CAQDAS 07 Abstracts

CAQDAS 07 : Abstracts

Here you can find abstracts of the papers presented, and links to downloadable full-papers where the authors have provided their permission.

Plenary Abstracts

1. The comparative method and computer-aided qualitative research
Udo Kelle, Institute of Sociology, Philipps University of Marburg, Germany

2. Honouring the past, scoping the future*
Ray Lee and Nigel Fielding, Directors of the CAQDAS Networking Project

3. Towards a Digital Visual Ethnography
Sarah Pink, Reader in Social Anthropology, Loughborough University

4. Comparative keyword analysis: a computer-assisted method for the qualitative analysis of text
Clive Seale, Department of Sociology & Communications, Brunel University, London

5. Beyond the Input Data Blob
Alan Stockdale, Center for Applied Ethics and Professional Practice, Newton, USA

Stream 1. Techniques for data collection and conversion

Technological Advances in the Recording and Transcription of Interviews in Qualitative Research
Gabrielle O’ Kelly, School of Nursing , Midwifery and Health Systems,University College Dublin, Republic of Ireland

Processes and Pleasures: digital technology in qualitative research
ML White, Middlesex University

Use of the Internet as a data collection tool: a methodological investigation of online synchronous interviews
Alison Evans, Jonathan Elford, Dick Wiggins, City University London

Using discourse analysis to analyse online interviews with rape survivors
Jennifer Yeager, Department of Applied Psychology, University College Cork, Ireland

Qualitative Data Exchange: Methods and Tools*
Louise Corti, UK Data Archive, University of Essex, UK

Stream 2. Techniques of qualitative analysis using software

Discourse parsing: analysing the non-linear organisation of text and talk*
Brian Torode, Trinity College, Dublin, Ireland

Sequence viewer: Qualitative and quantitative analysis of sequential data
Yfke Ongena and Wil Dijkstra, VU University Amsterdam, The Netherlands

Techniques of analysis using MAXQDA 07
Prof. Dr. Udo Kuckartz, Philipps Universitaet Marburg, Germany

International survey of users of qualitative data analysis and CAQDAS
Graham R Gibbs & Celia Taylor , University of Huddersfield, Ann Lewins & Nigel Fielding, University of Surrey,

Developing Computer-Aided Secondary Analysis : consequences of such an Innovation for Sociologists in an Industrial Context*
Dominique Le-Roux, Electricitie De France, Magda Dargentas, CETSAH, Paris & Mathieu Brugidou, Electricitie De France

To search or not to search? Issues raised in comparing differene secondary analyses of qualitative data
Joanna Bornat, The Open University, UK & Gail Wilson, London School of Economics

Secondary analysis of qualitative interviews: using NVivo to avoid the pitfalls of primary analysis
Dr. Cathy Murray, University of Stirling

Automated categorical coding of transcribed talk: a system that forecasts its future performance*
Richard Forsyth, David Clarke and Shaaron Ainsworth, School of Psychology, University of Nottingham

Using Atlas.ti for narrative research on informal learning
Giancarlo Gola, Dept of Education, University of Trieste

Automating the Analysis of Free-text Answers to Open-ended Questions*
Yuwei Lin, Peter Halfpenny, Jon Gibson, Firat Tekiner, James Nazroo and Rob Procter, University of Manchester

The Use of QL & QT software in a Mixed Method Research Design*
Marie-Hélène Paré, Department of Social Policy and Social Work, University of Oxford

Research Design, Units of Analysis and Software supporting Qualitative Analysis*
Silvana di Gregorio, PhD, SdG Associates, London and Boston & Judith Davidson, PhD, University of Massachusetts – Lowell

Computer-Aided Software for Qualitative Data Analysis: An Historical Overview and Contemporary Perspectives
Sharlene Hesse-Biber & Christine Crofts, Boston College, USA

A trip through arts: Analyzing children dialogue using NVivo 7
Héctor del Castillo, Universidad Autónoma de Madrid, Spain; Ana Belén García Varela, Rut Martínez, Mirian Checa, Universidad de Alcalá, Spain

Stream 3. Technologies and IT for the analysis of visual data

Qualitative Mapping, Qualitative Software: Using MAXqda2 to visualize social and spatial processes
César A. Cisneros-Puebla, Universidad Autónoma Metropolitana, Iztapalapa, México and Karen Andes, Emory University, Atlanta, GA, US

Visualizing the future of QDA with Transana 2.20
Nicolas Sheon, PhD, University of California, San Francisco and David K. Woods, PhD, University of Wisconsin, Madison

The World as a Document: GeoCoding with ATLAS.ti
Thomas Muhr, Dipl.-Psych., Dipl.-Inform.CEO, ATLAS.ti GmbH

Research results showed by a video and by a website
Dario Da Re, Department of Sociology, University of Padova , Italy

Virtual Qualitative Data
Ian Jones, Institute of Education, University of Warwick

Using Computer-Assisted Qualitative Data Analysis Software for Visual Sociology
Paolo Parmeggiani, University of Udine, Italy

And we’re rolling! VIA as a case for better software management of qualitative multimedia data*
Jennifer Patashnick, Video Intervention/Prevention Assessment (VIA), Children’s Hospital Boston

Digital Records: Synchronizing Visual Data with Diverse Resources
Andy Crabtree, Andy French, Chris Greenhalgh, Tom Rodden, Steve Benford, School of Computer Science & IT, University of Nottingham, UK

Creating visual transcripts: the role of computer software in the analysis of video data
Marie Joubert Gibbs, Graduate School of Education, University of Bristol

in the classroom using Transana: Children and parents learning together
Ana Belén García Varela, Universidad de Alcalá, Spain, Héctor del Castillo Universidad Autónoma de Madrid, Spain, Pilar Lacasa & Sara Cortés Universidad de Alcalá, Spain

Stream 1 or Stream 2

Stream 4. Impacts of software use on methodology and analysis

Enhancing quality and transparency; the role of Framework in supporting robust qualitative research
Kandy Woodfield & William O'Connor, National Centre for Social Research

‘Membership Matters’: How and Why Qualitative Researchers Should Consider Applying Membership Categorisation Analysis To Their Data Using CAQDAS
Andrew King, Department of Sociology, University of Surrey

On the Use of ATLAS.ti for the Reconstruction of Cognitive Maps from Coded Texts
Georg P. Mueller, University of Fribourg, Switzerland

CAQDAS, Context and the Recontextualisation of Qualitative Data
Libby Bishop, ESDS Qualidata, University of Essex, UK

The Use of CAQDAS in the UK Market Research Industry
Ruth Rettie, Helen Robinson, Anja Radke, Xiajiao Ye, Kingston University, UK

The contributions of QDA software to mixed methods research: a mixed success?
Normand Peladeau, Provalis Research

Gaining depth without drowning: How Framework can aid in the managementand interpretation of qualitative research data
William O'Connor & Kandy Woodfield, National Centre for Social Research

Coding for Human Language Tool Builders
Stuart W Shulman, University of Pittsburgh

Using Caqdas on Historical documents; some issues from the experience of the Health of the Cecils Project*
Caroline Bowden, University of London

Use and comparison of 2 software programs (ALCESTE / PROSPERO) in social representations’ analysis
Béatrice Madiot Université de Picardie- Jules Verne (ECCHAT), EHESS (GRSP et CETSAH-LPS), Magda Dargentas, Ecole des Hautes Etudes en Sciences Sociales (CETSAH-LPS)

CAQDAS: A ‘posthuman’ research experience?
Jo Haynes, Department of Sociology, University of Bristol

Mind-Software Interaction: Does the application of software cause an epistemological problem?
Dr. Susanne Friese, Method- und Media Centre, University of Hannover, Germany

Developing a model for the concept of Justice using a software : NVivo 2
Dr. Elif Kus, Ankara University, Turkey

Complexities, Levels, Quality and Ethical Issues in Qualitative Research: Do we Only Need Software?
Eugenio De Gregorio, PhD, Department of Developmental and Social Psychology, University of Rome

Benefits and drawbacks in a variety of approaches to the use of CAQDAS
Antonio Osorio, Altina Ramos, Ana Paula Martins, University of Minho, Portugal

Conversations about CAQDAS: An Analysis of Qual-Software Discussion List
Urszula Wolski, Royal Holloway, University of London

Stream 3

Stream 5. IT tools for collaborative working

Collaborative Video Data Analysis: The Mediational Effects of MiMeG and other Digital Technologies
Marie Joubert and Anne Manuel, University of Bristol

EdEt - Experimental collaborative CAQDAS project in cultural anthropology research*
Iwona Kaliszewska, Institute of Informatics and Institute of Ethnology and Cultural Anthropology, Warsaw University, Poland

Using IT to Conduct Collaborative Research Across Distributed Staff & Student Populations*
Denis Edgar-Nevill, Department of Computing, Canterbury Christ Church University

Introducing Advanced Qualitative Querying (AQQ) Project*
Sebastian Kaliszewski and Iwona Kaliszewska, Institute of Informatics, Warsaw University

Formal Technical Reviews for Research Projects
Andre Oboler, Computing Department, Lancaster University, UK

Evaluation of MiMeG in Use: Technical and Social Issues in Remote Collaborative Video Analysis
Dylan Tutt (King’s College London, UK) and Muneeb Shaukat (University of Bristol, UK)

The Role of Co-Interpretation of Routine Data in the Evaluation of a Virtual Research Environment
Richard Procter, Vito Laterza & Patrick Carmichael, Centre for Applied Research in Educational Technologies (CARET), University of Cambridge.

Collecting, Managing and Analysing Data within Multi-Disciplinary Research; Challenges from a four year collaborative research project
Sue Venn, Centre for Research on Ageing and Gender, University of Surrey

Stream 4 or Stream 5

The comparative method and computer-aided qualitative research
Udo Kelle, Institute of Sociology, Philipps University of Marburg, Germany

In this paper it will be argued that the most crucial methodological innovation provided by CAQDAS in the past 15 years is that it offers technical tools for the application of comparative methods. Furthermore it will be shown that the intelligent application of formal strategies of comparison included in CAQDAS packages may deepen our understanding of methodological hazards entailed in comparative analysis.

Therefore it is extremely helpful to draw on classical concepts of comparison, esp. on Mill´s methods of agreement and difference. In the paper earlier discussions about the use (and the misuse) of these methods in the works of Chicago School sociologists will be described and it will be shown that Mill´s methods also play a prominent role for Grounded theory methodology, although their application in qualitative analysis is almost never explicitly discussed in this context. The invention of CAQDAS did not only facilitate the use of comparative methods through ordinary code-and-retrieve techniques. CAQDAS also inspired the development of formal methods of comparison, for instance procedures for computer-aided hypothesis testing included in several packages and techniques which support the employment of “Qualitative Comparative Analysis” proposed by Charles Ragin.

Such formal strategies of case comparison can be used to demonstrate how two basic logical problems of comparative analysis – the problem of unknown background conditions and the problem of incomplete case selection – can exert disastrous effects on the interpretation of qualitative data.

Honouring the past, scoping the future
Ray Lee and Nigel Fielding, Directors of the CAQDAS Networking Project
Download Slides
Download Full Paper

The paper will chart the progress of Computer-Assisted Qualitative Data Analysis (CAQDAS) in general and the CAQDAS Networking Project in particular as they have developed in tandem since 1989, when the authors organised the world's first international conference on qualitative software. Among our themes will be the eternal recurrence of necessary if familiar concerns about the impact of CAQDAS on the craft of qualitative research and the canon of qualitative methodology, the distinctive nature of software development in this field, trends in user patterns, the role of CAQDAS in advancing more systematic, robust and formal approaches in academic research, and its role in the enhanced status of qualitative research in the applied research world. The historical account serves as a basis for a speculation on future developments grounded in the promise of a ubiquitous computational environment, Grid resources and cyber-research. The paper will also take the opportunity to sketch the contours of the conference overall and to indicate the vision around which the conference's intellectual dimensions constitute a coherent state--of-the-art for the CAQDAS field.

Towards a Digital Visual Ethnography
Sarah Pink, Reader in Social Anthropology, Loughborough University

In this presentation I will discuss the possibilities and potentials for greater use of software in a digital visual ethnography process. Drawing from some examples from my own visual fieldwork I shall demonstrate how visual ethnographers use digital media as part of their research. However the main objective of visual ethnographers is not simply to produce visual materials for storage and analysis but to also use these materials as ways of representing other people’s experiences in ways that are both theoretically informed and that can contribute to interdisciplinary theory building. I shall explore some of the potentials of audio visual materials for communicating about experience transculturally, discuss existing hypermedia representations, and tentatively suggest how, ideally, these might be supported by software.

Comparative keyword analysis: a computer-assisted method for the qualitative analysis of text
Clive Seale, Department of Sociology & Communications, Brunel University, London

I will describe a method which I have been developing for the comparative analysis of texts. So far, it has been applied in two studies: first, a study involving interviews and internet forum postings which compared text produced by men and women with concerns about cancer; second, a study of newspaper articles where the focus of interest has been a comparison of 'serious' versus 'popular' newspaper coverage of sleep and sleep disorders. The method uses Wordsmith software to identify words used more frequently in one text compared with another. Keywords are examined in their context and, in a process similar to conventional qualitative thematic coding, are placed into semantic groups to facilitate an interpetive and eventually qualitative analysis of key differences distinguishing texts. Like any scientific technology, the method is good for some things but not others and I will outline these advantages and disadvantages. Like conversation analytic transcription which extends the senses so that features of interaction normally hidden from sight are foregrounded, comparative keyword analysis allows the analyst to 'see' things that would otherwise go unremarked. In this sense it is a 'scientific' method, analogous to a microscope or a telescope, extending the capacity of human senses to apprehend otherwise hidden phenomena. By the time of my talk I may be able to report progress on further development of the method, funded by an ESRC grant. This involves the investigation of illness narratives found in interviews and web forums.

Beyond the Input Data Blob
Alan Stockdale, Center for Applied Ethics and Professional Practice, Newton, USA

The way talk is recorded, transcribed and analyzed has not changed significantly since before the advent of computer-based analysis. Recordings have become digital, transcription has become computer-based, and transcriptions are now input for analysis into qualitative data analysis software but the core of the process is largely unchanged. QDA software has developed to utilize audio data directly, potentially bypassing transcription, but analyzing audio directly without the benefit of transcription is hard. The potential to use computerization to work with audio data in genuinely new and beneficial ways has not been fully realized yet. In the past there have been significant technological barriers. It is only in recent years that technologies to record, manage and process digital audio data in new ways have become widely available. This talk will discuss those technological developments, particularly how transcription has developed into a process that can create structured data that remains closely linked to the digital audio data from which it was derived, and the current and future potential for aiding the management and analysis of qualitative data.

A common strategy for data collection in qualitative research is that of interviewing. This researcher had recently completed a qualitative study* of the possible impact of nurses’ unconscious mental processes on their work with patients. In this study, each research participant is interviewed on a number of occasions. The practicalities of recording and transcribing these interviews using modern technology are presented in this paper.

The rational for the use of electronic versus tape recordings of interviews is outlined. Detailed description of the use of a digital (electronic) voice recorder is provided, including the saving of the recordings onto a computer (PC) and making back up copies of recordings onto a CD.

Transcription of recordings is well recognised by researchers to be a very time consuming process. Sometimes as a means of saving time, professional typists are commissioned to carry out this work. However, this can raise both practical and ethical dilemmas. Clerical assistance for this work can be expensive. In addition, where sensitive personal information is being given by research participants it is especially important that the promise of confidentiality of material is fully respected by the researcher. As the latter issue was a priority in this study, the latest technology was used to make the transcribing process as efficient as possible. To this end, speech recognition software was used to ‘speak’ the interviews to the computer which transcribed them into print. A brief demonstration of this method of transcription is given and the advantages and disadvantages of its use will be discussed.

Finally, the transcribed data was inputted into Atlas ti data analysis software. Some key points about this will be discussed, and the advantages of using Atlas ti for this particular study will be outlined.

