university ethical review processes. An alternative approach would argue that safeguarding people's stories via an archive and allowing others to access them, albeit under restricted conditions, is a more appropriate way of taking care of them. This is particularly, although not exclusively, the case for Indigenous populations who remain owners of knowledge, rendering the researcher a guardian of that knowledge for current and future generations. In addition, research is often intrusive for participants, especially in cases where the research covers sensitive or personal issues. Some researchers are beginning to encounter difficulty in accessing certain populations, such as indigenous people and schools, who feel they have been 'over-researched' and are becoming less willing to tolerate further scrutiny for research activities, especially if they discern a duplication of topics and questions. Sharing data means the need to intrude into people's lives to collect more data is reduced and will help ensure full use is made of data already generated.

Finally, is also important to the scientific method that social science data generated from research projects, along with the methods and analytical processes adopted, are open to scrutiny by others. Many qualitative researchers working from an interpretivist paradigm reject the idea that the quality of qualitative research can be assessed by traditional criteria, such as reliability, replicability and validity, because they are based on an assumption that the social world can be readily captured, measured and recorded as objective truth by an independent observer (Guba and Lincoln, 1991). This is not to say, of course, that quality and analytic rigor are of no concern to qualitative research; simply that other criteria are required. These may require audiences to judge the success of qualitative research according to its methodological awareness (Seale, 1999), its transparency in the research process (Rubin and Rubin, 1995) and the consistency and credibility of its arguments (Guba and Lincoln, 1991). Archiving data for others to use will enhance this process. Access to an existing bank of data also means that researchers may be able to examine a larger number of cases in their research than would typically be feasible if relying on the generation of new data (Hammersley, 1997). As Mason (2007) points out, the political economy of social research is such that qualitative research has traditionally been regarded by funding agencies and government departments as the poor relation of quantitative research and that the way to change these perceptions is by advancing new ideas and innovations that push the qualitative research agenda forward. A national archive that facilitates data sharing and promotes a more collaborative approach to qualitative research, she suggests, will likely contribute towards this new agenda.

Important considerations in archiving qualitative data

Within current debates around qualitative data archiving, few researchers appear to hold any opposition to the idea of data archiving per se and, indeed, frequently acknowledge the utility of having a safe storage facility for their data, and of accessing others' data, especially for teaching purposes and historical or comparative analyses. Instead, their concerns tend to revolve around questions of how the archive will actually operate in a way that is sensitive to the kind of research they do and how it will deal with the special nature of qualitative data. In our focus group consultations, for example, most researchers felt entirely comfortable with the idea of an archive for quantitative survey-based data but viewed qualitative data as something quite distinct, both in epistemological terms where data were seen as embedded in particular sets of relationships and contexts that would be lost if the data were archived, but also in more practical ways which relate to research ethics and issues of trust between the researcher and the research participants (Broom et al, 2009). More broadly, it has also been suggested - both in this issue (see for example Travers, this issue) and elsewhere (Corti and Thompson, 2004; Fielding, 2004; Mason, 2007) that there are cultural and ideological reasons for why qualitative researchers are so reluctant to fully embrace data archiving. These arguments have all been outlined in detail previously but are summarised below, briefly.

Practical and ethical dilemmas

The most immediate questions raised about qualitative data archiving tend to be practical and relate to matters such as how much additional work will archiving create for the researcher, who will access the data and for what purpose, how will the participants feel about having their data archived, will this reduce the likelihood of them agreeing to participate in the research, and how assurances of confidentiality will be maintained when raw data, rather than simply published outputs, are shared with the broader research community. While many of these considerations can, and have been, addressed by the establishment of suitable policies and protocols for restricting data access, preparing data for deposition and obtaining informed consent, the ethical dilemmas of archiving data remain rather tricky because of the in-depth and contextually-rich nature of the data generated through qualitative research. This, as others have pointed out (see Corti, Day and Backhouse, 2000; Parry and Mauthner, 2004) means that where survey data can be readily de-identified by stripping it of identifying features and aggregating responses, qualitative data are imbued with the stories and life histories of individual research participants that render them readily identifiable even when direct identifiers, such as names, have been removed. Such considerations are not unique to archived data, since all qualitative researchers at some point need to discuss with participants whether and how their identities will be protected in the research account. However, this tends to be limited to the publications and other outputs arising from the research and does not need to be applied so rigidly to the treatment of raw data. Moreover, researchers readily acknowledge that these techniques for maintaining confidentiality are imperfect even when they are applied to research outputs only, particularly in cases where participants can be identified by the specialised or public positions they hold, or if their identities are obvious to other 'insiders' simply by virtue of the stories they tell (Tolich, 2004).

