CAQD Conference 2013. Marburg, Germany, 6-9 March 2013, Marburg Research Group for Methodology and Evaluation (MAGMA), Phillips University, Marburg

Abstract: Nestled on the banks of the river Lahn in central Germany, the 15th CAQD conference was held at Marburg. A beautiful provincial town, it is one of very few that was spared the bombings of WWII; now providing the perfect backdrop for meeting to discuss developments in qualitative technology. This was the second international conference in the series with more than 140 delegates from 14 countries, including: Canada, Brazil, Portugal, the UK, as well as Germany. Hosted by MAGMA, the Marburg Research Group for Methodology and Evaluation, in partnership with Philipps-University Marburg, CAQD prioritizes a user-focus which balances practical and methodological workshops with conference presentations.

CAQD conferences are characterized by the centrality of workshop programs offering students and researchers' opportunities to develop methodological and practical skills in qualitative and mixed methods techniques and applications using the software MAXQDA. This year there were a total of 24 workshops, covering topics from "Starter" and "Advanced" user workshops to qualitative evaluation and analytic visualization. Several were provided in English and Portuguese, as well as German, for the international delegation. All were led by expert methodologists and software trainers. [1]

The conference atmosphere was both relaxed and professional, with many opportunities for networking. The sign of good conference organization is the creation of spheres in which delegates working across disciplines and contexts and with different levels of expertise and experience can share ideas and debate topics in a friendly environment. [2]

Professors with established reputations were engaged in lengthy discussions with students new to the topic. Researchers working on opposite sides of the globe discussed similarities and differences in the use and teaching of qualitative methodology and software. Long-standing experts in software development and qualitative methodology reignited discussions in light of new technological possibilities and user trends. There was a buzz, and I was caught up in it. [3]

Nigel FIELDING, Professor of Sociology at the University of Surrey and Co-Founder of the CAQDAS Networking Project, kicked-off the conference with an informal evening lecture. [4]

Presenting a biographical and humorous narrative of his engagement with computer assisted qualitative data analysis over the past 25 years, FIELDING shared his assessment of the field and expectations for the future. This included relating his early ventures into technology (which involved avoiding gym class and eating chocolate cookies), "scrapes" with the authorities when researching right-wing political organizations in the 1970s for his PhD research (which involved an abrupt ending to his covert participant observations), and the origins of the acronym "CAQDAS" (a play on the word "cactus," the "prickly" plant, to reflect that using software for qualitative analysis was at that time (late 1980s) perceived by many as a "thorny issue"). [5]

He also reflected on the first conference ever held on the topic of computer assisted qualitative analysis, one that he co-organized at the University of Surrey in 1989. It was this event which—somewhat unexpectedly—sparked FIELDING's engagement with the field. The conference resulted in a publication (FIELDING & LEE, 1991) and led to the founding of the CAQDAS Networking Project in 1994. FIELDING has since published numerous articles and books on the methodological affordances and implications of the use of these software, and secured several tranches of funding to provide training and to research the area. In documenting the key stages in his subsequent career, FIELDING reflected on the contributions of early figures in qualitative computing, such as Mike AGAR and Renata TESCH, developers and methodologists such as Tom RICHARDS and Lyn RICHARDS (NUD*IST & NVivo), John SIEDEL (The Ethnograph) and Udo KUCKARTZ and Anne KUCKARTZ (MAXQDA). He also commented on some of the early resistance to qualitative technology—some of which centered on whether software would "do" the analysis—and thus, the extent to which using software meant you were being "scientific." Others pertained to the sense in which software acted as a "barrier" to analysts' engagement with data, or "homogenized" approaches to analysis. [6]

What struck me in listening to FIELDING's reminiscences in this regard— depressing me somewhat—is that these arguments (although common in the early years, given the "newness" of technology) still exist in some quarters. There are three camps in qualitative analysis—those that use software, those that do not and will not (largely as a result of ignorance of developments in technology), and those that have no idea of its existence. Indeed, much of the methodological literature concerning the use of CAQDAS packages still refers to those early criticisms—which have no relevance anymore due to the extent of technological development generally (e.g. digitalization of "data" and normalization of technology) and the development of analytic possibilities provided by specific software tools. [7]

In discussing technological developments, FIELDING noted IBM's claim at the 1992 CAQDAS conference in Bremen that "they would have accurate automated transcription within three years. We are still waiting." Technological development occurs in fits and bursts, ebbs and flows, influenced by a range of factors, not least commercial pressures, but also driven by methodological need. FIELDING's thoughts about future directions for the field included musings about what the "wonders of the internet and the online world" might result in; the way that "systematic analysis using CAQDAS can challenge the minds of people who believe that only quantitative methods and random control trials can tell us anything useful" and "the increasingly valuable archives of qualitative data that give us new possibilities for secondary analysis." FIELDING however, sees "citizen research" as amongst the "most exciting" developments. There is no stopping what is happening in the online world—as he puts it "you cannot put the genie back in the bottle." But does that mean that we, as social scientists, should be threatened by the democratization of social research? FIELDING thinks not. [8]

