A Pragmatic Approach to Qualitative Data Analysis

Beautiful 340
Artistic 239



Once such a listing has been obtained, it is then possible to exclude non-lexical words, such as ‘the’, ‘to’, ‘that’ and so on. What is left is an interesting array of the lexical words used in a transcript. Quite how useful such a listing is remains open to debate. If thought is given to how people talk, it will become evident that they may well repeat a given word many times. Consider, for example, this response to the question: ‘what do you think about nursing models?’



Example response


‘Nursing models? Well, I think some nursing models… well, a few nursing models, anyway, can be quite useful. You know, if you are talking about nursing models and not nursing theories…now I am not sure what a nursing model is!’


In this example, the phrase ‘nursing models’ is used five times. It is not difficult to see that there may not be a relationship between how often a word appears in an interview and the importance of that word. We have used this method once and can offer an example of how the method can be used to highlight issues in a research project that were missed with other forms of analysis.


Some years ago, PB undertook a series of telephone interviews about the topic of HIV/AIDS counselling. Each of these was recorded and then a program was used to count the number of occurrences of lexical words in the transcripts. Other forms of data analysis were also used. What became clear was that the word ‘anger’ headed the list of lexical words, following this analysis. The author then returned to reading the transcripts and realised, for the first time, that the notion of anger did have a large part to play in the interviews. As I have noted, this was missed with other forms of data analysis.


Used with extreme caution, the word-count form of content analysis has its place in the analysis of textual data. However, care should be taken in making any detailed sort of interpretation of the data that arise out of this activity. The objective nature of the word count, can, most usefully, be combined with a more subjective review of the data in the light of this word-counting process. Word counting, alone, is not a useful or valid activity in qualitative research. If we were to draw an analogy with quantitative research, this kind of content analysis would be, at best, the equivalent of the simplest kinds of descriptive statistics, but possibly no more than frequency counts. In other words, this is merely presentation, rather than transformation of the data. In both quantitative and qualitative contexts, no meaning is either clarified or added at this simplest level.


Grounded theory


Although often written about as a research method in its own right, grounded theory is also an approach to analysing qualitative data. The idea behind grounded theory is that social facts ‘emerge’ out of the collected qualitative data, rather than the researcher being able, in some way, to ‘divine’ meanings about those data. An analogy that may help here is that of a block of marble prior to its being sculpted. The sculptor may argue that his projected horse is already ‘in’ the marble and it is his job to help it ‘emerge’. Similarly, those who take a grounded theory approach to data may argue that meanings are already ‘in’ the data they collect and that it is their job to ‘release’ those meanings.


It is important to highlight the word ‘theory’ in this discussion. In the grounded theory literature, the aim is to use the emergent data to produce a theory which can then be further tested out. In our experience, quite a number of researchers who claim to do grounded theory seem to omit the theory. The whole point of grounded theory seems to be to collect rounds of data and to continue to refine the theory that emerges out of each round.


Grounded theorists, like ethnographers and those who use a phenomenological approach to research, still have to analyse their data into categories. There often appear to be great similarities in the ways that all qualitative researchers analyse their data. The rest of this chapter, then, is taken up with practical ways of analysing data and, we would argue, that can be used by those who do ethnographic, phenomenological or grounded theory approaches to research.


We appreciate that purists in each of these areas of qualitative research might argue that there are considerable differences in the ways they analyse data. If you are planning on doing a qualitative study, it is important to read widely around these topics.


Thematic content analysis


Frequently, in qualitative research, a form of ‘thematic content analysis’ is used to help organise and structure the data that accumulates from interviews. There are at least two opposing approaches to such analysis. The overall aim of both sorts of analysis is to first of all identify ‘themes’ or topics that occur in the data. Once these themes have been identified, the researcher can comb through the data for examples of utterances under each of these themes and cut and paste the data under these themed headings.


However, as noted above, there are at least two approaches to the identification of themes as part of such an analysis. The first might be called a ‘template’ approach. Here, the researcher asks questions of the dataset. Thus, in a series of interviews about health care education, the researcher might ask him or herself the questions: ‘What did the interviewees say about teaching?’, ‘What did they say about evaluation?’, ‘What did they say about reflective practice?’ and so on. In turn, each of the short versions of these questions becomes a category. Thus, in the example above, the first set of categories (or report sub-headings) are ‘Teaching’, ‘Evaluation’ and ‘Reflective practice’.


The researcher continues to ask questions of the data, in this way, until he can account for almost everything that was said, by all the interviewees. Once this has happened, the data can be cut and pasted under the various headings generated in this way. A variant of this approach – particularly for a very large dataset – is for the researcher to pose only questions to which he is looking for an answer. Thus, his questions might be: ‘What were the positive things said about teaching and learning methods?’ and ‘What were the negative things said about teaching and learning methods?’ In this way, the researcher ‘interrogates’ the dataset and selects from it data about the particular and specific issues he is interested in and which will help to answer his research question.


The other approach to thematic content analysis is to adopt almost the opposite position to the data. In this method, the categories of data are said to ‘emerge’ from those data. It is the researcher’s job, here, not to look for categories of information but to allow their emergence. What is more debatable, perhaps, is the degree to which categories ever do ‘emerge’ and to what degree the research ‘looks’ for them. As this is one of the most frequently used methods of content analysis of textual data, it will be described in more detail. The method described here is the one developed by one of us (PB) and which has been widely cited as a method of qualitative data analysis in other people’s reports. The method, as described below, refers to the analysis of interview transcripts. However, it is quite possible to adapt the method for analysing any other form of texts (e.g. newspapers, handouts, curriculum documents, government reports).


A pragmatic approach to schematic content analysis


Stage one


Notes are made after each interview regarding the topics talked about during the interview. It is useful if the researcher writes herself ‘memos’ during the research project. A memo, in this case, is a short note about an idea, theory or any other mental activity that has taken place within the researcher. The idea of a ‘memo’ is similar to that of keeping field-notes – as we saw in Chapter 9. These memos serve as memory joggers and are useful, at a later date, when the researcher is writing up a report of his or her data.


Stage two


The interview transcripts are read through and notes made throughout the reading on general themes that appear in the transcripts. The aim, in general, is to become immersed in the data and to get to know it very well. Examples of general notes made, at this stage, might be as follows:



Early general notes in qualitative data analysis



  • There seems to be a lot of discussion about student attitudes to teaching methods.
  • Many respondents seem to worry about being understood!
  • Some respondents are positive about reflective practice while others seem very against it ….

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Mar 24, 2017 | Posted by in NURSING | Comments Off on A Pragmatic Approach to Qualitative Data Analysis

Full access? Get Clinical Tree

Get Clinical Tree app for offline access