The dilemma from this lies in attempting to balance, on the one hand, the rights of participants to have their identities protected by removing all potentially identifying information from an interview transcript and, on the other, the need for subsequent researchers to have complete and unobstructed access to as much of the data as possible in order to analyse that data in their full context. More obviously, however, is the question of how one might de-identify audio or video data in ways that are neither time consuming nor costly. This dilemma for researchers is compounded by two additional considerations relating to the ethics of data archiving, specifically the requirement to obtained full informed consent, not only to collect the data but also to make it available for sharing. The first is that some groups or populations may be less likely to agree to participate in a research project if they know their data will be accessed by others. These include vulnerable groups; those engaging in illegal or anti-social activities such as drug-taking; those participating in research on particularly sensitive topics; elite groups such as politicians for whom confidentiality may be an issue; and organisations such as schools and hospitals where researcher access is already becoming difficult and subject to complex negotiations with gatekeepers. A second point is that informed consent to archive may also be more difficult to obtain because researchers cannot give assurances to participants about how that data will be used in the future, thus leaving open the possibility of subsequent researchers generating conclusions that run counter to the interests of the research participants. In its National Statement on Ethical Conduct in Human Research (2007) the National Health and Medical Research Council (NHMRC) refers to this open type of informed consent as extended (i.e. in the same general area of research) or unspecified (i.e. for the use of that data in any research project) and suggests that informed consent may be obtained from participants providing the full implications of granting unspecified consent are explained to them. In many ways, this issue is not limited to the collection of data for archiving as Bishop (this issue) points out. Indeed, she suggests that informed consent is always partial since the very nature of research is such that we do not know what our research findings are likely to be. What is important is to alert participants to this fact and to conduct research in a reflexive and ethical manner that allows for unanticipated findings from research without causing undue harm and embarrassment to research participants.

The epistemological debate: the problem of context

The second major concern over qualitative data archiving is more difficult to address since it gets at the very heart of what constitutes qualitative research and therefore cannot readily be addressed by the implementation of archiving policies. Such concerns stem from the philosophical roots of qualitative research and the suggestion that data derived from interpretivist approaches typically involve subjectivities and epistemologies that do not lend themselves to data archiving (Hammersley, 1997; Parry and Mauthner, 2004). There are three key elements to this argument. First, that qualitative research is based on a constructivist ontology that views the social world, and the research process itself, as socially constructed, and hence historically and situationally contingent. As Hammersley argues, depositing data into an archive ready for others to pick up involves stripping that data of any prior meaning or interpretation and indeed reverts back to foundationalist assumptions that knowledge and facts are pre-existing and lying around as common currency to be collected, rather than being co-created and value driven. Leading on from this is a second point about context in qualitative research since if meaning and knowledge are constructed under particular situational and historical conditions, researchers need to have full awareness of those contextual factors if they are to make sense of the data. This is seen to apply particularly to a semi-structured interview setting in which what was said can only be interpreted adequately if researchers have an understanding of how it was said and what was implied. The final and related component of this argument is that 'context' refers not only to the broader, socio-historic or situational conditions in which meaning-making occurs, but equally to the interactional and relational conditions of the research encounter between the researcher and the research participant. It is widely accepted in qualitative research that the researcher is integral to the research process and that his or her biography and subjectivity inevitably inform the research process and data analysis. Such influences should not be eradicated but form part of a reflexive epistemology for qualitative researchers which, for Parry and Mauthner (2004), is incompatible with attempts to archive data.

In an attempt to prevent these concerns from stalling the archiving agenda, one solution implemented by Qualidata has been to ensure that data depositors submit their raw data with an extensive set of accompanying documentation, such as research proposals, ethical clearance applications, published papers, field notes and other such material, to provide subsequent researchers with the necessary contextual information for informed analysis. Yet this has done little to ease the epistemological concerns of researchers such as Mauthner, Parry and Backett-Milburne (1998) who argue that background material also need to be treated as data in their own right, since they too are the product of a reflexive relationship between researcher and researched and therefore cannot be used merely to fill the interpretive gap. Others such as Fielding (2004), however, point out that the problem of context in data archiving is a practical rather than epistemological issue because it is something that confounds even the primary researcher who often has to make sense of data that are contingent and incomplete: perhaps the researcher did not follow up on something, or certain data were unavailable or even that the audio recording failed at a crucial moment. Thus Fielding (2004: 99) suggests that:

Moore (2007: para 2.3) also challenges the distinction between primary and secondary data, arguing that the secondary users working from an interpretivist perspective do not treat data as pre-existing but reflexively engage with it on their own terms, reconstructing and re-contextualising that data through new forms of analysis. Moreover, as we discuss elsewhere (see Broom et al, 2009), the partiality and contingency of qualitative data means that any claims to epistemic transcendence by the original researcher as one who 'knows best' are difficult to justify within an interpretivist paradigm. Not only is the reputation of qualitative research likely to be undermined if it is seen to generate data that are so idiosyncratic and particularistic that no one else is capable of making sense of them, it also suggests that the evidence on which a particular interpretation is based is tacit rather than explicit and requires the researcher to ask of the reader 'trust me, I was there'. Seale (2010) makes a similar point, suggesting that this reliance upon intuition, experience and memory in meaning making presupposes that those important contextual clues cannot be conveyed to others in writing. He also reminds us that the significance of context and reflexivity is limited to particular forms of data that involve face-to-face contact between the researcher and the researched (such as interviewing) and that there are many other qualitative methodological approaches that do not rely upon the presence of the researcher to collect the data and for which concerns over subjectivity are less important (see also Broom et al, 2009).

While this is certainly true, it does not lessen the need for an ongoing reflexivity among researchers who are involved in the co-creation of data through their interactions with research participants. The concern for Mauthner and Parry (this issue) is precisely that while users of secondary data may be recontextualising that data through their own biographical and analytical terms of reference, in doing so, they often ignore or eradicate the primary context in which those data were first generated. In other words, they are recontextualising (what is essentially viewed as) decontextualised data. One solution to this is to encourage researchers to convey more clearly the effects of researcher reflexivity in both primary and secondary data analysis, perhaps through diary keeping as some have suggested (Kelder, 2005). Nevertheless, with increasing demands on researchers' time and resources, the extra work required to more clearly document one's research for the benefit of others is likely to make archiving less, rather than more, popular.

Cultural impediments: a tradition of working alone

A third potential barrier to qualitative data archiving is the frequent claim that a culture of data sharing is lacking among the qualitative research community, with qualitative researchers often describing their work as a solitary, personal endeavour (Broom et al, 2009; Corti and Thompson, 2004; Fielding, 2004). This is not to suggest that qualitative researchers are anti-social beings who lack the skills for collaborative activity, but more that the institutional structures of research are such that while data sharing has become essential and accepted among quantitative researchers, their qualitative counterparts have not been exposed to the same kinds of pressures and opportunities. In survey data, the pursuit of large, nationally representative and longitudinal data has necessitated the creation of datasets such as the Household Income and Labour Dynamics in Australia (HILDA) at the University of Melbourne and the Negotiating the Lifecourse study at the ANU which are readily available to secondary users via the ASSDA website. This, in turn, has rendered secondary analysis of existing data a widely accepted methodology among quantitative researchers who regularly submit funding applications to the ARC and NHMRC for projects that involve the interrogation of existing datasets rather than the creation of new ones. Such practices are far more rare among qualitative researchers who typically rely upon primary data collection to address their research questions.

However, there are signals of change in this research culture, partly with the increased use of a team approach to research involving the use of colleagues, PhD students and research assistants to share the tasks of project design, data collection and analysis. In such cases, researchers may well find themselves working with data that were collected by others, and therefore doing so without the requisite contextual or reflexive resources that come from 'being present'. Nevertheless, as we discuss elsewhere (Broom et al, 2009), the removal of the researcher from the data collection process is often viewed as less problematic with team research given that collaborators are better able to convey important contextual information to one another through project meetings. In the case of a data archive, however, not only are the relationships between researcher and researched absent; so too are those between data collector and data analyst. In some cases, secondary users have been able to communicate with the original researcher for further insights into the data generation process (Kelder, 2005) but this will not always be possible in the long-term. Even so, there are other new trends in qualitative research that suggest a changing attitude towards data sharing and archiving. In the UK, for example, where Qualidata has been operating for over 15 years, the initial ambivalence of researchers towards archiving and re-using data appears to be giving way to a gradual acceptance. This is particularly evident in the growth of data re-use rather than merely deposition, with over 100 research and teaching projects listed on the Qualidata website which have been conducted with the secondary analysis of archived data. While there have been few published works based on secondary data analysis, this is also beginning to change as secondary analysis of qualitative data becomes an increasingly legitimate methodology for researchers, funding agencies and academic journals (see Fielding and Fielding, 2000; Kelder, 2005; and Seale and Charteris-Black, 2008 as examples).