The main conference program constituted three keynote papers, round table discussions, and a software user forum. [9]

The keynote sessions were chaired by Anne KUCKARTZ, CEO of Verbi software (developers of MAXQDA). Clive SEALE, Professor of Sociology at Brunel University, London, began by discussing the issues and potential of conducting secondary analysis of qualitative data. He began by discussing databases for archiving qualitative research data, and possibilities for exploiting naturally occurring qualitative records, and other information generated for non-research purposes. He then outlined advantages of and objections to qualitative secondary analysis and made the case for his own position—that the opportunities afforded by secondary analysis of qualitative data for scholarly inquiry should not be squandered. Clive went on to illustrate this using the example of a project he is currently involved in that is re-analyzing textual data about the administration of medication to people near the end of their lives in three European countries. He described a text-mining approach to the analysis of the large textual data corpus using the software WordStat. [10]

This involved the generation of dictionaries of words relating to the concepts under investigation (for example, discourses around "saying goodbye") and using them to compare groups of texts derived from the three countries. The approach includes several elements that incrementally build upon one another: 1. listing and comparing the frequency of common words and phrases likely indicative of pertinent content; 2. examining the context surrounding the most frequent of these occurrences to control for equivalence in meaning of usage; 3. developing a specific dictionary containing categories that group terms with semantic similarity; 4. using thesaurus functionality to identify the occurrence of similar words and phrases and inspecting surrounding context for semantic equivalence; 5. arriving at a dictionary for use with the particular data set, and applying its categories across the data set to compare their distribution in texts derived from the three countries. [11]

As well as giving the audience a tantalizing insight into some of the projects' early findings, Clive also discussed some differences in outcomes evident from the computer-assisted text mining approach and a manual approach to the same data set employed by a colleague. A rare exposition and discussion of the specific procedures employed in a research project utilizing qualitative software, Clive's presentation whetted my appetite for more—both in terms of findings and publications documenting computer-assisted analytic procedures. [12]

Margrit SCHREIER, professor of empirical methods in the humanities and social sciences, Jacobs University, Bremen, Germany, then provided an overview of qualitative content analysis (QCA), uncovering the extent of diversity in the field whilst drawing out synergies across approaches and presenting the case for a "toolbox" approach to their implementation. SCHREIER's demystification of the field pivots around her conceptualization of the "core and the periphery," in which she distinguishes between and prioritizes the procedures and methods employed by QCA variants (see for details SCHREIER, 2012). [13]

Her initial definition is broad: "QCA is a method for systematically describing the meaning of qualitative material. This is done by classifying sections of the material as instances of the categories of a coding frame." Versions of QCA include structural content analysis, thematic analysis, formal content analysis, scaling through content analysis, evaluative content analysis, summative content analysis, content analysis by extraction, typological content analysis and explicative/contextual content analysis. [14]

SCHREIER places structural content analysis at the "core," noting that it is the most frequently used type of QCA, corresponding closely to thematic analysis, but arguing that its procedures underlie all other types of QCA. In building this argument in her presentation, SCHREIER outlined procedures involved in the variant types, provided empirical examples of their use and illustrated the relation between different types of QCA in evidencing her placing of structural QCA at the core. [15]

SCHREIER's conclusion is that despite appearances, there are actually not many types of QCA, but that there is one core procedure (structural QCA) and many variations on this procedure. Her key message was that advocating a "toolbox" approach means that researchers are at liberty to "take what they need as long as it's suitable for your research question." Sound advice and not restricted to the application of QCA, but to scholarly research and the use of software to support analysis, generally. [16]

Professor Udo KUCKARTZ, Institute for Education & MAGMA, Marburg University, Germany, built on SCHREIER's presentation, illustrating potentials for conducting structural QCA through discussion about the new "summary grid" tool in MAXQDA Version 11. [17]

KUCKARTZ began by outlining the characteristics of types of QCA and how they may be performed in MAXQDA. In distinguishing between three main types of QDA (thematic/structural, evaluative, and type-building) he reflected on SCHREIER's presentation in prioritizing their systematic and replicable nature, their focus on category development, the sense in which they are guided by research questions, inter-subjectivity, and standards such as inter-coder agreement (or reliability). [18]