* The presenter gratefully acknowledges the Health Research Board, Dublin for funding of this project as part of a Clinical Research Fellowship in Nursing and would also like to express gratitude for support received from An Bord Altranais (The Nursing Board), Dublin

Processes and Pleasures: digital technology in qualitative research
ML White, Middlesex University

This paper reflects on ethnographic educational research conducted in 2005 and 2006 at Educational Video Center, a non-profit media education organization targeting hard to reach youth in New York City. The research considers how digital technology is used to create a participatory pedagogy and a learning environment that enables young people to reflect critically on themselves, their environment, and their digital video (DV) productions. The research brings together three parts of a story: understanding how meaning is made through a complex series of interactions (between documentary maker and teacher, text and technology and the audience); describing the impact and value of an experience and being a digital ethnographer in a community of enquiry. In this research the invisible research narrative is made visible in part through technology but also due to the research ethics which are informed and inform theory – the emic perspective which researches with, the collaborative nature of the account and the reflexive process for the researcher. In this research DV ethnography is an enabling methodology; an approach to experiencing and understanding a specific educational and cultural process.

This research is both anthropological in that it is focused on social relationships and culture and ethnographically styled as it attends to the social contexts of contemporary DV production. Although I use the term DV ethnography the research style has much in common with visual ethnography (Morphy and Banks 1997, MacDougall 1998, Ruby 2000, Grimshaw 2001) where video, film and photography along with other electronic media are created and viewed as cultural texts; representations of ethnographic knowledge and “sites of cultural production, social interaction and individual experience that themselves form ethnographic fieldwork locales” (Pink 2001, Introduction). This research attempts to understand the value and impact of specific educational experiences rather than generate a thick description (Geertz, 1973) of young people's cultures. Recognising that conventional writing presents a ‘thin’ description multi-media and multi-modal texts have been included to counter the representation of a single implicit point of view that does not do justice to the complexity of the experience.

This paper will focus on the use of digital technology as a research methodology and the implications of digital production, presentation and dissemination of educational ethnographic research. The presentation will offer DV examples and ways of analysis.

Use of the Internet as a data collection tool: a methodological investigation of online synchronous interviews
Alison Evans, Jonathan Elford, Dick Wiggins, City University London

Background Online synchronous interviews are conducted in “real time” using online venues such as chat rooms or MUDs, or by means of messaging or conferencing software. This paper examines the pros and cons of collecting qualitative data using online synchronous chat.

Methods Follow-up interviews were conducted with gay men who had already taken part in the Internet and HIV study, a study of high risk sexual behaviour among gay and bisexual men in London. Between July and September 2003, 31 gay men living in London were interviewed one-to-one for the Internet and HIV study by an experienced qualitative researcher (MD). MD conducted online synchronous interviews with 17 men and face-to-face (traditional) interviews with 14 men. AE then conducted follow-up interviews with 6 of the 17 online interviewees and 5 of the 14 face-to-face interviewees. The follow-up interviews were conducted in the same mode as the first interview. They were designed to explore the interviewees’ experience of their earlier Internet and HIV interview. All online synchronous interviews (both original, with MD and follow-up, with AE) took place in “private chats” between the interviewer and interviewee which could not be accessed by other people.

Results Online synchronous interviews appear to be suited to a more quantitative, structured format in order to reduce respondent burden, given the extra demands of typing in real time. The use of such interviews increases turnaround time through production of an instant transcript and the opportunity for respondents to slot their online interviews into other arrangements. This requires a flexible approach from the interviewer who also needs experience with online chat in order to adapt to the respondent’s style of communicating.

Conclusions Online synchronous interviews may be used in conjunction with or as a cost-effective supplement to face-to-face interviews. They may, for example, provide a cheap method for scoping out issues for future research or a tool for the rapid generation of data as part of a grounded theory approach, whereby data are collected from a variety of sources until emerging categories are saturated.

Using discourse analysis to analyse online interviews with rape survivors
Jennifer Yeager, Department of Applied Psychology, University College Cork, Ireland

The paper discusses the issues involved in using discourse analysis to analyse interviews with rape survivors conducted via Instant Messenger online. The data was collected as part of a larger study investigating how rape survivors attend to the role of their social relationships in coping with rape. The current paper considers the practicalities involved with conducting online interviews on such a sensitive topic and how well such data lends itself to subsequent discourse analysis.

Discourse analysis is a methodology that focuses on what is constructed between people in social interaction and how “facts” about the world are produced in discourse. The relationships and social networks of rape survivors need to be examined in conversation, and not through “forced-choice” questionnaires, in order to see how the respondents produce the concepts of coping and support in discourse.

If the Internet experience is a viewed as a process of negotiation it thus follows that the only way to understand it is to analyse the people involved in it. This, I argue, is why discourse analysis is the most beneficial type of analysis for examining online interviews. Such a dialogic approach can assess the online conversational interaction, especially when conversations are increasingly being performed in virtual contexts. Online chat does have a conversational structure in terms of turn taking and message length, which makes it suitable to be analysed using discursive techniques.

As a result, methodologies that analyse these accounts are essential and online data collection is a feasible and important option in the attempt to collect data on sensitive topics such as rape and sexual abuse. As D.A. recognises that coping is constructed in conversation it is a practical and useful methodology for analysing interviews conducted on the Internet in real-time in order to examine the discursive techniques used by respondents in conversation.

Qualitative Data Exchange: Methods and Tools,
Louise Corti, UK Data Archive, University of Essex, UK
Download Slides

This paper will present recent finishing from a project that is exploring the feasibility of developing data exchange models and data conversion tools for primary research data collected in the course of qualitative research. A standard format for representing richly encoded qualitative data is necessary because: it ensures consistency across datasets; it supports the development of common web-based publishing and search tools; and it facilitates annotated data interchange and comparison among data collections. Importantly, it should enable data and linked products to be imported and exported directly into and out of CAQDAS packages, avoiding the reliance on just a single product, and offering the opportunity to share analytic workings outside the confines of any particular software.

The project is developing, refining and testing models for data exchange for qualitative research data based on XML/RDF schema, while incorporating existing schema such as the Data Documentation Initiative (DDI), Text Encoding Initiative (TEI) and a draft Australian format (QDIF). The test data selected for this project are from the social sciences, but these formats are typically found across all domains of primary research. A small scale evaluation of the models and tools is being undertaken to scope out whether a functional and scalable facility or service where data formats can be submitted and seamlessly returned in a chosen, desired format might be possible.

Discourse parsing: analysing the non-linear organisation of text and talk
Brian Torode, Trinity College, Dublin, Ireland.
Download Slides
Download Full Paper

Both speech and writing are naturally linear (one-dimensional) in organisation in that turns are heard, and lines of text are read, in a single stream, one at a time. This homology justifies the view of the transcript as a ‘natural’ representation of talk. (“Conversation Analysis” [CA] has of course developed representations of overlapping talk, and other accompanying sounds such as laughter and applause, by showing two or more lines of transcribed text in parallel with one another.)

While such representation heightens awareness of adjacency features in text and talk, it diverts attention from remote relations. CA does acknowledge remote relations, for instance in insertion sequences where the response to a first pair part is suspended until an inserted pair has been completed; and in the case of story-telling, where a preface announces at the start what it will take for the story to be over, which announcement recipient holds in reserve throughout the subsequent narrative. But it is the responsibility of the analyst to hold these in mind: the transcription notation does not assist in this task.

Word processing and qualitative analytic software offer various means whereby remote relations can be marked, including:

(i) Coloured text/coloured highlighting/coloured underlining/coloured sidebars. Coloured text is supported by almost all word processors and qualitative packages, but is generally not supported in academic publication.

(ii) Tables – when used together with drawing devices such as arrows, shading, and text boxes these can generate detailed two dimensional maps of discourse in progress. With other software, such as Microsoft Powerpoint, such diagrams can be put in motion and precisely linked to transcript and audio-visual playback.

(iii) Hierarchical headings (outline view)

(iv) Tagging by way of bracketing, also brackets containing keywords which may be mechanically searchable.

This paper will discuss, with examples, whether any of these means can be regarded as a “natural” representation of the non-linear (two or more dimensional) organisation features of naturally occurring talk and text, just as “maps”, including contour lines and colouring, are generally accepted as a “natural” three-dimensional representation of the landscape.

A particularly significant use of such devices, if they are justifiable at all, would be in discourse parsing, i.e. the precise delineation of distinct discourses within the data concerned. In a recent critical review of “Discourse Analysis” [DA], written from a CA point of view, Robin Wooffit (2005) argues that “there is a clear lack of consistency as to what counts as evidence for the presence of a discourse: it could be a single word, or a short stretch of talk, or a slightly longer account or narrative” (p.183). A significant step towards greater consistency would be achieved if there could be agreement on acceptable ways of marking texts and transcripts to distinguish these discourses.

Wooffitt, R. 2005. Conversation Analysis and Discourse Analysis. London: Sage.

Sequence viewer: Qualitative and quantitative analysis of sequential data
Yfke Ongena and Wil Dijkstra, VU University Amsterdam, The Netherlands

Sequence viewer is a unique tool to combine qualitative and quantitative analysis. The program was developed to manage, code and analyse sequential data originating from video, audio and/or text. Such data can be behavioural data (e.g., verbal behaviour in interviews or other conversations, animal behaviour, etc.), but also other types of data (e.g., newspaper articles, job histories) that can be considered a sequence of events. The essential idea of Sequence viewer is that a particular event (an utterance by the interviewee, a paragraph in a newspaper article) partly depends on preceding events. Sequence viewer enables researchers to study relations between events at different time points and between events of different types.

The files in the program consist of sequences. A sequence consists of descriptions of events for a single case. For example, in studying interviews, a sequence may consist of all (non)verbal behaviour of interviewer and interviewees, related to a particular question or topic discussed in the interview. The events may consist of transcripts of actual verbal utterances and/or descriptions of non-verbal behaviour. The events can be individually described with alphanumerical code variables or with numerical event variables. Code variables, which are directly linked to individual events, are crucial for the possibilities of analyzing the sequential patterns in the data (e.g., which events precede or follow a specific event).

Texts can also be analyzed by means of text keys. A key may point to a single word, but also to a large piece of text, for example consisting of a number of utterances by interviewer and interviewee in an interview, discussing a particular topic. The assignment of codes and keys can be done automatically, using a fuzzy text search (e.g., assigning a particular code if a piece of text resembles a particular text string).

The program also has options for visual coding, which is especially suited for coding non-verbal behaviour (e.g., smiling or gestures) in video files. To each sequence a link to an actual sound or video file can be added. This file can be played directly from within Sequence viewer. Time keys can be inserted to a particular part of the sound or video file and labelled with appropriate keywords. The keys can be converted to code-, event- or sequence variables, which will allow quantitative analyses. In addition, the program offers a number of options to visualise sequential patterns of question-answer sequences, or interesting relationships; for example the relation between verbal utterances and non-verbal behaviour, and how such behaviour proceeds in time.

Each sequence may contain an unlimited number of sequence variables. Such variables concern characteristics of the whole sequence; e.g., the gender of interviewer and interviewee, the question number, the eventual score on the question posed, or a description of the sequence based on for example the frequency or order of codes or keywords assigned to (parts of) the sequence. With these variables more common statistical analyses are possible. In addition it is possible to export these variables for analysis with common statistical packages.

Techniques of analysis using MAXQDA 07
Prof. Dr. Udo Kuckartz, Philipps Universitaet Marburg, Germany

This paper will discuss the problem of data visualisation from two perspectives: in the first part I will focus on theoretical and methodological issues and argue that visual thinking is creative, helpful for exploration and for discovering patterns, relations and structures in social science data. Scholars like Tony Buzan or Lothar Krempel have shown conclusively that complex information and dependencies can be translated into easy-to-read graphical representations. The second part of the paper will focus the practice of research, particularly the new visual tools integrated in MAXQDA 07.

Data display and visualisation are areas of research methods that become more and more important, particularly in qualitative and mixed methods approaches. As Matthew Miles and Michael Huberman pointed out in their widely known book Qualitative data analysis– an expanded sourcebook better displays are a major avenue to valid qualitative data analysis“. Data displays may have multiple functions in the analysis process: data condensation, data reduction, data exploration, idea management and last but not least data presentation. According to Miles/Huberman, the dictum “You are what you eat” might be transposed to “You know what you display”.

In the practice of data analysis, it is useful to differentiate between two types of visualisations: case-oriented and cross-case visualisations. Case-oriented visualisation like MAXQDA’s TextPortrait and Codeliner offer a visual display of the codes applied to a text. These displays are interactive and directly connected to the source material. Cross-case displays like MAXQDA’s Code-Matrix Browser show more complex information and make it easy to identify relations.

International survey of users of qualitative data analysis and CAQDAS
Graham R Gibbs & Celia Taylor, University of Huddersfield, Ann Lewins & Nigel Fielding, University of Surrey,

As part of an ESRC funded project ( RES-333-25-0009) “Online QDA” ( http://onlineqda.hud.ac.uk/ ) that addressed the need for online materials for researchers learning qualitative data analysis (QDA) and computer assisted qualitative data analysis (CAQDAS) we undertook in 2005 an online survey of qualitative researchers. This investigated a range of issues including the subject background of the researcher: the main approaches to analysis they were using, the training (if any) they had received in QDA and CAQDAS, what they needed training about and what aspects of QDA they had found difficult to learn, whether they had used software (and why or why not), which software they had used, their experience of learning the software and their expectations about it.

Because it was an online survey, the representativeness of the sample is hard to judge. However, we went to some lengths to ensure that those we invited to respond were not just those currently using CAQDAS. However, most respondents had used CAQDAS at some time in their research.

251 researchers replied, with three-quarters of replies coming from universities. Nearly 45% of respondents were from the UK with just over a quarter from the US or Canada. The biggest group was PhD students (32%) with significant numbers of academics and researchers.

The most frequently used analytic approaches were, grounded theory (over half of the respondents used this approach), ethnography, action research, narrative, action research and discourse analysis. However, many respondents used more than one analytic approach.

Examining the use of approaches across disciplines, the popularity of grounded theory was preserved in most except, unsurprisingly in anthropology. Here, ethnography was popular but this approach was also commonly used in education, as was action research. Overall the picture was one where a few methods predominate in most disciplines, but at the same time there was a wide variety of approaches adopted by some researchers, with no discipline showing any exclusive methods.

Most of the sample had received some training in QDA, though over a quarter were self-taught. 70% of the postgraduate student respondents had received training in QDA during the course of their PhD.

Not everyone used CAQDAS all the time. Nearly three-quarters had done QDA without software assistance in some projects. This was often because there was no access to software at the time but also because the researchers considered the data set too small or it would take too long for them to learn the software. Although 18% had never used software at all, for those who had, the most popular programs by far were NVivo, ATLAS.ti and Nud.ist.

The overall picture amongst researchers was that CAQDAS use was a common and accepted part of QDA practice. Nevertheless, researchers were selective about software use, reserving it for projects where it could be most effective or provide the most benefit.

Developing Computer-Aided Secondary Analysis : consequences of such an Innovation for Sociologists in an Industrial Context
Dominique Le-Roux, Electricitie De France, Magda Dargentas, CETSAH, Paris & Mathieu Brugidou, Electricitie De France
Download Slides
Download Full Paper

Sociological research is an integral part of all marketing decisions enabling companies to undertake relevant analysis of their target markets. An electricity Company such as EDF needs to explore various topics : context of political and cultural transformation, poverty evolution, mobility, energy-related practices, estimating consumer willingness to pay for an efficient energy, social acceptability of nuclear plants or high tension wires ...

The authors will present how Verbatim project has been conducted in the Social Sciences Department of EDF R&D. The objective of this project, which began in 1998, under the inspiration of Qualidata model, is to archive qualitative data and make secondary analysis possible. Nearly 10 years after the beginning of this experience, quite new in France, it is interesting to take stock of the situation : a substantial database of sociological interviews has been developed, introductory courses to CAQDAS have been proposed to the sociologists of the company, several of them have done secondary analysis on their own data or in collaboration with other researchers, methodological research on these topics is being done in partnership with other institutions.