Ideological concerns: the erosion of academic freedom

A final explanation for the apparent reluctance of qualitative researchers to embrace the prospect of data archiving has been raised by Travers (this issue) who suggests that there are fundamental ideological objections to the practice based on concerns that academic freedom has been steadily eroded and that judgments about what constitutes good research are increasingly being underpinned by positivist or empiricist criteria that value impacts, outputs and policy-relevance over critical or theoretical engagement. From this perspective, data archiving is seen as yet another step in this process by rendering social research more publicly accountable, cost-effective and regulated. Such charges have previously been levelled at Qualidata in the UK, which some see as having been established by the Economic Social and Research Council as part of its drive for cost-effectiveness, value for money and increased accountability of 'public money' (see Moore, 2007). Here in Australia, there are similar concerns over the potential loss of academic freedom, particularly in terms of research funding, following the controversial vetoing by the former Federal Minister for Education, Brendan Nelson, of ten ARC grants which had previously been approved by the relevant ARC College of Experts. While researchers must, and do, accept some need for public accountability of research funding, there is a strong sense that archiving must be driven by intellectual and methodological imperatives. There is also a risk that AQuA may come to attract similar criticisms of managerialism although it is anticipated that the active involvement and leadership of social science researchers in AQuA and ASSDA more broadly will help reduce them.

Introduction to the papers

In moving the debate on qualitative data archiving forward, we have drawn together a range of stakeholders from Australia and the UK for this special issue who each have something new to say about the possibilities and barriers to the further development of qualitative data archiving and secondary analysis in Australia and internationally. In selecting these papers we sought to provide a global and interdisciplinary perspective on the state of play in qualitative archiving. Our approach as guest editors was to allow the authors to maintain their own conceptual lens and philosophical positionings, rather than setting a particular agenda for the issue. As such, this collection does not represent a set of papers that are, ultimately, in complete agreement. Rather, each has their own particular viewpoint on the value of current qualitative archiving initiatives, technological developments and disciplinary innovations. As such, they should be read as individual pieces situated within an ongoing debate. We hope that this issue will contribute to the emerging constructive conversations in Australia around the potential (and character) of a national qualitative archive. The contributions from British authors provide the opportunity to learn from previous innovations in the UK context, and the other Australian-based papers may begin to sensitise us to the concerns of Australian social scientists and some of the idiosyncrasies of the Australian academic and research context.

Drawing from her experience in the British context, Bishop engages in the debate around ethical practice and the contestability of the variety of ethical questions posed in relation to the archiving of qualitative data. Of particular interest here is the idea of a 'deontological ethical stance', encompassing the idea of duties as well as rights. Her paper seeks to unpack existing (and perhaps questionable) beliefs around the ethics of archiving, and importantly, offers a more nuanced and multifaceted view of the problematic of qualitative archiving; it furthers the debate by avoiding the unfortunately common process of merely objecting to contemporary scholarly objections. The essence of Bishop's argument is that the various critics of re-using qualitative data have maintained a very narrow view of 'what is ethics', focusing largely only certain agents' rights (i.e. participants, not other scholars and the community) and not 'duties' (i.e. what one should do, rather than what one should be protected from). These 'duties' include such things as avoiding duplication, and ensuring openness and transparency, and the community and public interest--arguments central to the move toward qualitative archiving. This paper also provides insightful and timely critique of the more practical 'barriers' to qualitative archiving including those of consent, anonymity, the researcher/participant relationship, and participants' influence within (and over) data analysis. Ultimately, the paper provides a convincing and refreshing broad overview of how a regressive and simplistic view of ethical practice will limit the development of qualitative archiving internationally.

