Analysing qualitative data

Once you have collected your data, you need to analyse it! I have already touched on two types of qualitative analysis (content and thematic analyses) in an earlier post, but there are many more types of analysis than just those two. In this post, I will talk about some of the other types of qualitative analysis, but this list is by no means exhaustive.

Hierarchical content analysis

Hierarchical content analysis, like most other types of qualitative analysis, requires some kind of coding. In this type of analysis, codes are usually short, simple and precise key words. These key words are chosen to represent segments of primary content and "essence" of the data. Once codes have been identified, the researcher then identifies patterns and the way that the patterns interplay with one another, make comparisons and contrasts, divide the data into categories, and so on.

Hierarchical content analysis can be summarised in six main steps:

1. Immersion
2. Searching for, identifying and labelling themes
3. Connecting and ordering themes
4. Cross-checking to verify that all themes and categories are represented
5. Confirmation (possibly by another investigator)
6. Presenting results (usually in table form)

Grounded theory analysis

In grounded theory analysis, researchers generate theory from the data, rather than imposing theory on the data. Grounded theory analysis emphasises analytical techniques, such as constant comparison, in which emerging data is interpreted in relation to data that has already been collected. There are various kinds of grounded theory analysis, such as Glaserian, Straussian, and constructivist. Below are the main steps in Straussian grounded theory analysis:

1. Open coding with short, specific codes. Intense line-by-line coding is performed. Concepts are examined in terms of their properties and arranged. During this stage, it is important to remain open-minded.
2. Axial coding. Categories are related to sub-categories and a core category (or perhaps categories) is identified. During this stage, researchers may develop a conditional matrix, which visually depicts the relationship between concepts and conditions under which they occur. Researchers using the "paradigm model" may focus their attention on the context and "interactional" strategies that underpin the core category.
3. Selective coding / theoretical integration. During this stage, there is further linkage of categories to the core category in the pursuit of a coherent story. This stage continues until there is evidence of theoretical saturation, which in the case of grounded theory analysis, means that there are no new ideas or insights that can assist in the development of the theory.

Thematic analysis

I have already covered thematic analysis here. In thematic analysis, there is an emphasis on interpretation as researchers identify, analyse, interpret and report patterns. Writing up the data is also considered to be an essential part of the analysis. Here are the six main steps of thematic analysis:

1. Immersion
2. Generating initial codes
3. Searching for and identifying themes to create initial "thematic maps"
4. Reviewing themes
5. Defining and naming themes
6. Writing the report (which is another opportunity to refine the analysis)

Interpretative phenomenological analysis

Interpretative phenomenological analysis aims to understand and "give voice" to the concerns of participants, and contextualise and make sense of such concerns. The focus is on the understanding and experience of the participants. The main steps in interpretative phenomenological analysis are as follows:

1. Searching for themes in one individual's case (rather than across all cases)
2. Identifying themes
3. Connecting themes
4. Producing a table of themes
5. Continuing analysis with other cases
6. Writing- translating themes into a coherent account

Narrative analysis

Narrative analysis, as the name suggests, focuses on stories, which are a primary source of qualitative data. Narrative analysis aims to work with the whole story intact and examine how stories are put together. There are many types of narrative analysis:

• Thematic narrative analysis: Examines the core pattern by looking at the content
• Structural analysis of narrative types: Examines how stories are put together
• Performative analysis: Examines how talk is relationally and interactively produced and performed as a narrative
• Analytical bracketing: Analytically moves back and forth between the "whats" and "hows" of the narrative

Discourse analysis

Discourse analysis has a broader focus on language, and aims to attend to the "business" that is done in talk. There are several types of discourse analysis. Foucauldian analysis understands discourse as a process of constructing objects and subjects, and is primarily concerned with ways in which discourse constructs subjectivity, selfhood and power relations. Critical discourse analysis examines the social and political context of discourse.

Visual analysis

Visual narrative analysis focuses on the production of images, the images themselves, and how different audiences interpret images.

Computer-assisted data analysis

Computers can also be used to assist with coding. I'm not sure whether this truly counts as a separate type of analysis, or rather just a tool to aid in the other types of analysis.

References

Sparkes, AC & Smith, B 2013, Qualitative Research Methods in Sport, Exercise and Health: From Process to Product, Taylor & Francis Ltd, London.