After some years of stagnation of the project due to different factors, the experience seems positive. This paper will analyse, from the “innovation sociological theory” point of view, how this new practice in a French industrial context has progressively diffused and how it is changing the way of working (collaborative work, resolution of methodological issues..). It will also show how it is getting recognition both from the scientific researchers and from the industrial managers point of view.

To search or not to search? Issues raised in comparing differene secondary analyses of qualitative data
Joanna Bornat, The Open University, UK & Gail Wilson, London School of Economics

he secondary analysis of qualitative data sets has won legitimation through theoretical and methodological argument and is now an area of lively development and debate. There are, however, few worked examples to draw on and so still much to be said about process and product. One such area of debate is raised by questions of ethics. To return to someone else's, or even one's own, data raises issues for all involved, from original interviewees to the secondary analyser.In this paper we will consider some questions raised by current developments in ethical guidelines, comparing examples from two different searches of the same data set. We used N6 to search for various categories that led to text relating to ethnicity, community, family. Our paper will discuss some of the advantages and disdavantages of using archived data which we had not generated . In the first example we begin to analyse changes in the attitudes to family and community held by pioneers of geriatric medicine. The second example considers their use of an Asian workforce and evidence of personal and institutional racism. Comparing with other disciplines, we ask why secondary analysis remains ethically contentious solely, it seems, for social scientists.

Secondary analysis of qualitative interviews: using NVivo to avoid the pitfalls of primary analysis
Dr. Cathy Murray, University of Stirling

While the practical benefits of cost and time of secondary analysis of qualitative data are rarely contested, it is the challenges (e.g. decontextualisation and whether the data are fit for purpose) which frequently dominate the discourse. Using a recently completed re-analysis of 112 qualitative interviews with teenagers, entitled Quest for Identity: Young People's Tales of Resistance and Desistance from Offending, as an exemplar, I highlight the benefits as well as the pitfalls of re-visiting the data for a doctoral thesis. I discuss the mechanisms for avoiding some of the pitfalls and end with a broader question as to what follows secondary analysis.

With access negotiations, fieldwork and the agenda setting constraints of government funders removed, secondary analysis arguably enables researchers to engage more creatively with datasets. In this case, the re-analysis of the interviews using NVivo led to an exploration of the how rather than the why questions. Drawing particularly on the work of Derrida on identity and binary oppositions, new perspectives emerged as to how young people maintain their resistance to offending. A pitfall encountered in the original study was a ‘coding fetishism' and mechanisms to avoid this in the re-analysis were instituted. For example, a self-imposed limit was set in advance on the number of nodes, reduction sessions were systematised and conceptualisation was reinforced. With regard to the latter, a feature of the primary study had been a disproportionate number of descriptive compared with conceptual nodes. The effort to avoid a repetition of this began with a decision to take a ‘clean slate' approach to re-analysis, which was made possible by the fact that the original transcripts of the interviews were available in raw form. They were therefore introduced into NVivo, rather than re-worked in NUD.IST (the programme used in the primary study), with its attendant noding system. A further strategy adopted was to create a visual separation between the descriptive and the conceptual nodes, which was achieved by having the former in the ‘free' section of NVivo and the latter in the ‘tree' section.

On completing secondary analysis, researchers may consider further re-analysis of their datasets, although they may be dissuaded by the practical and ethical constraints associated with this. For example, in this study, would the original team, funding body and participants agree to the dataset going beyond the original researchers? Would it in any case be ethical or feasible to contact the young participants to renegotiate consent? However, these are dilemmas with which the secondary analyst is familiar. Of more interest is the status of further re-analysis. Does it merit different terminology, such as tertiary analysis? Should it be regarded as emanating from the primary analysis or does secondary analysis constitute a separate status to which further re-analysis relates? And is there infinite scope for multiple re-analyses or can databases be analytically exhausted?

Automated categorical coding of transcribed talk: a system that forecasts its future performance
Richard Forsyth, David Clarke and Shaaron Ainsworth, School of Psychology, University of Nottingham
Download Slides

Despite recent proliferation of extralinguistic electronic information (audio files, video recordings, event logfiles etc.), verbal data is still a fundamental resource for researchers in a wide variety of fields. Social scientists, for example, must frequently spend a great deal of time and effort recording, transcribing, segmenting, and categorizing segments of natural language. Each segment may often be only a few words in length. Linguistic data, whether written language or transcribed talk, is gathered in situations as diverse as interviews, online communities, email, face to face conversation, and chatrooms. Researchers use such data in many ways. For example: they examine self-explanations in learners' verbalizations to assess whether students are learning effectively (Chi et al., 1989), elucidate the hierarchical and sequential structure of discourse acts in dialogue (Carletta et al., 1997), or try to identify communicative strategies used by participants in a collaborative problem-solving task (Green & Gilhooly, 1996). Most of these tasks involve assigning categorical codes to short stretches of natural-language text.

The present paper reports on the progress on the CODELEARNER project, which applies machine-learning techniques to assist in the categorical coding of short text segments. We focus particularly on two issues raised by the attempt to automate this analytical process: firstly, the inescapably iterative nature of the process of 'training' a computer system to emulate an analyst's coding decisions; secondly (and consequently) the fact that such a system must predict its own future performance in order for this iterative process to terminate without wasted effort. Given that most coding schemes are generated and/or refined during the analysis, adding self-prediction to the learning system is a necessity: it enables the researcher to re-train it at any convenient interval -- for instance, after coding every 20 or 50 or 100 new instances -- and have it estimate its own accuracy at various future points. This allows early exit from problems where the learning system proves incapable of the categorization task or finely-judged effort in situations where it is capable of reaching a sufficient accuracy level. In either case, human effort is saved.

Empirical trials on a corpus of self-explanations (Ainsworth et al, 2007) as well as the HCRC MapTask corpus (Anderson et al., 1991) suggest that a simple 3-parameter model enables the computer system to predict its own future "learning curve" with sufficient accuracy to make iterative computer-assisted coding a practical proposition. It is noted that success in this enterprise will have methodological implications: it will make feasible an incremental, exploratory approach to machine-assisted coding, thus opening up areas of research where the coding task would previously have been dismissed as too arduous.

Ainsworth, S.E., Forsyth, R.S., Clarke, D.D., Robertson, L. & O'Malley, C. (2007). Automatic coding of learner's self-explanations when learning from diagrams. EARLI 2007 Conference, Szeged, Hungary, 28 August - 1 Sept 2007.

Anderson, A.H., Bader, M., Bard, E., Boyle, E., Doherty, G., Garrod, S., Isard, S., Kowtko, J., McAllister, J., Miller, J., Sotillo, C. & Thompson, H. (1991). The HCRC Map Task corpus. Language & Speech, 34, 351-360.[http://www.hcrc.ed.ac.uk/maptask]

Carletta, J., Isard, A., Isard, S., Kowtko, J.C., Doherty-Sneddon, G., Anderson, A.H. (1997). The reliability of a dialogue structure coding scheme. Computational Linguistics, 23(1), 13-31.

Chi, M. T. H., Bassok, M., Lewis, M. W., Reimann, P., & Glaser, R. (1989). Self-explanations: How students study and use examples in learning to solve problems. Cognitive Science, 5, 145-182.

Green, C. & Gilhooly, K. (1996). Protocol analysis: practical implementation. In: Richardson, J.T.E. (ed.) Handbook of Qualitative Research Methods for Psychology and the Social Sciences. British Psychological Society, Leicester.

Automating the Analysis of Free-text Answers to Open-ended Questions
Yuwei Lin, Peter Halfpenny, Jon Gibson, Firat Tekiner, James Nazroo and Rob Procter, University of Manchester
Download Slides

Open-ended questions are crucial in surveys as they have the advantage of greater freedom of expression that allows respondents to qualify their answers without being restricted to limited response ranges. However, traditionally, researchers undertaking analysis of open-ended questions have read all the answers word by word and coded them manually.

This is true whether or not researchers use so-called computer-assisted qualitative data analysis packages. Coders might also misclassify responses. Coding open-ended questions is therefore time consuming and as a consequence its scope is commonly small scale, and its results are prone to human errors.

This paper provides a preliminary report on a pilot study that has the aim of exploring whether the analysis of free-text answers to open-ended questions can be efficiently and accurately automated through data and text mining techniques. These mining techniques offer the possibility of avoiding human errors and of processing large amounts of textual and numerical data in a short time. They also offer the possibility of revealing previously undiscovered relationships in the data. Our pilot study has acquired a corpus of full text answers to some open questions and the codes applied manually to those answers by researchers.

Interview transcripts inevitably contain spelling, grammatical and typing mistakes. When researchers code by hand, they deal with these errors as they go along, because they can normally correct them using their common sense and their knowledge of the context. However, before text mining techniques can be applied, these mistakes need to be corrected in a separate, pre-processing stage. Our first question is, therefore, the extent to which the mistakes can be automatically corrected through the application of computer techniques. A parallel issue is the degree of tolerance that mining techniques have to these sorts of mistakes, which we address by evaluating the results of mining after cleaning the data against those achieved using the same mining techniques on the original, uncleaned data.

There are numerous text and data mining techniques. Our second topic is to identify which of these techniques are suitable for coding free-text answers to open questions. We are focusing first on classification and clustering techniques. Addressing this topic involves comparing the results of applying mining algorithms with the results produced manually by researchers. This itself requires that we establish methods for comparing the two results, bearing in mind that the researchers will have made some mistakes in applying codes. The outcome of the evaluation will be used to refine the mining techniques, which will then be applied to the data again, and the new results evaluated, in a series of iterations.

Overall, the pilot will provide evidence of the contribution that data and text mining techniques might make towards automating the process of coding free-text answers to open-ended questions, which is currently a slow, labour-intensive and error-prone part of the social research process.

The Use of QL & QT software in a Mixed Method Research Design
Marie-Hélène Paré, Department of Social Policy and Social Work, University of Oxford
Download Slides

This doctoral research intends to develop the foundation of a theory of participation as it applies in mental health interventions and research conducted in communities affected by natural or man-made disasters. Although it is widely assumed that local participation helps to maintain low project costs and increases the effectiveness of aid, the construct of participation lacks rigorous theoretical examination and clear definition. This study aims to convert the experience of mental health practitioners and researchers working with disaster affected communities into a coherent set of hypotheses, definitions and variables embodying participation theory. In attempting to answer the research question: In disaster mental health programmes what forms does local participation take, what constructs accounting for its use and what impacts does it have on project outcomes?, an exploratory, sequential mixed methods design was adopted. The research process was supported by various QL & QT software which suited the eclectic nature of the data collection, the variety of the sources analysed, and the selection of a sampling frame.

Phase 1 of the data collection was qualitative and aimed at identifying the forms which local participation takes, and the conditions under which it occurs, in disaster mental health interventions and research. A metasynthesis (Dixon-Woods et al., 2005) of the evidence of participation in these programmes was conducted, using text and survey data as primary sources. Grounded theory led the analysis and interpretation procedures (Corbin and Strauss, 1990). The metasynthesis findings revealed the variations in participatory practices in disaster mental health interventions and stressed the lack of such practices in research. These findings were used to formulate a hypothesis which was tested in Phase 2. NVivo 7 supported the process of Phase 1.

Phase 2 was quantitative and used Concept mapping, a structured conceptualisation process that uses multi-dimensional scaling and cluster analysis to describe how respondents understand, and rate in terms of their importance, the different constructs pertaining to a phenomenon (Kane and Trochim, 2006). Concept mapping aimed at eliciting the factors that promote and restrict local participation in the work of 61 clinicians and researchers identified in the sample of text and survey data of Phase 1. Concept mapping’s results provided a series of juxtaposed multidimensional and cluster maps which gave a clear account of the variety, and the distance amongst, the various constructs accounting for the promotion and restriction of local participation. Concept Systems with QDA Miner-WordStat softwares supported the work of Phase 2.

Phase 3 consisted of qualitative interviewing with the 10 most influential mental health figures in the field of disaster mental health. Interviews explored the actors’ perceptions on how local participation may impact on intervention and research outcomes, using the results of Concept Mapping as interview guide. To draw the sampling frame of the 10 most influential figures, citation frequency of authors was used with the sample of text and survey data of Phase 1. Citation network analysis (Hummon and Doreian, 1989), with the software UCINET, demonstrated the extent of the connectivity between authors who published, and have being cited the most, by their peers.

The presentation will highlight the key software maneuvers used in the research with a focus on how a mixed methods architecture design should keep their use under the control of the researcher.

Corbin, J. and Strauss, A (1990) "Grounded Theory Research: Procedures, Canons, and Evaluative Criteria", Qualitative Sociology, vol.13 (1): 3-21

Dixon-Woods, M., Agarwal, S., Jones, D., Young, B. and Sutton, A (2005) "Synthesising qualitative and quantitative evidence: a review of possible methods", Journal of health services research policy, vol.Jan 10 (1): 45-53

Hummon, Norman P. and Doreian, Patrick (1989) "Connectivity in a Citation Network: The Development of DNA Theory", Social Networks, vol.11 (1): 39-63

Kane, M. and Trochim, William. (2006) Concept Mapping For Planning and Evaluation, Series Applied Social Research Methods. Thousands Oaks: Sage

Research Design, Units of Analysis and Software supporting Qualitative Analysis
Silvana di Gregorio, PhD, SdG Associates, London and Boston & Judith Davidson, PhD, University of Massachusetts – Lowell
Download Slides

Many people approach software use without much thought of the design of their research project. The fact that they are constructing a database for their project is not immediately obvious. It is only in later stages of the analysis that they may feel frustrated at not being able to get at the analysis they want and realise that they need to re-think the organisation of their work in the software. In this paper, we propose that the design of the research is the first consideration people should have when setting up a project in a software package. The information they need should be in their research proposal. It is a common misapprehension that attending to research design (whether with or without software) is nothing new. Yet, shifting to understanding good research design in qualitative research software requires new knowledge, skills, and techniques. A framework for issues they need to consider is proposed according to the complexity of the research design. Of particular importance is to understand what the unit of analysis is in the project. This aspect of the design is often implicit in qualitative studies but if this is not made explicit in the setup of the project, it will lead to difficulties with analysis. In our experience consulting and supervising projects, many qualitative studies have more than one unit of analysis. These need to be made explicit and be reflected in setting up a project using CAQDAS software. We illustrate this point through three case studies – one using ATLAS.ti 5, one using QSR NVIVO 7 and one using QSR XSight 2. The same questions about design are relevant irrespective of the software package used.

Computer-Aided Software for Qualitative Data Analysis: An Historical Overview and Contemporary Perspectives
Sharlene Hesse-Biber & Christine Crofts, Boston College, USA

We examine the history and impact of qualitative data analysis software on qualitative research in the social and behavioural sciences. We challenge traditional categorisation of software products that takes an "expert approach" to assessing their utility for qualitative researchers, one that relies on a "features-based” approach. We instead offer a "user's perspective" approach to utilizing and assessing computer-assisted software. For these reasons, and because of the increasing diversity of programs and features, attempts at categorizing or grouping CAQDAS programs seem limiting. We propose a different way of thinking about CAQDAS programs that further develops Renata Tesch's imperative to acknowledge the user's role in determining how qualitative data analysis software can be used. This new approach can be thought of as reflective, user-based and non-hierarchical evaluation. From this perspective, each individual CAQDAS user takes on the role of “expert” in deciding which program he/she would like to use. Before simply selecting a program based on its performance of particular function or its “rank” along the hierarchical continuum, the user asks him/herself a number of questions that will encourage reflection about how the software will fit into the larger scheme of the user's research agenda. Some of these questions would include the following:

1. What is my research and analysis style? - How do I conduct my analysis, and how might computers fit into that style? How might they enhance (or detract from) my analysis?