Max Travers takes a rather different angle, engaging with the broad idea that archiving 'promotes and institutionalises a narrow empiricist version of qualitative research'. While in many ways 'sitting on the fence' regarding the value of archiving qualitative data in Australia, Travers provides an interesting and reflective paper that warns against repeating the perceived mistakes of previous proponents of qualitative archiving in other countries. Engaging with Thompson's claim regarding the apparent lack of enthusiasm for archiving in the qualitative sociological community ('a strange silence'), Travers posits that ideological tension and even resentment may underpin and strengthen ethical and practical objections to qualitative archiving. In actuality, he suggests no such silence exists, rather, qualitative researchers vary hugely in their willingness to engage with data archiving and secondary analysis. In the first part of the paper, he positions the archiving agenda beside other contemporary initiatives to assess, regulate and bureaucratise academic research (research assessment exercises and funding bodies' requirements), arguing that a similar 'empiricist agenda' runs through these processes. He then turns to what he sees as the 'flaw' in arguments around interpretive multiplicity and interpretive analysis, pointing out that theoretical disagreement in sociology presents various challenges to the depositing of (non-contextualised) data. As Travers points out, the idea of an archive containing 'objective data' that can be re-analysed and re-interpreted, sits uncomfortably with many sociologists given that the data are in fact a product of diverse interpretations. Finally, the case of conversation analysis is used to bolster the idea of the problematic of intra-disciplinary diversity (in tradition, interpretation and current forms of practice within the social sciences) in terms of working towards effective archiving practices. As part of this discussion, he reports on his own exploration of the Qualidata site and archived material, arguing that the practical processes of archiving, in the British context, lend themselves to a more positivist view of data--something that should be avoided in the Australian context.

In their contribution Natasha Mauthner and Odette Parry take a critical approach to the deployment of the notion of 'opponents' of qualitative data archiving, suggesting instead that the archiving agenda has been driven by foundationalist principles and a monolithic conception of 'what is qualitative'. Engaging with the rationales of those seeking to 'facilitate' researchers with the 'practical' problems of archiving data, they re-focus the debate on the plural epistemologies evident in qualitative research and the multiplicity of stances evident in the interpretive project more broadly. They argue that the so-called 'opponents' are largely not against data preservation and sharing, but rather sceptical toward philosophical assumptions underlying the current archiving agenda--much of which is driven by empiricist orientations of quantitative researchers. This agenda, it is argued, is imposing normative expectations about the philosophical orientation of the qualitative research community, and threatens to work against epistemological diversity that remains core to the character of qualitative research. This paper does not discount the potential of archiving nor the value of sharing data, rather, the authors seek to revisit the clear diversity of philosophical traditions within the qualitative research community and the need to retain different positionings and standpoints within the deposition (and secondary analysis) of qualitative data.

In the next paper Peter Waiters develops a very different position from that of Travers, arguing that re-interpreting this data is consistent with interpretive and constructivist epistemologies. He argues that debates about re-use and re-interpretation of data should be based on professional skill rather than epistemological barriers. Moreover, he suggests that any loss of privileged interpretive insights of the 'original' researcher would be more than compensated by the ability of future researchers (he assumed some time delay in reuse) to challenge past and present assumptions and reflect upon the process of social change. Waiters takes a more relativist ontological stance, arguing that neither interpretation is inherently weaker or stronger than the other--the secondary analysis requires the earlier one for its coherence. Furthermore, he suggests that researchers who re-interpret can argue for a more convincing version (of the 'truth'), rather than a 'more correct' interpretation, based on access to wider contextual data, the 'wisdom of hindsight' and even more sophisticated method. The debate over reflexivity and qualitative data forms an important part of this paper, with Waiters suggesting that by refusing the opportunities of secondary analysis researchers are reifying one version of reality. If the reflexive project is true to its ideals, he argues, then the full reflexive revelations of a research project can only add to the value of this data in the future. The issue, he argues, is thus not epistemological but rather practical, with researchers needing to systematically document contextual pointers (e.g. original research questions, funding sources, research participants, stage management and so on), for the future analyst.

In the final paper, John Southall draws on his experience as a senior archivist at ESDS Qualidata (UK) to reflect on some of the practicalities and difficulties in negotiating the deposition of data in the UK archive, with a focus on researcher concerns around audio recordings. Focusing on both emphasising key barriers and debunking myths around the problematic of depositing audio data in an archive, he highlights the tendency of actual audio recordings to be relegated once text based material is produced (and thus undervalued). Given the issues around disjunctions between actual interview interaction and transcribed interaction/conversation, this paper is a pertinent contribution to the debate around the form and structure of qualitative data deposition. Strategies for getting past the issue of confidentiality and ensuring anonymity, even when depositing audio recordings, are examined, and it is argued that audio data can (and should) be preserved and shared alongside text transcriptions.

References

Broom, A., Cheshire, L. & Emmison, M. (2009) 'Qualitative researchers' understandings of their practice and the implications for data archiving and sharing', Sociology, 43(6), 1163-80.