In distinguishing between his identification of these three main types of QCA KUCKARTZ illustrated how a content analysis of AIDS brochures would proceed differently in taking each approach using the category of "false-positive probability" as an example. In a thematic QCA analysis the category would be identified in the brochures, coded, summarized, systematized and relevant text passages presented to illustrate the category. In an evaluative QCA analysis the treatment of the category "false-positive probability" would be evaluated according to its presence as "extensive," "incomplete" or "omission"—through an identification and analysis of frequencies of occurrence and the presentation of examples. In a type-building QCA analysis different types of treatment of the category would be identified and thereby typologies developed. [19]

The possibility of conducting cross-case analysis was explored in the context of these approaches whereby KUCKARTZ illustrated the affordances of constructing matrices of cases by themes and presented options for analysis and presentation by describing different forms of analysis. He then went on to demonstrate the MAXQDA summary grid and summary table tools and discussed how they offered new options for computer-assisted qualitative data analysis. This included a description of the influence of MILES and HUBERMAN's (1994) "case-level display for meta-matrices" on the development of MAXQDA's tools. He illustrated the potentials of the "Code Matrix Browser" for displaying matrices of themes by cases and discussed the sense in which such general overviews, although useful as high-level analytic tools, represent a "monster-dog" approach. [20]

The new summary grid function was thus presented as a means of integrating the quantitative representations of coded qualitative data provided by matrices alongside the qualitative records themselves, whilst offering the concurrent opportunity of further integrated analysis through the development of interpretive summaries. He thus presented the methodological principles behind summary grids and their analytic utility as a means of "reading segments and writing summaries." [21]

Amongst the main characteristics of summary grids KUCKARTZ outlined the following: 1. their operationalization as a new level of analysis; 2. their role as second order data; 3. how their role in enabling aggregated, abstract, and evaluative level work facilitates analysis; 4. the move away from "monster dog matrices" to more usable formats; 5. the full integration with and access to primary data whilst working at the higher level; 6. the dynamism of summary grids, in facilitating incremental analysis. KUCKARTZ concluded his presentation by showing grid tables as a means of taking work done in summary tables further and comparing the characteristics of structural QCA with other analytic approaches, and alluding to ways in which the MAXQDA development team are currently working on additional ways of facilitating within and across case analyses. [22]

Keynote presentations were followed by roundtable user forums at which delegates were invited to discuss one of five topics relating to the use of software in qualitative and mixed methods research. I joined the "teaching and learning" table along with several other methodology lecturers and software trainers working across academic disciplines and national boundaries. [23]

Having co-led two workshops earlier during the conference using a different set of materials and teaching methods than I usually employ, I was keen to hear about the way others' impart technical, methodological, and practical aspects of MAXQDA in their teaching. This proved a very fruitful discussion which highlighted similar challenges across settings as well as opening up my thinking to alternative ways of teaching. I relished the opportunity to learn from others in this way, and to challenge my established methods and assumptions. Other roundtable discussions covered the topics of "mixed methods and complex data analysis"; "implementing research methods (e.g. qualitative content analysis) using MAXQDA"; "using visualizations for presentations and analysis"; and "organizing data, project structure, working in teams and general/technical questions." I heard from delegates who participated in these discussions that they resulted in similar thought-provoking conversations. [24]

Following the roundtable discussions was a formal opportunity for delegates to ask questions of the MAXQDA development team about specific aspects of the functioning of the software and to add to their "wish list" for new features. Given the relative currency of the latest release (Version 11 was released in December 2012) this was a particularly timely opportunity for Verbi to canvas user-response and respond directly to specific queries. [25]

Listening to others' wishes also enables oneself to think differently about software functionality and potential for methodological applications. Amongst functions discussed in this feedback session were the practical use and potential role of the new "emoticodes" functionality, particularly in terms of possibilities for democratizing coding processes, facilitating cross-national collaboration and opening up CAQDAS technology for lay use (citizen research). Users expressed a desire for the proliferation of color into the new summary grid functionality in similar ways as is enabled across MAXQDA functions. In addition, options for converting user-generated summaries into data documents in such a way that links to originating "raw" data are maintained whilst simultaneously being able to integrate interpretation with data would add value to summary grid functionality. This sort of capability would provide "system-closure type" completion of a circle of data-interpretation—taking potentials for treating ones' own interpretations as data to another level, and thereby more fully attending to reflexivity, researcher biases, and quality. [26]

Several themes resonated throughout the three days I was at the conference, not least the role of software in analytic processes. Each keynote presentation alluded to integration—be it related to materials, analyses or contributions. The background context of software support for qualitative and mixed methods analysis of course was the main undercurrent at CAQD 2013, and the workshops, presentations, and user discussions all alluded to the relationship between technology and methodology—whether explicitly or implicitly. [27]