2. How do I want the program to assist me? - What tasks do I want to mechanize? What are my expectations of what the program will be able to assist me in doing? Are my expectations realistic?

3. What resources are available to me? - Which programs can my computer support? Which programs can I afford? What resources (time, personnel, material) necessary for learning how to use this program are available to me?

4. What are my preconceptions about these programs? - How have other users' opinions, product marketing or other sources of information about CAQDAS programs influenced my preferences? Are my assumptions about programs accurate? What more would I like to learn about particular programs?

Reflecting on these types of issues before attempting to select a qualitative data analysis program puts users in the position to critically evaluate for themselves how each program might integrate into their unique research programs. By examining detailed descriptions of program features and examples of how other users have used the software, users can develop their own sense of how various programs can be utilized. In addition to assisting them in the selection process, this reflective approach also reminds users that they are in control of the analysis process and that no technological tool, regardless of its features, can independently “perform” analysis. Through this type of program evaluation, both user empowerment and accountability are strengthened.

We examine the perils and promise of software packages for the analysis of qualitative data and offer specific future directions in the development and usage of software for qualitative researchers.

A trip through arts: Analyzing children dialogue using NVivo 7
Héctor del Castillo, Universidad Autónoma de Madrid, Spain; Ana Belén García Varela, Rut Martínez, Mirian Checa, Universidad de Alcalá, Spain

In this presentation we analyze an educational innovative practice, in which children participate working and thinking together in the analysis of paintings. Adopting an ethnographic, action research and socio-cultural approach in our study, we play the role of participant observers working in an extra-curricular workshop, sharing with teachers the task of introducing children to the analysis of arts. In this context, we explore how the software for qualitative data analysis NVivo7 can be used to facilitate the process of organizing, storing, retrieving and analyzing data in this research project.

The discussion of this paper focuses on the data analysis. In the process of analysis we used NVivo7, which allowed us to work directly with the transcription of video recordings in the discourse analysis. In this sense, the first step was to organize the data considering our data corpus, and then, we segmented the data in significant ‘sections’. But the main issue here was the unit of analysis. In a first phase, our unit of analysis was the classroom. Then, each document (as summary of the activity developed during one lesson - 4 paintings) transcribed and coded, was a unit of analysis. However, in a second phase, the unit of analysis was the painting, because there was what makes sense in order to analyse the children’s thinking in front of each painting. Then, the section turns into the unit of analysis: Every lesson (1 session) was segmented in four sections (1 section by painting) and coded and analysed as an independent unit.

Also, we added codes to these sections to group them in free nodes. This was a first approach to the categories that was relatively descriptive but provided an important infrastructure for later data retrieval and for the following analysis and interpretation. On the other hand, Nvivo7, helped us in a following step to create an analytical system of nodes for coding and interpreting the data. We took as starting point the methodology of the educational programme and we consider the dimensions they define as nodes for the analysis of children thinking.

In this process, our starting point is creating two different kinds of documents that are linked in the NVivo project: single ‘analysis’ documents, which are the transcriptions of every working session in the classroom, and ‘memo’ documents, which are the summaries of each researcher who participates in every session (could be more than one by session). These documents are linked by session, and it let us combine the narrative and analytical approach to the data analysis, considering the summaries (narrative interpretation) and the transcriptions as documents ‘to be coded’.

Finally, we use Boolean and Proximity searching modes to find coincidences and non-coincidences in the coding structure to analyse the different research questions above mentioned which are related to the discourse and the activity in the classroom. We created co-occurrence and difference matrix and start working with the text. So we use the matrix and the text for the analysis and interpretation of the data and, then, in the process of creating the evaluation report.

This presentation will address conceptual and methodological considerations related to qualitative mapping as an approachto data collection. Qualitative mapping is a discussion group technique designed to collect spatial data in the form of drawings and maps with supplemental textual data. Mapping activities have always sought to represent relationships between different levels and elements in space; however qualitative mapping also includes human emotions, culture, language, memory, and so on. Modeling these human relations and their multiple connections to space is not an easy task; the presentation focuses on how the visualization, coding, and annotating functions in MAXqda2 can be used to build a multifaceted representation of space and lived experience and link multiple types of data.

The presentation draws its examples from data collected on adolescent reproductive health behaviors in Asuncion, Paraguay in 2006 through small group discussions, qualitative mapping activities, and digital mapping activities. The presentation outlines the analytical tasks that were conducted using MAXqda2 software to develop a typology for adolescent romantic and sexual relationships that takes into account youths’ perceptions of safety in the community and supervision by parents and other adults. The approach worked to visualize discussion group data (text) from the perspective of spatiality and view visual geographic data (drawings and digital maps) in terms of youth’s discussions of their lived experience.

The discussion then turns to the challenges of integrating the potential for visualizing qualitative data that MAXqda2 and other CAQDAS applications offer with the representational, modeling and projecting capabilities of GIS, particularly in the context of the digital divide. It concludes that more work must be done in order to integrate CAQDAS and GIS and see advances in computing and qualitative geographic methods. Similarly, we consider the extent of methodological influence that CAQDAS may have on GIS applications in the realm of mixed methods, and the potential for smoothing over the fuzzy boundaries between Geography and Sociology.

Visualizing the future of QDA with Transana 2.20
Nicolas Sheon, PhD, University of California, San Francisco and David K. Woods, PhD, University of Wisconsin, Madison

Choosing a QDA package can be both daunting and expensive. Packages typically provide certain key features, but are often found lacking in others, such as support for multimedia data, cross-case analysis, or remote collaboration with other researchers. One feature hitherto missing from CAQDAS packages has been the ability to visualize broad patterns across an entire data set. Because most packages use textual fragments as the primary unit of analysis, qualitative researchers have focused their research at the cross-sectional and micro-level, leaving macro analysis of longitudinal data the purview of quantitative researchers. Unlike quantitative research, textual analysis does not lend itself to representing actual data using a chart or matrix.

Transana is a free, open-source software program for the transcription and analysis of large (or small) collections of video and audio data. This presentation will describe new data visualization tools recently unveiled in version 2.20. These new features came about through collaboration between a researcher, Nicolas Sheon, and the Transana developer, David Woods.

Transana’s Keyword Maps provide a bird’s eye perspective of the coding of multi-media data. Transana 2.20 extends the capabilities of the Keyword Maps by making the maps available as part of the main Transana interface, and adding the capacity for cross-case comparison of any number of media files within a single Matrix. The ability to visualize coding patterns over a series of media files makes Transana particularly useful for longitudinal evaluation of naturally-occurring data, such as clinical, educational or service encounters.

Figure 1: Series Keyword Sequence Map

The colored bar charts also provide an intuitive way to assess inter-coder reliability, compare the validity of different coding systems, as well as the co-occurrence, and density of codes. In the collaborative version of Transana, coding changes made by one user show up immediately and automatically in the maps of other users.

We will use data from the popular television series, Lost, to demonstrate the utility of Keyword Maps in providing an overview of large, complex data sets. Then we will show how we are using Transana for Sheon’s research on counseling interaction. This research combines conversation analysis and ethnographic interviews to triangulate the analysis of three sources of data: recordings of counseling sessions, follow-up interviews with counselors and clients, and focus groups.

The World as a Document: GeoCoding with ATLAS.ti
Thomas Muhr, Dipl.-Psych., Dipl.-Inform.CEO, ATLAS.ti GmbH

A fascinating new type of data is becoming readily accessible from our computer desktops: the world rendered from satellite and aerial images by specialized software such as Google Maps™ or Google Earth™. Virtual globe software permits virtual travels to any place anywhere, enriched, if needed, with multiple layers of auxiliary information.

Locative information is of increasing significance for all scientific, governmental and commercial fields concerned with data including or associated with locations or areas (some examples would be: Anthropology, Archaeology, Historical Sciences, Criminology, Customer Relations, Forestry, Geography, Tourism, Ethnography, Epidemiology, Urban Planning, Real Estate, Logistics, and many more).

The ability to code, memo and hyper-link physical locations to text passages, video sequences and image sections provide additional insights into the data.

The upcoming next major release of ATLAS.ti will offer seamless integration with Google Earth for viewing the world as a document, treating geographical locations as “quotations.” ATLAS.ti’s new GeoCoding facility combines a powerful vehicle to navigate the world with the semantics of a proven analysis framework.

The paper aims at three objectives: demonstrate this powerful feature in action, discuss its implication for the work of researchers in various fields, and suggest the necessity of transcending a traditional notion of "qualitative data" as merely a manifest set of physical documents.

Research results showed by a video and by a website
Dario Da Re, Department of Sociology, University of Padova , Italy

This paper will present research methods and techniques applied in a project on the c onditions, needs and expectations of rural pensioners in Veneto Region ( Italy ). Although all the phases of the research project were structured and realized using many different software and in general using many digital technologies, I prefer to focus this presentation on two unusual approaches to the analysis of results.

Normally the representation of data resulting from a study is reported in a paper or in a publication. In this case both a dvd video and a website have been implemented.

The website (www.raccontiditerra.it) aims to make the results of this study available to the widest possible audience and the video seeks to make these results available to a part of the audience that would never dream of reading any written report or of opening a browser on the Internet.

This latter group refers particularly to the rural pensioners themselves, the fulcrum of our study. And even though the video is as methodologically rigorous and valid as any written report could be, it seemed to be the best way of giving back the results of the study to those people who gave the precious information used in the first place.

The website has been built using a CMS (Content Management System) that facilitates the organization, control, and publication of documents (text) and other content, such as images (photos) and multimedia (video) resources. The main advantage is the possibility of publishing information in real time without any knowledge of Internet programming languages.

Various technologies were used to create the video documentary. For recording: digital video camera, digital photo camera and digital audio recorders and for editing: Final Cut software on Mac platform. But the most innovative aspect of the video is the use of Atlas Ti software, not only to analyse the data contained in the filmed materials but also to choose which clips to place in the final video report.

Atlas Ti was using to select 300 quotations out of the 900 that had already been used for qualitative analysis. In short, out of the 45 hours total of filmed material, 300 clips, lasting about 2 hours were selected. These clips varied from 10 seconds to one and a half minutes in length. These 300 were then sorted again, this time on the basis of technical features, such as poor light, audio fuzzy, unclear etc. In the end 191 clips remained and editing began.

www.raccontiditerra.it (available in English)

Virtual Qualitative Data
Ian Jones, Institute of Education, University of Warwick

I report on a use of technology to support the design and analysis of interventionist experiments into the learning of mathematics. Interventionist experiments involve developing novel educational tasks and tools and assessing the conceptual changes that arise when learners interact with them. The tasks and tools are trialed in messy, real-world settings such as classrooms and iterative cycles of systematic design play an analogous role to the systemic control of variables in clinical experiments. Interventionist approaches seek to bridge the gap between educational theory and practice by drawing on pragmatic philosophy and have been likened to the iterative engineering practices found within the aerospace industry (diSessa and Cobb, 2004).

An intervention comprises cyclic phases of development and formative analysis. The development phases involve generating trial specifications by imagining tasks and tools being manipulated by typical learners; the analysis of the subsequent (movie) trial data then informs the next cycle. My focus here is on the use of Microsoft PowerPoint as a storyboarding tool for generating, communicating and selecting trial designs. I have found this technique particularly valuable for resolving the tension between learner-centric and techno-centric concerns in the design of software-based educational tools. The essential idea is that the development of tasks and software tools is expressed as conjectured learning trajectories rather than as technical schematics. Such conjectures are represented as annotated storyboards in which learners are imagined manipulating hypothetical software tools towards a particular goal.

I have found this PowerPoint storyboarding technique to be invaluable during the early iterations of an intervention when initial ideas are getting off the ground. Simplified graphical images of resources can be created and then, using the copy and paste tools, new slides cloned and small stepwise adjustments made to the graphics. In this way ideas can be tried out quickly by considering how typical learners might interact with the resources. When design issues and problems are encountered an earlier slide can be returned to, copy-pasted anew, and an alternative possible trajectory of events considered. The “click to add notes” facility of PowerPoint allows slides to be annotated as appropriate. These annotations can be added retrospectively and date-stamped to build up a time-line of evolving ideas, theories and hypotheses.

Storyboards might be considered as ‘virtual data’ of thought experiments and videoed trials as ‘actual data’ of real-world interventions. A key achievement of this approach is that design decisions made during development phases, which conventionally are opaque in nature, can be made available for scrutiny and discussion with other designers. Furthermore, the representational similarity of storyboards and movies has helped me to blur the distinction between development and analysis phases, and between technical and educational design tensions. This in turn has afforded a more holistic approach to the design of educational interventions.

diSessa, A. and Cobb, P. (2004). Ontological Innovation and the Role of Theory in Design Experiments. The Journal of the Learning Sciences, 13(1), 77-103.

Using Computer-Assisted Qualitative Data Analysis Software for Visual Sociology
Paolo Parmeggiani, University of Udine, Italy

Visual sociologists employ images to analyse society and culture. They often use methods such as Grounded Theory [ Suchar S. 1997 ] or photo elicitation [Collier and Collier 1986]. Unfortunately, gathering, analysing and presenting visual and audio data with CAQDAS is often time-consuming. To transcribe time-based media, their integration with still images and text codes are complex. Commercial CAQDAS is often expensive.

I partly disagree with these opinions. The first hypothesis is that sociological photo elicitation research, once done with traditional tools, can now be done more effectively with IT. The second hypothesis is that it isn't necessary to rely on only one commercial package; a combination of freeware and open source software could lead to a satisfying practice.

I describe the methodologies, techniques and merged use of different hardware and software to help a visual sociology's project, which was to analyse social changes and meanings of landscape images in a little Italian village. I used a digital camera, a laptop and software which supported the collection to display and analyse still images and audio data. I choose freeware software, which helped me to design the research and map the ideas, open the coding process, record photo elicitation interviews, transcribe and analyse data and finally present the results.

Instead of photographing and printing "analogically", I made thousands of digital shots. This allows the pictures to be immediately available for observation.

A “zoomable” image browser allows for an "overview first, zoom and filter, then details-on-demand" [Shneiderman 1993] and allows for dragging and dropping notes and label placements on photos. This procedure is more suited for open and axial coding [Strauss 1998] compared with a traditional database. The software Photo Mesa uses the methodology of semantic regions, which provides a formal model of spatial and dynamic labelling and reorganization of photo data based on users' mental categories (following grounded theory method). This helped in observation, sorting by category, revising definitions of categories and selecting convenient images for interviews.

Instead of showing pictures printed on paper and video-recording the photo elicitation interview, a software (Camstudio) replaced the work of a TV camera recording the slide show screen and the comments of people. The use of laptop with this software is less invasive compared to a traditional TV camera and a mike. The screen recorder software speeds the recording and digitalizing stages; it therefore guaranties the synchronization with voices. It makes possible to record also the interviewed mouse's moves according his point outs.

A CAQDAS (Transana), mainly designed for transcription and analysis of video data, was conveniently used for analysing interviews and still images (already synchronized by screen recorder).

I found out that is useful to consider the methods, the hardware, the software and the practices like a compound digital environment where different tasks could be performed in a personal integration. In this way visual sociologists can use freeware software in a way that could compete and even lead better results (faster, cheaper and more visual) than a single commercial CAQDAS.

And we’re rolling! VIA as a case for better software management of qualitative multimedia data
Jennifer Patashnick, Video Intervention/Prevention Assessment (VIA), Children’s Hospital Boston, USA
Download Slides

Video Intervention/Prevention Assessment (VIA) provides participants tools to document their lives as they experience them. Since 1994, young people at Children’s Hospital Boston have generated thousands of videotapes to teach their clinicians about living with a chronic health issue. Our challenge: review and analyze each video, eliciting themes affecting participants’ everyday lives with which clinicians can improve healthcare communication and delivery. Each recorded moment must be viewed, coded, and compared to the rest of the individual’s data and across all participants.

