INTRODUCTION TO MANAGING AND ANALYSING QUALITATIVE DATA Flashcards
What is the first stage of any qualitative interview?
Description of the data. This relatively superficial approach can be useful in producing a rapid (if simplistic) account of a particular phenomenon of interest. There are three main pieces of work required:
Read and annotate transcripts: this is the most basic stage. Here you do not provide an overview of the data, but make preliminary observations. This is particularly useful with the first few transcripts, where you are still trying to get a feel for the data.
Cut and paste: probably most common method of descriptive data presentation. Identify themes that run across many different interviews, cut them out and paste them into a document that contains all quotations related to that theme. Note that by doing this, it can mean the context of the data is lost. The identification of themes is part of the process of CODING. You can make a physical mark in the margin of the transcript to indicate text that belongs to a particular code.
Summary of the interviews: make a summary of each interview as an aide memoire for the rest of the analysis.
GROUNDED THEORY
This involves the “discovery of theory from data” (Glaser and Strauss, 1967:1). It is a more in-depth approach than the descriptive methods described above. What exactly this involves is a matter of some debate. However, core to most grounded theory approaches are a meshing of theorising and data collection (i.e. formulation of question / collect data / generate categories / more data collection / adjust categories / data collection etc). The process continues until the categories are “saturated” (i.e. that collecting more data will not add any new knowledge and researcher feels assured of the categories’ meaning and importance).
The presumption in grounded theory is that the codes move from the particular to the general, the simple to the complex, the concrete to the abstract, the thematic to the conceptual, and the descriptive to the analytical. You should end up with a dense formal theory which is grounded in the data.
Open coding - grounded theory
To do this, you begin with OPEN coding. This aims to ‘open up’ enquiry. Data are broken down, conceptualised and grouped together into subcategories. During open coding, you identify different types of data of interest such as: topics discussed; the ways people talk about something; or things relating to existing theory, such as existing sociological constructs. Guidance to doing open coding often emphasises the following: (i) look for ‘in vivo’ codes (the language and concepts participants use); (ii) undertake some detailed line-by-line analysis to get going; (iii) provisionally name each code; (iv) question the code thoroughly; (v) look for comparative cases; and (vi) ask yourself ‘what’s going on here?’.
Selective coding, second phase coding
After this phase, you move to SELECTIVE coding. This consists of intense coding/analysis of the categories which are central to the research project. All subordinate and subcategories become linked with the core category. Throughout the process, you need to go back and forth continually, reading the pieces of transcript identified with specific codes, then going back to the interviews, ensure the text is not misunderstood because of its lack of context, then returning to the specific coded material that will comprise extracts from several interviews etc. The process is ‘iterative’ – that is, you repeat the stages many times, continually refining your codes as you go along until they represent the data adequately. This can take a very long time.
FRAMEWORK ANALYSIS - why was it developed?
Framework is an approach designed by the National Centre for Social Research specifically for policy orientated projects, where the research question is determined at the outset, and tends to be less inductive than grounded theory.
Stages of framework analysis
Framework analysis comprises a number of distinct phases:
(i) Familiarisation (reading transcripts/listening to tapes);
(ii) identifying a thematic framework (and developing a coding scheme);
(iii) indexing (codes systematically applied to the data);
(iv) charting (rearranging data according to the thematic content in a way which allow within and between case analysis; and
(v) mapping and interpretation (defining concepts, mapping range and nature of phenomena, creating typologies, finding associations, providing explanations, developing strategies).
SUBJECTIVITY AND QUALITY IN QUALITATIVE ANALYSIS
It is vital to see the researcher as part of the creation of the data (collection and analysis), and for this reason to emphasis ‘reflexivity’. This discussing and accounting for the researcher’s role in the collection and interpretation of data, alongside recording researcher experiences and interpretations in field notes.
The quality of analysis will be improved by:
(i) making sure your analysis is systematic and comprehensive;
(ii) using colleagues to help analyse, code and discuss findings; and
(iii) taking a critical approach that constantly looks for contradictory evidence and does not ignore “deviant” cases.
Key objectives of qualitative analysis can include:
(i) identifying concepts;
(ii) describing typologies;
(iii) describing the form and nature of phenomena (e.g. processes, systems, attitudes, behaviours);
(iv) exploring associations (e.g. between attitudes, behaviours, experiences);
(v) forming explanations (explicit or implicit) of why phenomena occur;
(vi) generating new ideas, theories, strategies; and
(vii) illuminating accounts.