Margrit SCHREIER's presentation, which focused on methodological application irrespective of software, anchored discussions by prioritizing analytic strategy. In discussing QCA in general methodological terms rather than commenting on specific ways in which its variants might be achieved through the manipulation of software tools, she promoted the idea that analytic strategies can be created for particular research studies by drawing on the principles and procedures of multiple approaches. Flying in the face of methodological determinism such arguments place the requirements of research questions at the center of research design and analytic strategy development. Not only speaking against age-old distinctions between and debates about qualitative and quantitative approaches, this sort of methodological creativity is also promoted by virtue of the continued existence of a range of qualitative software packages and plethora of tools within them. [28]

Clive SEALE illustrated how software can be employed to achieve specific analytic ends; demonstrating in practical terms how analytic requirements can shape the way software tools are used, and alluding to differences between analyzing the same set of data with and without the use of software. In discussing work-in-progress he reported on how software-supported text mining had been conducted in parallel with a more traditional qualitative analysis by a different researcher on the same extensive secondary data set. These separate analyses were being brought together subsequently, enabling processes to be compared, and differences in findings identified. His example clearly illustrates how software allows far greater quantities of data to be systematically analyzed when a quantitative text mining approach is being employed. But there are also possibilities for integrating discrete analytic approaches through the use of software, for example, using deductive techniques such as those afforded by text mining to identify broad themes which can then be more inductively analyzed—with or without the use of software. [29]

Udo KUCKARTZ explicitly illustrated the relationship between technology and methodology operating in the opposite direction, showcasing how the development of new software tools offers alternative potentials for data analysis. Using structural qualitative content analysis as an example, his presentation illustrated clearly how methodological needs underpin the development of MAXQDA's software tools, resulting in new analytic possibilities. Indeed, developmental focus on mixed methods tools in MAXQDA over the past few years—including the various matrix functions, cross-tabs, typology tables, as well as summary grids and grid tables—has contributed to the furtherance of analytic possibilities. In some areas, such as visual analysis, I would argue that software still lags behind the practical and analytic needs of some methodologies. In the context of mixed methods analyses, however, software tools such as those provided by MAXQDA offer more potential than are discussed in the methodological literature. [30]

Nigel FIELDING's evening lecture framed the keynote presentations, providing a historical context within which contemporary methodological debates and technological developments can be interpreted. As a pioneer in framing the methodological scenery from the late 1980s and in establishing a center for training and debate in the UK which has significant reach internationally, FIELDING is amongst the best placed to comment on the scope of CAQDAS' history. His reflections concerning the pace of technological change during his career indicate that more significant change is on the horizon. Technological development is occurring at an unprecedented pace. It seems to me that the task of methodologists, teachers, and software developers, is to attend to the dynamic relationship between technology, methodology, and teaching in order that we may serve the next generation of students and researchers in providing professional tools which offer user-friendly and analytically appropriate tools that serve to stretch methodological boundaries. [31]

Rather than homogenizing qualitative analysis—as some critics of CAQDAS still contend—software opens up possibilities for integration at all stages and levels of analysis. The lectures, workshops and discussions at CAQD 2013 are testament to this. [32]

Fielding, Nigel G. & Lee, Raymond M. (1991). Using computers in qualitative research. London: Sage

Miles, Matthew B. & Huberman, A. Michael (1994). Qualitative data analysis: An expanded sourcebook (2nd edition). London: Sage.

Schreier, Margrit (2012). Qualitative content analysis in practice. London: Sage

Christina SILVER is the manager of the CAQDAS Networking Project, for which she has worked since 1998. The project provides information, advice, and training in a range of software packages designed to facilitate the analysis of qualitative (and increasingly, mixed methods) data. She leads the training and capacity building activities of the project and is involved in research into the use and implications of qualitative technology. Christina's particular interests relate to the relationship between technology and methodology, the integration and analysis of non textual data, using software and the impact of technology on the teaching of methodology. Christina is co-director of Day Courses in Social Research, a program of methodological training provided by the Department of Sociology at the University of Surrey, England. She is also the co-founder of Qualitative Data Analysis Services (QDAS), which provides customized training and consultancy services for individuals and groups engaged in qualitative analysis.

CAQDAS Networking Project
Department of Sociology
University of Surrey
Stag Hill
Guildford, GU2 7XH
United Kingdom

Silver, Christina (2013). Conference Report: CAQD Conference 2013 [32 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 14(2), Art. 24,
http://nbn-resolving.de/urn:nbn:de:0114-fqs1302249.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License