At first, the VIA research team logged videos manually. Adding trained interns increased the amount we logged, but the knowledge was distributed across many individuals and hundreds of log pages. Synthesizing themes without the full team present was difficult. We soon realized word-processing software would allow us to use multiple loggers, but maintain a single, searchable record. We searched for keywords, but when loggers used different vocabulary for similar ideas or if an idea was generalized as opposed to point-specific, searches were less effective and connections could not be made.

Sophisticated qualitative analysis software enabled more thorough exploration of logs. Instead of isolated word searches, we coded text to inductive and deductive themes. Memos and annotations were maintained within the project frame. Collaboration and coding, synchronously and asynchronously, were achieved through active team management, and we are overcoming our reliance on analog video data: particularly revealing video is digitized and referenced directly from the electronic logs.

However, audiovisual analysis remains suboptimal. Most qualitative analysis packages claiming to incorporate video data relegate it to a peripheral role, with text remaining central. Researchers sometimes complain that analysis software takes them away from their data. Given textual data, this is arguable; with video, no such argument exists. No programs offer the ability to code with breadth, depth, and flexibility directly with video. All require transcription, a process which constitutes a requisite analytic “pass”; future analysis becomes a “generation” removed. Video is more complex than text; every study collecting audiovisual data has developed a different method of analysis, with diverse aims and constraints. Software that can handle the volume of video VIA participants produce is rare. In part, the challenge of direct analysis of VIA audiovisual data lies in our examination of macro-behavioral themes through grounded theory-based analysis. We code narrowly, investigating the behavior itself, and broadly, examining the context of that behavior and the physical, psychosocial, and historical contexts in which it occurs, to create a database of coded segments that can be analyzed in response to specific research questions.

While qualitative analysis software has progressed dramatically for textual analyses, it has light-years to go before it adequately manages multimedia data. At minimum, video needs to replace text as the data to be structured, analyzed, coded and recoded in response to specific avenues of inquiry. Particularly in research where the participant collaborates in collecting data, analytic software must acknowledge the perspective of the “gaze,” as well as both objective and subjective content to be analyzed. Most importantly, software must provide time-efficient ways to manage, display, and analyze audiovisual data in formats that can be considered and conceptually structured by researchers.

The National Centre for e-Social Science Research Node DReSS - www.ncess.ac.uk/research/nodes/DigitalRecord/ - is exploring the development of digital records to support qualitative study of life in the digital age. Digital records consist of 2 core components: 1) resources gathered by a field worker (video and audio recordings, photographs, field notes, etc.), and 2) resources internal to computational environments (such as text messages, voicemail, email, etc.). Digital records therefore enable qualitative researchers to combine visual data – especially video – with diverse resources to support qualitative study of life in the digital age.

The combination of visual data with diverse resources is supported through a suite of software tools called the Digital Replay System (DRS). The DRS enables researchers to synchronize visual data with diverse resources and replay them side-by-side. Thus, and for example, a video recording detailing someone’s computer-mediated interactions may be combined with a system-based recording that makes the digital medium (e.g., instant messaging) and content of interaction (e.g., text) available for inspection too. The different recordings may then be synchronized and replayed side-by-side to provide a comprehensive view of social interaction across physical and digital environments.

The DRS enables researchers to record resources from multiple digital media, including mobile location-based applications (e.g., GPS–enabled phones equipped with location services and audio, visual, and textual resources) and supports the study of asymmetrical forms of communication and interaction where, for example, people interact via audio messaging on the one hand and text messaging on the other. Uniquely, it enables the researcher to bring such resources together for the first time.

The DRS also supports analysis of the diverse contents of digital records. It enables the researcher to assemble and structure the contents of digital records to develop their analysis of social interaction in the digital age. Annotations may be added and the digital record may be coded to support the production of distinct analytic insights. Exploiting semantic web ontologies, the DRS enables the qualitative researcher to exploit pre-existing coding schemes or to develop completely new ones from the ground up.

The emergence of digital records and the DRS takes qualitative research beyond the possibilities of ‘hypermedia’(1), to capture and present resources never available before and combine them with visual data. Together digital records and the DRS provide unprecedented access to interaction in the digital age and open up the possibility for qualitative researchers to develop new insights into the nature of social interaction in contemporary society.

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

Creating visual transcripts: the role of computer software in the analysis of video data
Marie Joubert Gibbs, Graduate School of Education, University of Bristol

Video data is notoriously difficult to transcribe because of the richness of the data, (Rose 2001), which includes not only speech, but also other audio, eye gaze, gesture and action. However, most analytical techniques require some form of transcripts to work from. This paper reports on an approach to creating such transcripts and explores questions relating to the potential and role of the software and the mediational effects of the software on the process of analysis.

The data is taken from a PhD study concerning the use of computers in mathematics classrooms, in which students worked in pairs or small groups at computers. The question at the heart of the study was to investigate the mathematical learning of the students in terms of the mathematical processes in which they engaged; the emphasis therefore was on process as opposed to product. Video was used to capture these processes, together with screen grabbing software to capture on-screen activity.

The situated action perspective framing the study uses the interactions between people and the setting as the unit of analysis, and pays attention to the ‘flux of ongoing activity’ and to the ‘unfolding of real activity in a real setting’ (Nardi 1996), and action is seen as ‘an emergent property of moment-by-moment interactions’ (Suchman 1987), implying that the inquiry should take place at a very fine-grained level of minutely observed activities. Further, each interaction is situated in time, and related to all the other interactions taking place as the students work.

The first implication of this point of view is that, to make sense of interactions, it is necessary to understand the unfolding of all these interactions; or to ‘tell the story’ of all the interactions. The second implication is that, to understand the story, an appreciation of the situational variables surrounding the interactions is required; to give them context and provide them with meaning.

This paper describes how computer software was used first to code the data in terms of the students’ mathematical learning and then to create a visual transcript of the data which represents the coded student interactions over time, thus creating a visual narrative of the lesson. The transcript also includes situational variables such as teacher interventions and on-screen computer activity. Woven into the description is an analysis of the ways in which the potential and limitations of the software mediated the process of analysis and led to the construction of the visual transcriptions.

Nardi, B. A. (1996). Studying Context: A Comparison of Activity Theory, Situated Action Models, and Distributed Cognition. Context and Consciousness: Activity Theory and Human-Computer Interaction: 69-102.

Rose, D. G. (2001). Visual Methodologies: An Introduction to the Interpretation of Visual Materials, Sage Publications.

Suchman, L. A. (1987). Plans and Situated Actions: The Problem of Human-Machine Communication. Cambridge, Cambridge University Press.

Software used
For coding the data and creating the timeline: Studiocode http://www.studiocodegroup.com/
For creating the visual transcript: Omnigraffle http://www.omnigroup.com/applications/omnigraffle/

Analysing an educational project in the classroom using Transana: Children and parents learning together
Ana Belén García Varela, Universidad de Alcalá, Spain, Héctor del Castillo Universidad Autónoma de Madrid, Spain, Pilar Lacasa & Sara Cortés Universidad de Alcalá, Spain

In this presentation we analyse how we can build bridges between family and school that will enable us to make school learning a significant experience, bearing in mind that in both settings children acquire symbolic instruments that concretise human relationships and facilitate the transformation of the physical world into a set of intellectual tools. We show how we use Transana in this research process from the first steps of the data analysis.

In this study we adopted an ethnographic perspective (Atkinson, Coffey, Delamont, Lofland & Lofland, 2001, Castanheira, Crawford, Dixon & Green, 2001), which helped us to understand school as a cultural or institutional context in which the activity of individuals acquires meaning. In this context, we adopted a case study methodology which enabled us to go more deeply into the analysis of the activity developed by one of the families (Yin, 2003; Stake, 2006). From this approach, we analyse how children who participated in the workshop with their families progressively constructed a community of learners.

In this presentation we take some examples from a large research project that involves children in primary education and their families. We were working in a multimedia workshop that took place in a Spanish public school where children (6 to 12 years old) and their families work together in the construction of a digital newspaper of the school to be published on the Internet (http://es.geocities.com/web_lahuella). In this context, we worked for a total of ten two-hour sessions, in which six families participated, as well as the researchers themselves (as participant observers in all the sessions).

We used a number of instruments to register and analyze the information: audio recordings, pictures, diverse materials handled and generated by the participants in every session, field notes, summaries of the sessions, etc. All these materials were stored, organized, transcribed, constructed and analyzed using Transana.

We focus on some of the conversations that take place during the workshop, as a good example of how children and adults learned together to write an article for a digital newspaper in the context of a community of learners. More specifically, we wish to show the type of strategies of support which a mother used to help her children to elaborate a narrative using contents related to their everyday lives.

The family that we analyze consists of a mother and her two children (11-year-old boy and 6-year-old girl). None of them had much knowledge of computers when they came to the workshop; the children sometimes used the computer at school, and at home they used it to paint. The girl was learning to read during this academic year . The mother did not use the computer much and used it only as a typewriter, and, sometimes, to look for information in a digital encyclopaedia. In spite of this, they felt very motivated to work on the newspaper, and they had a lot of interest in learning to use new technologies and many ideas that they wanted to share in the production of the newspaper.

Atkinson, P., Coffey, A., Delamont, S., Lofland, J., & Lofland, L. (Eds.). (2001). Handbook of Ethnography. London: Sage.

Castanheira, M.L., Crawford, T., Dixon, C., & Green, J. (2001). Interactional ethnography: an approach to studying the social construction of literate practices. Linguistics & Education, 11(4), 353-400.

Stake, R.E. (2006). Multiple case study analysis. New York: Guilford Press

Yin, R.K. (2003). Case Study Research Design and Methods. Thousand Oaks: SAGE Publications.

Using Atlas.ti for narrative research on informal learning
Giancarlo Gola, Dept of Education, University of Trieste

The paper draws on work done for a Doctorate of Research which is still in progress. The objective of the research is to examine the experiences in professional practice of social workers and, in particular, to explore their informal learning at work (work-based learning, learning on the job, experiential learning, evidence-based learning). The research is based on the paradigm of narrative inquiry to explore, into the knowledge, action and subjective reconstruction of experience (Bruner 1988, 1992, Polkinghorne 1988, 1996, Clandinin J.D., Connelly M.F., 1980, 2000, Taylor C., 2003).

The tool for the data collection has been the narrative interview (Atkinson R., Bertaux D., Riessman, C.K., Bichi R., Mantovani S.), through a predisposed trace both to maintain a rigorousness in the method of harvest of the story-life, both with the purpose to allow a comparative analysis of the qualitative data. It founds him upon two initial generative questions, (Flick 1998, Bichi, 2002), that are characterized for a different cognitive orientation, the first trace contemplated to the acquisition of a story of professional life, the second mostly set the attention in comparison to the reconstruction and reflection on a story of learning, understood as event on which the person feels to be able to tell a situation that has left a sign, has produced a chance of learning in the work. The narrative interviews at 30 social workers in different position in the job : c hildren's services, elderly care, mental health, young offenders, drugs and alcohol, adults & children with disabilities , have been recorded and transcribed in electronic format.

The analysis of the narrative contents used the most recent version of Scientific Software's ATLAS.ti (5.0). The analysis of the narrative contents focussed on narrative themes connected with learning and experience. This was undertaken by segmenting the texts into simpler units then underlining their meaningful contents and examining their mutual interrelations (Cicognani, 2002). During this analysis, the software ATLAS.ti, helped in exploring the complex meanings hidden inside the retrieved text, assisted the researcher in all the phases of the job and allowed a general view of the task by rapid search and recovery of the material being analysed. In particular, the tool has allowed and facilitated the following phases of the process of textual analysis: segmentation of the oral transcripts of the "interview-biographies" into meaningful quotations for comparison to the object of our investigation; coding of the quotations through the assignment to each of a principal code indicating the content of the quotation; construction of networks that graphically illustrate the relationships among the codes, among the families or among other data considered meaningful.

The coding of the narrative interviews has been effected rereading the texts individualizes in the narrative discourse of every interview a nucleus (core), (Reissman, 1993) and introducing an aggregation "for families" with software ATLAS.ti of the nucleuses, it's reported to the forms of the informal learning, informal practice-based workplace learning: deliberative learning, reactive learning, implicit learning, incidental learning, (Eraut, 2000, Marsick and Watkins, 1990, , Cheetham and Chivers, 2001).

Use of the ATLAS.ti in the process of analysis has allowed connection of the elements of the narration (events and actions) with the formal characteristics of the text (the used language and the textual structure) and the extra-textual elements (context of production of the narrations). The treatment of the qualitative data through the software has allowed to recognize some data and to facilitate comparative analysis on the elements of interest for research, the elaboration of a network of correlations between all the elements analyzed of every interview, and to check of the relationships on the narrative data elements with the Query Tools of Atlas.ti, has allowed particularly to specify the characteristics of the informal learning of social workers, the contents of informal learning in the workplace, and the implicit learning in social-care organizations.

Ph.D. Student in “Sociology, Social and Education Science” Department of Education - University of Trieste - Italy (2005-2007).

Enhancing quality and transparency : the role of Framework in supporting robust qualitative research
Kandy Woodfield & William O'Connor , National Centre for Social Research

Qualitative researchers need intelligent tools that can assist them in understanding large bodies of unwieldy data. This paper will introduce a pioneering new approach to computer assisted qualitative data analysis called ‘ Framework' due for release in 2007. Developed by the Qualitative Research Unit at the National Centre for Social Research in the mid 1980's ‘Framework' is a matrix based tool for qualitative data management.

The development of this software version of the tool has allowed the team to consider the ways in which a new generation of software tools might help to enhance the quality and transparency of the qualitative analysis process.

he presenter(s) will introduce the package to the audience and discuss various features of the new software that can assist researchers in producing robust, credible qualitative evidence for social research. In doing so they will also explore wider issues relating to quality in qualitative research and how far the use of software packages in the analytical process can support and enhance transparency.

‘Membership Matters’: How and Why Qualitative Researchers Should Consider Applying Membership Categorisation Analysis To Their Data Using CAQDAS
Andrew King, Department of Sociology, University of Surrey

Membership Categorisation Analysis (MCA) is a form of conversation analysis that has its origins in the work of Harvey Sacks. It explores how individuals make sense of and order their social worlds, particularly how they constitute their own and others identities. Recently, this somewhat neglected area of Sacks’ work has experienced a resurgence of interest. MCA has been applied to the analysis of gender, crime, organisational structures and stigmatised identities (Eglin and Hester 2003; Horton-Salway 2004; Housley and Fitzgerald 2002; Stokoe 2003a; 2003b). However, it remains a somewhat esoteric methodology that is not widely recognised amongst qualitative researchers in general, especially those who use Computer-Assisted Qualitative Data AnalysiS (CAQDAS) software. Furthermore, it can be misinterpreted, and therefore rejected, as a methodology that is purely descriptive.

In this paper I argue that MCA is neither esoteric nor purely descriptive. Indeed, my aim in this paper is threefold. Firstly, I will introduce the key concepts of this methodology and note recent critical extensions. Secondly, I will explain why it is a pertinent methodological approach that, although not ‘interpretive’ in the traditional sense used by many qualitative researchers, is nevertheless ‘critical’. Thirdly, I will demonstrate how MCA can be applied to interview data using NVivo software by outlining two coding strategies that I have employed to analyse qualitative interview data. These strategies are “pinpoint coding” and “collection coding”. The former makes use of NVivo’s text searching tools in order to locate key categories and attributes. The latter uses a more methodical, and perhaps more orthodox approach, to reading and coding interview transcripts in order to identify key collections. However, I will explain that when these two coding strategies are combined they provide a valuable procedure for undertaking a MCA. I thereby conclude the paper by arguing that MCA is an important methodology for undertaking qualitative data analysis; a methodology that those interested in using NVivo, in particular, and CAQDAS, more generally, should consider utilising and extending.

References

Eglin, P. and Hester, S. 2003 The Montreal Massacre: A Story of Membership Categorization Analysis, Waterloo, Ont.: Wilfrid Laurier University Press.