Buscher, M. (2005) Social life under the microscope? Sociological Research Online, vol 10, 1 http://www.socresearch.org.uk/10/1/buscher.html

Cheshire, L. (2009) Archiving Qualitative Data: Prospects and Challenges of Data Preservation and Sharing among Australian Qualitative Researchers, Discussion Paper, Canberra, Australian Social Science Data Archive.

Cohen, S. (1972) Folk Devils and Moral Panics: The Creation of the Mods and Rockers, London, MacGibbon and Kee.

Corti, L., Day, A. & Backhouse, G. (2000) 'Confidentiality and informed consent: issues for consideration in the preservation of and provision of access to qualitative data archives', Forum: Qualitative Social Research, 1(3), [Online], http://qualitative-research.net/fqs/fqs-eng.htm.

Corti, L. & Thompson, P. (2004) 'Secondary analysis of archived data', in C. Seale, G. Gobo, J. Gubrium and D. Silverman (Eds) Qualitative Research Practice, London, Sage, 327-43.

Fielding, N. (2004) 'Getting the most from archived qualitative data: epistemological, practical and professional obstacles', International Journal of Social Research Methodology, 7(1), 97-104.

Fielding, N. & Fielding, J. (2000) 'Resistance and adaptation to criminal identity: using secondary analysis to evaluate classical studies of crime and deviance', Sociology, 34(4), 671-89.

Goldthorpe, J. Et Al (1969) The Affluent Worker in the Class Structure, London, Cambridge University Press.

Guba, E. & Lincoln, Y. (1991) 'What is the constructivist paradigm?' in D. Anderson and B. Biddle (Eds) Knowledge for Policy: Improving Education Through Research London, Falmer Press, 158-70.

Hammersley, M. (1997) 'Qualitative data archiving: some reflections on its prospects and problems', Sociology, 31 (1), 131-42.

Kelder, J. (2005) 'Using someone else's data: problems, pragmatics and provisions', Forum: Qualitative Social Research, 6(1) [Online] http://www/ qualitative-research.net/fqs/

Mason, J. (2007) 'Re-using qualitative data: on the merits of an investigative epistemology', Sociological Research Online, 12(3), [Online] http://www. socresonlinr.org.uk/12/3/3.htm

Mauthner, N., Parry, O. & Backett-Milburne, K. (1998) 'The data are out there, or are they? Implications for archiving and revisiting qualitative data', Sociology, 32(4), 733-45.

Moore, N. (2007) '(Re)using qualitative data', Sociological Research Online, 12(3), [Online] http://socresonline.org.uk/12/3/1.htm.

National Health and Medical Research Council (2007) National Statement on Ethical Conduct in Human Research, Canberra, Government of Australia.

Parry, O. & Mauthner, N. (2004) 'Whose data are they anyway? Practical, legal and ethical issues in archiving qualitative research data', Sociology, 38(1), 139-52.

Parry, O. & Mauthner, N. (2005) 'Back to basics: who reuses qualitative data and why?', Sociology, 39(2), 337-42.

Rex, J. & Moore, J. (1967) Race, Community and Conflict, Oxford, Oxford University Press.

Rubin, H. and Rubin, H. (1995) Qualitative Interviewing: The Art of Hearing Data, Thousand Oaks, CA, Sage.

Seale, C. (1999) 'Quality in qualitative research', Qualitative Inquiry, 5(4): 465-78.

Seale, C. (2011 forthcoming) 'Secondary analysis of qualitative data' in D. Silverman (ed.) Qualitative Research, 3rd Edition, London, Sage.

Seale, C. & Charteris-Black, J. (2008) 'The interaction of class and gender in illness narratives', Sociology, 42(3) 453-69.

Silverman, D. (1998) 'Qualitative research: meanings or practices?' Information Systems Journal, 8, 3-20.

Stacey, M. (1974) 'The myth of community studies', in C. Bell and H. Newby (Eds) The Sociology of Community, London, Frank Cass, 13-26.

Tolich, M. (2004) 'Internal confidentiality: when confidentiality assurances fail relational informants', Qualitative Sociology 27, 101-6.

Townsend, P. (1979) Poverty in the UK- A Survey of Household Resources and Standards of Living, Harmondsworth, Penguin.
... since an essential part of qualitative research has always
   involved monitoring the effects of reflexivity and taking account
   of these effects in the analysis, there is no logical
   incompatibility between assessing the influence of contextual
   features in primary data analysis or in secondary data analysis.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.