For best experience please turn on javascript and use a modern browser!
Our Research

Text and Talk

Any handbook on the analysis of spoken or written discourse will stress that the investigation thereof does not start only after fieldwork has been done and data has been collected, but during fieldwork and data collection.

Rather, it is one of the main advantages of doing qualitative fieldwork to be able to analyze while collecting. This adds not only flexibility to the analysis, but also the possibility for serendipitous findings as well as better data collection informed by previous findings.

After stressing the iterative nature of most qualitative analyses, many introductory texts start out with basic decisions to be made at the start or on the way, leading to typologies of research. The over-emphasis on typologies, often based on obsolete philosophical ideas, constrains many researchers, however, and creates obstacles to flexible adaptation to the object of study, which almost never lives up to academic typologies. It is better to acquire a broad knowledge of, and then use, multiple qualitative methods.

Types of qualitative analysis: interview conceptualization

An important example of interweaving data analysis and data collection is in the conceptualizations of open interviews. There, the conceptualization of the interview has salient consequences for the analysis as well as for the interviewing itself.

For instance  Silverman (2011) lists three different conceptualizations: positivist, emotionalist, and constructionist. In the positivist approach the analyst attempts to find facts in the data. In the emotionalist approach the focus lies on finding and describing the emotions of respondents. In the constructionist approach the analyst puts the emphasis on the social (inter)action during the interview, as conversation analysts do. It's obvious that the interviewer's probes and prompts will vary across these three different approaches.

Besides Silverman’s simplifying yet useful threefold typology, others, such as  Roulston (2010), have created more fine-grained typologies on interview conceptualizations, which might be used for other methods of data collection as well.

CAQDAS or not?

A basic decision to be made is to either use specialized CAQDAS software (Computer Assisted Qualitative Data AnalysiS), or to just use a text program (or a notebook, which is fairly rain-proof when substituting a pencil for a pen). There have been debates on the usability as well as the desirability of using software in qualitative analysis, or specific forms of qualitative analysis. See for instance  Coffey  et al. (1996)Macmillan (2005), or  Goble  et al. (2012) for critical appraisals, and the debate around the paper “Illumination with a dim bulb?” in Sociological Methodology 42 (2012). 

CAQDAS enthusiasts can be found at  Surrey. People who already use R can use its specialized packages; see the related page on digital and digitized media. There are many specialized (non open source)  programs, each with its advantages and disadvantages; some are suitable for coding video material as well.

Types of qualitative analysis: Specific approaches

Specific interest(s) of the researcher lead to a typology based upon specific approaches that match those interests. In  Tesch (1990) 27 different types of analysis were described, based on four different research interests:

One of the experts at the AISSR is Gerben Moerman. Questions about Q-methodology in particular can be posted to Michael Deinema.

Although there is nothing against pencil and paper in the field (still readable after being soaked by rain), Computer Assisted Qualitative Data AnalysiS  (CAQDAS) is frequently used. People who already use R can use its specialized packages; see also the related MEC page digital and digitized media. There are also many specialized (non open source)  programs, each with its advantages and disadvantages; some are suitable for coding video material as well.

One of the experts at the AISSR is Gerben Moerman. Questions about Q-methodology in particular can be posted to Michael Deinema.

Textbooks for a general overview; take the typologies with a grain of salt:

  • Gibson, W., & Brown, A. (2009).  Working with qualitative data. London: Sage Publications. 
  • Roulston, K. (2010).  Reflective interviewing: A guide to theory and practice. London: Sage Publications. 
  • Ryan, G. W., & Bernard, H. R. (2009). A nalyzing qualitative data: Systematic approaches. London: Sage Publications. 
  • Seale, C. (1999).  The Quality of Qualitative Research. London: Sage Publications. 
  • Silverman, D. (2011).  Interpreting Qualitative Data. London: Sage Publications. 
  • Tesch, R. (1990).  Qualitative Research: Analysis Types and Software Tools. London: Routledge. 

Journals on qualitative research

A warning on careless jumping from stories to science

  • Gelman, A. and Basboll, T. (2014) When do stories work? Sociological Methods & Research 43: 547-570. They argue that stories should represent aspects of social life that are not well explained by current theory, and that the details and context should be well-established—use thick descriptions!—to make them robust against misinterpretation. 

Computational approaches such as machine learning make inroads into the analysis of text and talk, can do it quickly, but not nearly as good as humans can.