Horton-Salway, M. 2004 'The Local Production of Knowledge: Disease Labels, Identities and Category Entitlements in ME Support Group Talk', Health 8(3): 351-371.

Housley, W. and Fitzgerald, R. 2002 'The Reconsidered Model of Membership Categorization Analysis', Qualitative Research 2(1): 59-83.

Stokoe, E. H. 2003a 'Doing Gender, Doing Categorisation: Recent Developments in Language and Gender Research', International Sociolinguistics 2(1): 1-12.

— 2003b 'Mothers, Single Women and Sluts: Gender, Morality and Membership Categorization in Neighbour Disputes', Feminism & Psychology 13(3): 317-344.

On the Use of ATLAS.ti for the Reconstruction of Cognitive Maps from Coded Texts
Georg P. Mueller University of Fribourg, Switzerland

One of the special features of the ATLAS.ti software is the graphical representation of content analysed texts in so called network views, where text segments (quotations) and user defined codes are displayed as nodes and the attribute relations between the mentioned quotations and codes are represented by graphical links (see fig. 1, upper part). Unfortunately, network views become complex and puzzling when in an empirical project the number of text segments and codes increases. Thus, this paper proposes a methodology of logical abstraction and synthesis of network views of coded texts. This is especially useful for researchers interested in theory construction, e.g. with the methodology of Strauss & Corbin 1), where theoretical saturation is an important goal.

Fig. 1: An exemplary transformation of a network view of a coded text into a cognitive map.

The method of this paper is based on the idea that the analysed text segments as well as their coding are inferences from the cognitive map of the author of the text, which can be represented by a relatively abstract semantic network of logical propositions, conceptualised in the intellectual tradition of Norman & Rumelhart 2). This network of propositions is composed of codes,negations of codes, and implication operators from Boolean logic, which function as links between the first two types of elements (see fig. 1, lower part). Hence the goal of this methodology is the reconstruction of an abstract cognitive map from the original network view of a coded text. The technique proposed for achieving this goal is based on the replacement of the text segments and attribute relations by appropriate Boolean implication operators, which define new links between the codes of the original network view (see fig. 1).

This transformation bears the risk that the resulting cognitive map is either contradictory or incomplete with regard to its logical implications. Hence the paper also discusses heuristics for avoiding these risks. Contradictions in cognitive maps can e.g. be escaped by framing or contextualising the logically incompatible parts of a cognitive map. Similarly, missing codes in the original network view of a coded text can be substituted by appropriate negations of codes, such that the derived cognitive map becomes more complex and consequently also more complete. These and a few other heuristic principles will be illustrated by a content analysis of a Swiss cooking book, which describes the typical dishes form the different regions of the country.

As a matter of course, the mentioned heuristic principles do not guarantee that the reconstructed cognitive map is really complete and reproduces all codings of the analysed text. This obviously points to some limits of the proposed methodology. It has however the advantage that apart from its instrumentality for classical content analysis, it can also be used for studying transcripts of qualitative interviews. In this case, there is even the possibility to discuss the reconstructed map with the interviewees and to correct misinterpretations of the researcher.

1) A. Strauss & J. Corbin (1998). Basics of Qualitative Research. Techniques and Procedures for Developing Grounded Theory. Thousand Oaks: Sage Publications.

2) D. Norman & D. Rumelhart (1975). Explorations in Cognition. San Francisco: Freeman and Company.

CAQDAS, Context and the Recontextualisation of Qualitative Data
Libby Bishop, ESDS Qualidata, University of Essex, UK

This paper explores to what extent use of CAQDAS enables, hinders, or does not affect the process of recontextualising qualitative data. The paper explores issues that arise when using Atlas.ti to support secondary analysis of qualitative data. It has been argued elsewhere that secondary analysis should not be seen as distinct from primary analysis, but rather secondary analysis is a form of recontextualisation of new data. When using CAQDAS packages, context is typically considered at the level of the text. Those emphasising the integrity of an interview—its narrative flow—are rightly concerned that technology can make it too easy to strip a word or phrase free from its surrounding contextual signifiers in a sentence or paragraph. But “context” can and should be considered more broadly. There are multiple layers of context. There are two points in time at which context needs to be considered: the period when the original project was done, and the period for the contemporary project. In each period, there are at least three levels of context: the interaction, the situation, and the cultural. This paper uses data from two archived collections to explore the role of CAQDAS in the multi-faceted process of recontextualising qualitative data.

The Use of CAQDAS in the UK Market Research Industry
Ruth Rettie, Helen Robinson, Anja Radke, Xiajiao Ye, Kingston University, UK

This paper explores the use of CAQDAS in commercial market research in two ways. Firstly, we report a quantitative survey of CAQDAS usage in the UK market research industry and secondly, we describe a case study, which explored the use of CAQDAS to supplement a more traditional manual analysis.

A questionnaire was sent to a sample of 400 UK market researchers exploring awareness, usage and attitudes to CAQDAS. The survey response rate was reasonable at 38%, but reported usage of CAQDAS was very low at 9%. There was, however, significant awareness of both CAQDAS and of the various software packages, with Nudist as the brand with the highest awareness. This is the first large scale survey of CAQDAS usage in UK market research, although the findings are consistent with the much smaller survey conducted by Nancarrow, Moskin and Moskar (1996) ten years ago, and with a survey conducted in Ireland by Bezborodova and Bennett (2004) .

The low usage of CAQDAS within market research is usually explained in terms of tight deadlines, often within days of the interviews, which may preclude the production of full transcripts and computer coding (Ereaut, Imms, & Callingham, 2002) . We therefore explored an alternative approach, which utilized CAQDAS in a supplementary analysis. The group moderator produced an initial report and debrief using paper-coding, and we followed this with a second analysis based on CAQDAS coding of the transcripts.

We report a case study based on focus group research into online grocery shopping. Four 1½ hour focus groups were conducted by an experienced moderator of commercial market research. The analysis was in two stages. Firstly, the group moderator analysed the transcripts using traditional paper coding, producing a research report and presentation. This was followed by an analysis of the transcripts by a second analyst using Qualrus software. Using Qualrus we were able to 1) mine the data for more detail;
2) clearly identify minority views; 3) and produce a useful resource for future research.

Repositioning CAQDAS as supplementary, rather than as an alternative, circumvents arguments about time pressure, and highlights its role as a long-term resource. We conclude that promotion of CAQDAS as a supplement to, rather than as a replacement for, more traditional analysis would increase its usage within the market research industry.

Bezborodova, B. & Bennett, B., (2004), An investigation into the non-usage of qualitative software in the marketing research industry, Irish Marketing Review, 17, 1 and 2, 1-5.

Ereaut, G., Imms, M., & Callingham, M., (2002), Qualitative Market Research: Principle & Practice, (vol. Four) Sage Publications.

Nancarrow, C., Moskin, A., & Shankar, A., (1996), Bridging the great divide - the transfer of techniques, Marketing Intelligence & Planning, 14, 6, 27-37.

The contributions of QDA software to mixed methods research: a mixed success?
Normand Peladeau, Provalis Research

Despite the growing popularity of mixed methods research and the exponential development of new text mining and information retrieval techniques in the past two decades, one may consider the evolution of existing QDA tools today to be somewhat conservative. As a consequence, researchers often bear most of the burdens of performing this integration of various techniques. While some philosophers and researchers may have doubt about the feasibility or even the desirability of such an integration of quantitative and qualitative methods, we will take a more pragmatic approach and attempt to identify the potential benefits of further integrating a variety of text analysis and statistical techniques in QDA software. We will present some of the principles and ideas that were at the origin of the design of QDA Miner, Provalis Research's qualitative data analysis tool, and show you three areas in which such an integration takes place: 1) the structure of projects and data files, 2) the inclusion of graphical and statistical tools for exploratory data analysis of coded data, and 3) the tight integration with an existing content analysis and text mining module (WordStat). We will also propose a quick look towards the future and identify potentially new ways to combine techniques for diverse areas in order to provide further benefits both for qualitative and mixed-method researchers. We will likely use this occasion to offer to the audience a glimpse of the new features in the upcoming version 3.0 of QDA Miner.

Gaining depth without drowning: How Framework can aid in the managementand interpretation of qualitative research data
William O'Connor & Kandy Woodfield, National Centre for Social Research

Working with large amounts of textual data continues to be a challenge for qualitative researchers, despite the establishment of sophisticated software packages for the analysis of textual data. Despite the ‘analytic support' (Coffey and Atkinson, 1996) that such tools can provide, analysts are still in danger of being overwhelmed by the task at hand. This can result in an analytical process that is confined to the descriptive whilst never really reaching the required level of depth or sophistication needed to understand or explain the complexities of the social world.

This paper will discuss how Framework can be used to aid data management and interpretation of qualitative textual data. Framework is a long standing tool used by qualitative researchers in all sorts of research environments. It is soon to be released as a windows based software package.

The paper will focus on two valuable aspects of the package that can aid the analyst on their analytical journey. The first of these is ‘data summarisation', a process by which verbatim data is synthesised and located within a thematic framework while retaining a direct link back to the original source. The authors will argue that this can aid the user in data reduction and, more importantly, in beginning to make sense of the data. The second is the software's facility to construct user defined matrices of summarised data, by theme and by case. These matrices, it will be argued, are an important way of injecting much needed perspective into the analytical task. They enable an analyst to see clearly how different themes map across a dataset whilst simultaneously detecting linkages between themes at the individual case level. In so doing, they are an important support to the interpretive process.

Coding for Human Language Tool Builders
Stuart W Shulman , University of Pittsburgh

Computer scientists in the eRulemaking Research Group focus on text clustering, text searching, near-duplicate detection, opinion identification, stakeholder characterization, and extractive summarization. Social scientists in the group are studying the impact of such tools and the Internet more generally on the process of rulemaking in the United States . Over the last five years, our group has collected 16 public comment datasets comprising in excess of 1,000,000 public comments on federal regulatory actions. Our research explores the use of information extraction and information retrieval to assist rule writers and analysts in managing large volume public comment periods. Information extraction techniques strip off email headers, salutations, signature lines, and advertising text. Text clustering algorithms identify exact duplicates, group together comments that are similar but not identical, and organize them hierarchically for browsing by rule-writers. Text-differencing algorithms identify where a person has edited a form letter so that a rule-writer's attention is drawn immediately to the unique part of an edited form letter. This talk will focus on the role of manual coding and new tools we have built, and are building, for refining and reporting the work of independent coders in relation to the development of these new human technologies.

Using Caqdas on Historical documents; some issues from the experience of the Health of the Cecils Project
Caroline Bowden University of London
Download Full Paper

This paper will consider the impact of using QDA software (in this case NVivo 2) on historical research methods by focussing on one group of sources used in the project; that is manuscripts in State Papers held in the National Archives at Kew. The purpose of the project was to examine the experience of health care in the Cecil family between 1550 and 1660 using the extensive manuscript collections surviving relating to the family in this period.

At the outset it was clear that the research would focus on studying attitudes and experience which is often the purpose of sociological research. This suggested that software designed originally for social and political scientists could be appropriate for use in this project. However since at the time of starting the project no support literature had been written with historians in mind, the methodology of the project and the resulting organisation of the documents in the software were founded on experimentation at a basic level. While this was often challenging it did lead to new insights.

As the project concludes, the paper will consider the impact of QDA software on the Health of the Cecils project by focusing on two issues: the analysis of the documents and on the organisation of the research process.

Use and comparison of 2 software programs (ALCESTE / PROSPERO) in social representations’ analysis
Béatrice Madiot Université de Picardie- Jules Verne (ECCHAT), EHESS (GRSP et CETSAH-LPS), Magda Dargentas, Ecole des Hautes Etudes en Sciences Sociales (CETSAH-LPS)

In qualitative research in social psychology in France the use of CAQDAS programs (e.g., NVivo, Atlas-Ti, MAXQDA) is limited. This does not mean that researchers do not employ programs to analyse qualitative data. In fact, three types of software programs are used: ALCESTE, PROSPERO, TROPES.

In this paper we present the differences between CAQDAS and these three software programs. Their main difference is their aim. If the first one help the researchers to carry a classic content analysis, the other ones offer more technical possibilities to analyse textual data. They are also based on specific theories of language and possibly on specific sociological traditions. We focus on the use of ALCESTE and PROSPERO, in order to analyse a corpus on hygiene, within the framework of secondary analysis.

Secondly, we present the theoretical framework we used to analyse our data: social representations (Moscovici, 1984; Jodelet, 1989). Social representations study knowledge of common sense. In other words, this theory analyses how members of a group/culture conceive an object and how these conceptualisations vary in different groups or contexts. It is an alternative to the cognitive social psychology and to positivist methods of research (Farr, 1987; Hepburn, 2003). This theory uses findings and approaches from various disciplines: history, sociology, ethnology. It carries a holistic approach of a social phenomenon viewed in its complexity (Markova, 1999). Because of these characteristics, qualitative research is fundamental in this approach. Consequently, the use of software programs becomes precious, especially in the case of large corpus.

Our corpus is composed of 37 non-directive/ in-depth interviews. 16 are part of a primary study conducted in 1996. The 21 have been realized in 2006-2007. Both cases concerned collective research.

According to the definition of secondary analysis and Heaton’s typology (1998), we are doing a secondary analysis with multiple qualitative data sets (with 1 primary data set), with additional in depth analysis around a dimension of analysis (interpersonal issues), and the use of new methods of data analysis.

Thirdly, we present our findings following three axes: a description of the contents, a comparison of data sets for the year of collection, presentation of specific contributions of the programs.

We discuss validity of qualitative analysis through triangulation methods. The use of several software programs strengthens the validity of our analysis. In fact, results are acquired through different methods focusing on diverse topics and strategies of analysis. The validity of research is also enhanced through secondary analysis, by comparing an actual data set with an older archived data set. We finally discuss the contribution of secondary analysis combined with the use of software programs to the theory of social representations, regarding the study of social change in the course of time.

CAQDAS: A ‘posthuman’ research experience?
Jo Haynes, Department of Sociology, University of Bristol

The relationship between human beings and new technologies is often characterised by anxiety or fear about whether people will be marginalised or intimidated by technology. This anxiety characterised some of the initial speculation about the implications of the use of computer software in an increasing amount of intellectual and creative activity. The introduction of software into qualitative analysis is, amongst many things, perceived as potentially eroding the researcher’s control over the analytical process by for instance, increasing the distance between researcher and data and making research an artefact of technology, or that it will be used uncritically to produce ‘quick and dirty’ results.

Such concerns however, have not heralded a decline in their use within qualitative research. Indeed, the development and use of qualitative software is arguably becoming more widespread within academic research, particularly amongst the more recent generations of researchers. Researchers are not however, accepting the uncritical use of software like Nvivo, rather their wider use and acceptance is indicative of a social and cultural shift in the way technology is conceptualised. The anxiety that once determined populist beliefs about technology as a ‘Frankenstein’s monster’, i.e. what technology will inevitably do to us, is transforming into an acknowledgement that technology is an extension of us, i.e. they are an extension of human intelligence and skill. This is understood within what is referred to as the ‘posthuman condition’, whereby humans are viewed as embodied within an extended technological world.

By drawing on experience within research teams conducting qualitative cultural and social research, this paper explores the posthuman standpoint in relation to the use of qualitative software such as NUD*IST and Nvivo. In doing so, it not only highlights the creative use that can define its analytical support, but also that what should be feared in research relates more to wider and familiar political and economic processes shaping the research context.

Mind-Software Interaction: Does the application of software cause an epistemological problem?
Dr. Susanne Friese, Leibniz University, Hannover, Germany

In regular intervals, in mailing lists, discussion groups or papers the question is posted whether one runs into an epistemological problem by using software to analyse qualitative data. Back in 1999, I gave a talk at the First International Conference, Advances in Qualitative Methods, in Edmonton, Canada that was entitled: ”Does the use of QDA software influence the ways we analyse data?” In a self-test, I conducted a small scale analysis using a number of different software packages and concluded that there is no influence. With hindsight, the title of the presentation should have been: “Does the use of different software packages influence the ways we analyse data?” My answer to this question still is: generally no, apart from a few nuances.

However, if one asked me today whether software influences the analysis of qualitative data, I would answer yes, it does. But this does not cause an epistemological problem. Rather it offers the chance to get out of the black box of alleged subjective interpretation that qualitative researchers are often accused of by making the analysis process more transparent and comprehensible. Still, a lot happens inside the mind when a researcher is immersed in data, as Fielding and Lee (1991) write: “[R]esearchers engage in a continual process of inferring categories, testing their inferences against subsequent data, revising the inferred categories, and retesting the revised inferences against subsequent data. The ultimate outcome of this sequential process is well conceived theory (or conceptual propositions) that is heavily grounded in the data”.

It is difficult to document the entire analytic process but software at least offers the possibility to do more of it. Prerequisite is that the researcher builds up the analysis in a systematic manner. Software, if used appropriately, facilitates a systematic approach leading to improved reliability as well as validity of the research process and the findings (Friese, 2007).

The mind can interact with the software by documenting the steps of the research process, by using it as memory board, by following up on exploratory ideas, by setting up experiments and running tests. The technical functions of the various software packages, the existence or non-existence of certain tools, may hinder or further certain exploratory pathways. Software, however, is only one of the two players in the interaction; when looking at the human part it is important to consider skill level, technical understanding and creativity of the researcher. WE need to ask: How well or how badly is a researcher trained in using the tool “CAQDAS”?

If I may draw a comparison, first attempts with hammer and nails, electric screwdrivers or percussion drills are likely to lead to worse results than such tools operated by an experienced craftsman. CAQDAS well used are more likely to lead to better results, i.e. true (adequate) knowledge rather than false (inadequate) knowledge, thus offering reassurance rather than doubt with regard to the epistemological question.

According to my view, software does exert an influence on the analysis as it offers the mind different and new ways of handling data. In fact, when software is used the analysis can be approached differently than described in most method books as most are based on manual practices. Only a few books explicitly focus on computer supported qualitative data analysis methods and procedures (see for example Kuckartz, 2005; Friese, 2007).

In the presentation it will be shown how the use of software influences the qualitative data analysis process and that the mind-software interaction bears more chances than threats.

Fielding, N. G. and Lee, R. M. (1991). Using Computers in Qualitative Research. London: Sage.

Fries, S. (2007). The NCT-approach: a computer supported qualitative data analysis method, forthcoming.

Kuckartz, U. (2005). Einführung in die computergestützte Analyse qualitativer Daten. Wiesbaden: VS Verlag für Sozialwissenschaften.

Developing a model for the concept of Justice using a software : NVivo 2
Dr. Elif Kus, Ankara University, Turkey

In this study, we aim to develop a model for concept of justice by using qualitative data analysis software (NVvio2).

Justice is a concept which is defined differently in various times and spaces throughout the history. Beside this, views of jurists about justice also differ based on their philosophical stance. As individuals, we can find our selves confused in deciding “what is just or what is unjust” while acting in everyday life. In modern societies, justice is tried to be brought in an institutionalized legal structure. Despite this status of law in modern societies, the function and aim of law is still under discussion. In this study, we examined the meaning of justice according to jurists. Based on Gidden’s (1990) definition of “reflexivity in modern societies”, we assume that jurists’ views about justice have an influence on views of people in society. Therefore we limited our sample, with jurists’ views in this study.

The data were Turkish jurists’ articles about justice written in the period of 1933-2005. We treated these articles as a text like those which are derived from interview transcripts. Our sample consisted of jurists from two different universities’ faculty of law (Ankara University and � stanbul University) as a representative of various philosophic stances: positivist law, natural law and sociological law. There were 22 articles in our sample.

We made our analysis by following grounded theory coding schema. Our analysis was based on grounded theory because, we just want to derive the meaning of “justice” from the articles in our sample; we do not accept any theory or definition of justice at the beginning. During our analysis, firstly, we made open-coding to discover meaning of justice according to jurists. Our main category was justice (but there were no passages coded to it). We created sub-categories from our data. Then, by using software’s Model Tool we draw taxonomic models for each jurist to show his view of justice by using in-vivo categories. This first step represented the process of “decontextualization” At the second stage; we create new categories in order to reorganize our in-vivo categories. This second step was the process of “recontextualization”. These new categories are developed according to semantic-epistemic operations which are proposed by Allwood (1999). These were: 1. Basic semantic-epistemic categories 2. Basic conceptual structure 3. Anchoring in time and space 4. Relations 5. Processes 6. Roles derived from relations and processes 7. Properties 8. Quantity, modality and evaluation. Based on these eight category we developed sub-categories related with justice, as stated below:

Basic semantic-epistemic categories of jurists: a. positivist law b. natural law c. sociological law
Basic conceptual structure used by jurists: a. morality b. law c. social reality
Anchoring ‘justice’ in time and space: a. dependence to time and space b. independence to time and space
Relations cited by jurists: a. relation of justice and morality b. relation of justice and individual c. relation of justice and society d. relation of justice and law
Processes of justice: a. creating codes of law b. applying codes of law
Roles derived from relations and processes: a. moral role b. legal role c. social role
Properties of justice: a. subjective b. objective
Evaluation of justice: a. confirmation b. reputation

At the third step of our analysis we use these reorganizing (semantic-epistemic) categories in drawing taxonomic models. There were three taxonomic models at the end of our analysis representing “definition of justice” among jurists who have different view of ‘law’. In the process of drawing these taxonomies, we also used our “memos”, and other taxonomies showing each jurist’s view of justice as well as in-vivo categories. Using software, made our analysis process easy to gain a circular or spiral form. We went back and forth at any time by using software’s tools.

These three taxonomic models let us compare three different view of law and see the common points used by jurists when they define the concept of justice. Creating models by using the software’s model tool brings to analyst capture “the bird’s eye view” after all the process of splitting data.

Complexities, Levels, Quality and Ethical Issues in Qualitative Research: Do we Only Need Software?
Eugenio De Gregorio, PhD, Department of Developmental and Social Psychology, University of Rome

This paper presents the main results of a vast survey on training requirements for Police officers engaged in recording and checking testimonies. This study is part of a wider project involving the University of Portsmouth ( UK), the University of Bergamo ( Italy) and the University of Sassari (Italia): each research unit focused on a specific aspect.

The University of Sassari has been focusing on training requirements by employing innovative techniques of qualitative research, such as focus group techniques ( fcts), analysis of critical cases, brainstoring sessions.

FCTs represent a new challenge in terms of qualitative analysis: in fact not many studies exist on how to manage interactions and contents in order to reach results. For the most part scientific literature on fct draws attention to management modalities and their benefits (Krueger & Casey, 2000; Puchta & Potter, 2004).

This article aims to illustrate some options and puts forward some leading questions in relation to

the information format (transcribing vs. videorecording: Brown, 2002);
the role of software (namely, ATLAS.ti: Muhr, 2004);
the functions, advantages and disadvantages of a specific context and the role of the researcher (intrusion and biases, reflexivity);
reflexivity as a way to present such complexities to the scientific community and improve the quality of the study;
the potential reasoned integration between quantitative and qualitative approaches in gathering information as well as in analysing data and interpreting results (Bazeley, 2002; Morgan, 2007; Bryman, 2007).

With reference to the above issues, some critical discussions and operational alternatives will be advanced in order to highlight the following:

the mere use of software cannot be exhaustive in elucidating the results of good qualitative research (especially when this is methodologically complex or if it makes use of non canonical tools and it has implications for the participants),
qualitative research also needs “brains” capable of responding to the challenges that inevitably arise from quality requirements and ethical dilemmas.

In conclusion, this article will put forward arguments in favour of the thesis that qualitative research cannot simply rely on software or complex technological innovations, but it needs a more aware, context sensitive and relevant use of such innovations and technologies. One could think of software programmes as an extension of the person who uses them (Mecacci, 2000).

The following metaphor could well elucidate this concept: «The blind man uses his stick to know his way. His mind controls actions that might depend on how he uses his stick. This acts as a prosthesis that filters certain information and helps the mind re-elaborates such information. In a way we are all blind and we explore the reality with the help of certain instruments» (Mantovani, 1998, p. 121). Mind, instruments and objectives are not separate areas but they're all part of one system within a unitary process.

Bazeley P. (2002), Issues in Mixing Qualitative and Quantitative Approaches to Research, Paper presented at the 1 st Int.Conference “Qualitative Research in Marketing and Management”.

Brown D. (2002), Going Digital and Staying Qualitative, Forum: Qualitative Social Research, 3 (2).

Bryman A. (2007), Barriers to Integrating Qualitative and Quantitative Research, Journal of Mixed Methods Research, 1 (1), 8-22.

Krueger R.A. & Casey M.A. (2000), Focus Group. A Practical Guide for Applied Research , Thousand Oaks: Sage.

Mantovani G. (1998), L’elefante invisibile [The Invisible Elephant], Giunti: Firenze.

Mecacci L. (2000), La mente umana e il suo mondo artificiale [Human mind and its Artificial World], in G. Mantovani (eds.), Ergonomia: Lavoro, sicurezza e nuove tecnologie, Il Mulino: Bologna.

Morgan D.L. (2007), Paradigms Lost and Pragmatism Regained, Journal of Mixed Methods Research, 1 (1), 48-76.

Muhr T. (2004).

Puchta C. & Potter J. (2004), Focus Group Practice, London: Sage.

Benefits and drawbacks in a variety of approaches to the use of CAQDAS
Antonio Osorio, Altina Ramos, Ana Paula Martins, University of Minho, Portugal

The purpose of this paper is to analyse and discuss the way in which Computer-Assisted Qualitative Data Analysis Software (CAQDAS) supported the process of conducting research within a naturalistic paradigm.

We provide an overview of how three PhD researchers have differently approached and used CAQDAS for literature review, data gathering, data analysis, and report preparation. Researchers' experiences and approaches to their work included: i) ajo; ii) ar; iii) apm. Combining and discussing those experiences, we provide important information that can be used to inform research practice. Especially we highlight the many benefits and drawbacks that ensued from adopting, learning how to use, and using such software when making a detailed examination of a single subject, a group or a phenomenon. Additionally, we briefly illustrate the role of this research software in motivating and encour ag ing researchers to complete the report (dissertation) writing, in three different moments in time and approaches adopted: i) the first researcher used earlier versions of Nud*ist , mostly for data storage and management, in a final stage of his research; ii) the second, had a deep experience with Nud*ist 4.0 in a grounded theory based research project; iii) the third researcher used NVivo 2.0 to support data management throughout the research project, from data storage to data analysis and findings presentation.

It is argued that the use of CAQDAS as opposed to manual analysis had benefited the three researchers in terms of speed, consistency, motivation, accuracy, transparency, and access to data-analysis procedures. Furthermore, the process of “ play around with data ”, data man ag ement, writing up the final report, peer debriefing, and editing was a lot easier. Moreover it helped and motivated the researchers in the process where features and relationships were revealed within the data. Using our own experience for illustrative purposes, we provide guidelines about how to deal with the fact that the use of computer software for qualitative research may also add to a number of difficulties, such as: i) finding access to adequate software and to expertise in software use; ii) dealing with complexity in data man ag ement and data analysis; iii) temptation to keep rolling with data analysis.

Finally, we underline that CAQDAS does not carry out the researcher's work for us. On the contrary, it facilitates the achievement of more complex and, hopefully, better quality work: development of human knowledge, imagination, insights and hypothesis development, creative reflection, experience, adaptability to the variety of realities that are encountered in the research field. In a word CAQDAS may have something to add to the work of art embedded in the task of communicating research findings.

Conversations about CAQDAS: An Analysis of Qual-Software Discussion List
Urszula Wolski, Royal Holloway, University of London

In 1994 the CAQDAS Networking Project set up a discussion list, qual-software, the aim of which was to: ‘provide an on-line forum for the debate and information concerning the general usage of qualitative data analysis software packages’ (http://caqdas.soc.surrey.ac.uk/qualsoftware.htm).

The content of the discussion list was analysed from 1994 to 2005. The paper uses this analysis to assess how its members use the list; to identify the major themes and debates that occur and reoccur on the list, and examines how the list has changed over the twelve years of its existence.

The list is used in a number of different ways which include; a means for announcing forthcoming events such as conferences and workshops; a means of disseminating knowledge, for example about different software; the sharing and exchanging of problems and offering advice and solutions that arise in an individual’s or group’s research project, as well as providing a forum in which to engage in theoretical and epistemological debates that surround CAQDAS and qualitative research in general.

A frequently occurring debate on the list was the issues surrounding grounded theory and it can be argued that these same issues which occur elsewhere in the literature re-emerge on the list and so perhaps one of the functions of the list is that it is a reflection of what is happening and has happened in the field itself. Other frequently discussed issues raised on the list included problems with coding and in particular coding in teams, which software is most suitable for a particular analysis and discussions surrounding software for transcription and voice recognition.

What makes the list so interesting to study is that it provides an arena in which researchers of different disciplines and backgrounds can exchange thoughts, problems, advice and suggestions, whether it is software-related or methodologically related issues (or both).

The list has changed over the twelve year period, not only has it grown in terms of the number of members, but also in the growing diversity of its user base, with members coming from a number of different disciplines and backgrounds. The CNP (CAQDAS Networking Project) was always and still is an international community, but it can be argued that this international user base has extended, as has a national one.

Therefore, it can be argued that the list has certainly performed the aim of the CNP, but as well as so much more, both of which have been anticipated and unanticipated consequences of the medium it uses.

Collaborative Video Data Analysis: The Mediational Effects of MiMeG and other Digital Technologies
Marie Joubert and Anne Manuel, University of Bristol

This study grows out of two converging developments in contemporary social science: the emergence of digital video as a widespread resource for social and cognitive scientists to capture and analyse a range of human activity; and the increasing amount of research undertaken by distributed teams. This frequently requires researchers to share video data and to collaborate on the analysis of this video data remotely. Tools are being developed for the purpose of supporting such distributed collaborative work but the ways in which they mediate it is little understood.

This study aims to make a contribution to developing such an understanding by examining the use of one of these tools, MiMeG, in a research setting. MiMeG is a recently developed software tool designed to support the distributed, collaborative and real-time analysis of video data. Drawing on a sociocultural perspective, which suggests that human activity is historically and culturally situated and emphasises the mediational effects of tools on human activity, we explore in this paper the effects that using MiMeG and other digital technologies have on the collaborative research process.

MiMeG was used, along with a range of synchronous and asynchronous digital technologies in a qualitative collaborative research project, Trinity. The resulting collaborative work was filmed and then analysed to uncover the mediational effects of the tools used. Further, because our sociocultural perspective places an importance on context, the research was conducted in a range of settings including working face to face and remotely, and with different combinations of tools, which provides bases for comparison.

The data were analysed using a framework which maps the phases of the research activity against the software used and the modes of collaboration. Findings are discussed in terms of the potential and limitations of the range of tools used in each phase to mediate collaborative work.

EdEt - Experimental collaborative CAQDAS project in cultural anthropology research
Iwona Kaliszewska, Institute of Informatics and Institute of Ethnology and Cultural Anthropology, Warsaw University, Poland
Download Slides

The paper describes a new lightweight tool for analyzing, categorizing and archiving field material (especially interviews) dedicated for individual and collaborative work at home and in remote places. Moreover it presents results of its introduction to researchers at Institute of Anthropology and Cultural Anthropology and further ideas of development.

At our Institute there is a tradition of field research in Post-Soviet Countries. Researchers, alone or in project groups often spend long time in remote places. Such field research is already done by undergraduates. At the end of the field work students should have an insight into other participant's interviews and can use them as inspiration for further inquiry. This approach implies the following needs, not fully implemented in existing CAQDAS:

ability to effectively search in other researchers' material for ideas, inspiration and further analysis,
easily accessible centralized archive of field materials,
better and easier control over students' work,
ability to save materials in a secure repository.

Based on those, the requirements for a software tool for the Institute could be devised. The tool should be:

lightweight,
easily manageable by more conservative researchers and BA students,
adaptable to different methodologies preferred especially by older generation,
have an Internet repository for archiving and for group work,
feature coding and building codes hierarchy,
have automatic formatting and structure creation functions, tuned for interview processing,
have simple but powerful query system for easy search, retrieval and hypothesis building and testing,
store its data in XML format, allowing easy interoperability with other software,
freeware – accessible to students at home (working on interviews at lab did not work out), to use on computers in a field, without need to pay per copy.

EdEt (Editor for Ethnographers) described in this paper is an answer to those needs. I created it working on a joint project of Institute of Ethnology and Cultural Anthropology and Faculty of Mathematics, Informatics and Mechanics of Warsaw University.

Besides basic CAQDAS functionality of text editing, coding and querying EdEt puts the emphasis on a collaborative work, even between remote researchers. Using Internet repository they can share data or have an insight in others' data with assigned coding and annotations.

Other features of EdEt are connected with facilitating work on interviews. EdEt has functions which automatically divide a text into statements. One can assign user defined characteristics to persons and interviews.

The first part of the paper describes EdEt, its features and how it fulfills Institute's needs. The second part of the paper contains a detailed report about EdEt use. It was initially tested by a small postgraduates group doing research in Dagestan. Then, in October 2006 EdEt was introduced for wider use. Undergraduates working in 7 topic groups were to categorize their field interviews and then perform secondary analysis . The paper discusses the way EdEt actually affected researchers' work.

Finally the paper briefly describes future directions of development. EdEt is a starting point for theory building and hypothesis testing database and query language currently developed at our University.

Using IT to Conduct Collaborative Research Across Distributed Staff & Student Populations
Denis Edgar-Nevill, Department of Computing, Canterbury Christ Church University
Download Slides

This paper describes the work in progress on a funded project to study the development of Research Informed Teaching [DFES 2004] [JENKINS et al 2005] in new developing HE programmes. The effective use of IT is critical for communication between academics and students as well as peer group discussion between academics themselves and within student groups.

The research project is focused on how research as concept develops within students and staff in a new discipline; which is, to a very great extent still being defined and explored. It brings together academics, professionals working in a cutting-edge training centre with an international reputation and students (many of whom are themselves professional practitioners working in the private and public services). The qualitative research discussed here involves the use of technology to facilitate open-ended in-depth interviews which take place in number of different forms. Additionally focus-groups are used to compare and evaluate different experiences, in different settings. The intention is to use a holistic Systems Thinking approach [JACKSON 2003].

This project is based on students studying on the MSc in Cybercrime Forensics [CCCU 2005] and will be followed by a further comparative study on students completing the BSc in Forensic Computing [CCCU 2006] run by The Department of Computing at Canterbury Christ Church University. Developing these awards has been made possible by collaboration with the national police training organisation CENTREX and the specialised High Tech Crime Training Unit [CENTREX 2006]. Involvements with international developments in this subject have also been important [AGIS 2003]. Cybercrime Forensics has proved to be one of the fastest growing areas of new study for Universities in the UK the last three years as a response to the serious need for specialists to address this expanding area [EURIM 2004] [GHOSH 2006].

IT hardware and software is being used in a number of different ways to support the operation of the course programmes and measure how staff and students are engaging with research. It is used use to overcome barriers of geographical separation between groups and temporal barriers (where individuals and groups have limited availability. There are also privacy and security barriers to groups working together which also need to be addressed. It seemed a natural extension to make use of the familiar IT environment to broaden the purpose and nature of communication to facilitate the open-ended qualitative research required to gain a deeper insight into the students experience.

Understanding how pedagogy develops as a direct consequence of the introduction of new technology is crucial. Creating new learning infrastructures based on the use of IT is expensive and it is important to achieve the most effective blended balance between direct and indirect contact. Quantitative research gives some understanding but is limited by the numbers of students involved. The use of qualitative research methods is intended to

References
[AGIS 2003] European Union AGIS Programme 2003-2007 to help legal practitioners, law enforcement officials and representatives of victim assistance services from the EU Member States and Candidate Countries to set up Europe-wide networks, exchange information and best practices http://ec.europa.eu/justice_home/funding/agis/funding_agis_en.htm

[CCCU 2005] “ Canterbury Christ Church joins forces with central police training and development authority to deliver masters degree in cybercrime forensics”, Canterbury Christ Church University website, 24 th June 2005, http://www.canterbury.ac.uk/news/newsRelease.asp?newsPk=517

[CCCU 2006] “ New degree in Forensic Computing”, Canterbury Christ Church University website, 4 th October 2006, http://www.canterbury.ac.uk/news/newsRelease.asp?newsPk=740

[CENTREX 2006] Centrex, the Central Police Training and Development Authority, http://www.centrex.police.uk/cps/rde/xchg/centrex/root.xsl/home.html

[DFES 2004] The Relationship Between Research and Teaching in Institutions of Higher Education, Department for Education and Science, UK Government, 2004, dfes.gov.uk/hegateway/uploads/Forum's_advice_to_Ministers_on_ Teaching_and_ Research...

[EURIM 2004] “Supplying the Skills for Justice: Addressing the needs of law enforcement and industry for investigatory and enforcement skills”, European Information Society Group (EURIM) third discussion paper, IPPR E-Crime Study drafted by UK MPs, 2004

[GHOSH 2006] “ Computer industry 'faces crisis” Pallab Ghosh Science correspondent, BBC News website, Friday, 17 November 2006, http://news.bbc.co.uk/1/hi/technology/6155998.stm

[JENKINS et al 2005] Institutional Strategies to Link Teaching and Research, Jenkins A., & Healey M., Higher Education Academy, October 2005

[JACKSON 2003] Systems Thinking: Creative Holism for Managers (Wiley), 2003

Introducing Advanced Qualitative Querying (AQQ) Project
Sebastian Kaliszewski and Iwona Kaliszewska, Institute of Informatics, Warsaw University
Download Slides

AQQ is a project run at Warsaw University, aimed at creating textual database system with advanced qualitative querying capabilities, designed from the ground up to share data between many users allowing concurrent and remote access (via network).

The goal of the project is to create backend database system which then supports actual end user (frontend) applications. This is in fact an analogue of business relational database systems like Oracle TM, MySQL TM or SQL Server TM which are just backbones of actual custom applications. The difference is that AQQ database is not going to be a relational database with SQL querying language. In that way AQQ is an open system, which differentiates it from majority of today's theory building & hypothesis testing software. The other difference is an ability for users to work on the same data concurrently. But on the other hand the openness of AQQ combined with it's backend role (thus lack of advanced user interfaces, visualisation and graphical interaction) means that it is not meant to be a competition for current qualitative analysis tool but rather a complement and a mean to future development. As a case example it is planned to integrate EdEt program (also developed at our University) with AQQ.

The key part of AQQ project is a design of a specialised computer language to manage the database and write all the queries in. The language is going to be the only way by which external world communicates with AQQ database.

In a first part of this paper we present motivation behind AQQ project and detail its place on the qualitative software arena, further elaborating what has been written in the above paragraphs of this abstract. Then follows a presentation of a general design of AQQ system, describing its components and how actual users data is going to fit there - i.e. to be stored, accessed and manipulated. Third part introduces our initial design of AQQ computer language, with example queries and data manipulation commands. Finally future prospects are discussed, including frontend systems integration with an emphasis on EdEt project.

Formal Technical Reviews for Research Projects
Andre Oboler, Computing Department, Lancaster University, UK

Formal Technical Reviews (FTRs), where a software developer and a team of reviewers walk through a piece of code to assess its quality, have been used in the computing industry for many years. In this paper we look at the suitability of FTRs in an academic research setting.

We investigated the use of FTRs combined with other approaches in a set of case studies over a three years period. MSc students were used as test subjects with their dissertation projects being used as the research products. In all about half the student volunteered to participate each year, though very few took part in FTRs initially. Changes to the experimental setup through the introduction of additional artefacts to lower adoption slowly increased the participation in FTRs.

Our initial intervention included the introduction of a documentation style for software coding, this was largely about the approach to commenting and the specific type of data that should be captured for a research project. We also introduced the use of a software tool to extract comments from computer source code and generate a set of hypertext documentation complete with reverse engineered design diagrams. The generated documentation aimed to facilitate both the FTRs and discussion between researchers, for example office colleges, supervisors, or researchers at remote locations. We also found it necessary to create an installation and configuration guide to lower the barriers to adoption. A side effect of the generated documentation was the ability of the software to create a website for the participants with very little manual effort. This increased the benefits associated with the cost of learning the tool, and effectively reduced the barrier to FTR participation by making the tool more obviously cost effective.

While computer science researchers in universities typically shy away from quality improvement processes and have a high level of resistance to the introduction of new tools and techniques, the combination of approaches we used is shown to reduce the effort of participation required by both researchers and review participants to acceptable levels. This level of acceptance increased as the experiment progressed and more of the tools became available, and could with future work be improved further.

Past research shows a low level of adoption of similar software engineering approaches, and the current research repeatedly highlights the initial scepticism of both students and supervisors, however our data suggests that these barriers can be overcome with the right combination of tools and this can lead to improved collaboration between researchers in computer science or related areas.

Evaluation of MiMeG in Use: Technical and Social Issues in Remote Collaborative Video Analysis
Dylan Tutt (King’s College London, UK) and Muneeb Shaukat (University of Bristol, UK)

One of the major ways in which video materials are analysed in the social sciences and beyond is through the practice of group data analysis sessions (or video ‘data sessions’). With research projects increasingly comprised of inter/national teams that are distributed geographically and institutionally, MiMeG has been working on technological solutions to some of the problems of remote collaboration with video data.

Mixed Media Grid, a Research Node of the ESRC’s E-Social Science programme, is a collaborative project between the Computer Science and Education departments of the University of Bristol, and the Work, Interaction and Technology Group at King’s College London. Our MiMeG software, recently released under GPL, enables multiple participants in geographically distributed sites to control and view video data synchronously during the session. It allows social scientists to make (and log) freeform analytic annotations over the video using a pen-based input system, to interlink textual transcripts or other relevant mixed data sources, and provides audio communication through VOIP. From a technical perspective MiMeG is a customizable distributed framework that allows for the development of collaborative media systems.

A programme of research has been conducted to study our prototype MiMeG system in use by various groups of social scientists, who collaboratively analyse video as part of their everyday work. The distributed data sessions were video recorded at each participating site, to capture data concerning the embodied conduct of participants (use of media and objects, as well as talk, gesture and body movement etc.) and the screen is use.

The paper will discuss the advantages of MiMeG compared with the current range of technologies for distributed collaborative work, before evaluating the user experience of MiMeG – drawing on both social and technical issues. It will illustrate the problems facing participants during remote collaboration over video (such as the synchronisation of playback and annotations across sites), how problems were faced/solved in practice (through the use of talk, gesture and playback control), and will also document some of the methods developed by participants for effective remote collaboration over video, including the interplay of ‘local’ (designed for co-present site) and ‘general’ (designed for the remote site) talk and action.

The Role of Co-Interpretation of Routine Data in the Evaluation of a Virtual Research Environment
Richard Procter, Vito Laterza & Patrick Carmichael, Centre for Applied Research in Educational Technologies (CARET), University of Cambridge.

This paper reports a novel research approach in which routine server data collected from online environments provides a 'focus' for semi-structured interviews in the course of which research participants are involved in a process of 'co-interpretation' and sense-making (Weick, 1993) about emerging trends in individual and group behaviour online. The context for this approach is a two year development and research project under the Joint Information Systems Committee (JISC) Virtual Research Environments Programme, in which we have worked with ten projects in the ESRC Teaching and Learning Research Programme (TLRP). One of the key research foci of this project was to better understand the needs of research groups during the entire research 'life cycle' from project initiation through to reporting, publication and dissemination; another was to explore the ways in which project roles were defined and manifested in the online environment.

The TLRP Virtual Research Environment (VRE) is based on the Sakai platform which provides an access controlled online suite of interchangeable modular collaboration tools that geographically distributed research teams can use to support their work. It is highly customisable and tools can be deployed within the environment to suit the needs of individual research projects. Sakai generates very detailed 'logs', and the aggregation of the data from these server logs and database tables provided insights into how specific users and groups were engaging with the environment.

These largely quantitative data are useful in determining trends in behaviour, broad patterns and sequences of events, but tell us little about context other than 'who', 'when', and 'from where'. In order to explore further what contributes to these observed patterns, we designed a focused interview protocol in which, following appropriate data reduction, we engaged research participants and VRE users in a set of co-interpretation activities in which they provided interpretations of specific incidents and offered rationales for short-term and long-term trends in access, use and online interaction.

We present this protocol together with some typical data used as the focus of interviews. We show that it is possible to construct rich and illuminating qualitative accounts which help to 'make sense' of: single events (such as sudden increases, decreases or changes in online activity); repeated sequences and 'activity-flows' (such as the online publication of documents followed by moderated periods of peer-review, or sequences of data collection or analysis); and long-term trends in activity over the project life-cycle. Our findings suggest that the approach of using suitably reduced quantitative data supports reflective, concrete accounts which are more useful to evaluators, software developers and project managers than less focussed, more generalised and summative evaluation approaches which tend to overemphasise general notions of 'satisfaction' and 'affordance'.

Online environments are coming to play an important part of the work of many research projects in education and other scientific disciplines. Our findings and the approaches we have used have implications for those involved in commissioning, setting up, using and evaluating online environments.

Weick, K. E. (1993) "The collapse of sensemaking in organisations: The Mann Gulch disaster", Administrative Science Quarterly, 38(4), pp 628-652.

Collecting, Managing and Analysing Data within Multi-disciplinary Research: Challenges from a four year collaborative research project
Susan Venn & Sara Arber, Centre for Research on Ageing and Gender, Department of Sociology, University of Surrey

This paper will raise questions about the pertinence of using IT within a recently commenced multi-disciplinary, multi-institution project. The aim of this four year ESRC funded project is to extend and ‘join up' strategically targeted areas of sleep research, and to address practice and policy relevant issues arising from the nature, impact and management of the sleep-wake balance in later life.

Within this paper, I will focus on a workpackage which aims to understand the quality of sleep of older people living in the community, to discuss the challenges facing the project team in terms of multi-faceted data collection as follows:

Standardised sleep questionnaires

Interview transcripts

Audio interview data

Video data

Audio sleep diaries

Actigraphy output (a watch-like device measuring 24 hour activity/light exposure)

Daily written diaries of activities, medication use, and food/drink consumption

Analysis of qualitative data, including interview transcripts, audio diary and video data, will be undertaken using NVivo 7, actigraphy output by Sleep Analysis software, and regression models will be undertaken on data from activity diaries. The challenge will be to link these data, both within the analytical framework of the workpackage, and to other workpackages within the project that will be physiological (light therapy in care homes and the community), engineering (sensor-based products to optimise sleep for residents in care homes and the community) and psychological (non-pharmacological treatment options for sleep problems). One potential solution to the challenge of analysing mixed-method data such as actigraphy output (via sleep analysis software) and qualitative interview data, will be demonstrated through the use of algorithms previously developed for analysing couples' sleep (Meadows et al, 2005).

Two further challenges remain, the first being the presentation of findings from this workpackage as a module on older people and sleep on a web-based health and illness resource in textual, audio and video format (www.dipex.org). The second being the management of this large scale, collaborative research project in terms of project management and multi-platform encrypted data sharing. The suggested solution to this final challenge, Microsoft Office Groove 2007, will be demonstrated.

Meadows, R., Venn, S., Hislop, J., Stanley, N. and Arber, S. (2005) Investigating Couples' Sleep: An evaluation of actigraphic analysis techniques. Journal of Sleep Research 14:4. 377-386

CAQDAS 07 : Abstracts

Search

Links: