Qualitative Data Analysis: Concepts and Approaches (W2)

Definition of qualitative analysis

Qualitative analysis is the search for patterns in textual, visual, or artefact data and the explanation for why these patterns exist in the first place. While statistics rely on the use of probability theory to estimate population inferences, qualitative analysis uses theory to decontextualise the raw data in segments and recontextualise them in categories to generate concepts, identify relationships, and formulate hypotheses for theory development. Qualitative analysis may be conducted inductively by identifying conceptual categories directly in the data, or deductively by applying predefined theoretical notions onto the material. A mixed approach to analysis - when both induction and deduction are used in different points in time in the analytical process – is increasingly popular since it makes use of the researcher’s theoretical sensitivity and allows space for meaning making and category generation. 

Criticisms about qualitative analysis

Historically qualitative data analysis has been criticised for being opaque and subjective given that it is sometimes difficult to see how researchers went from hundreds of interviews pages to a handful of conclusions, since no discussion is provided about what the researchers actually did when they analysed their data. For this reason, qualitative analysis has been associated with a ‘black box’ problem: we know what data were collected and what results these yielded, but nothing in between. There is also a misconception that qualitative analysis merely involves the identification of themes in the data. The real analytical tasks researchers should be concerned about – that is, examining commonalities and differences across units of analysis, discovering what patterns exist amongst these and synthesizing those in a storyline, model, or schema - is most frequently than not absent in qualitative studies. Conversely, the widespread myths that the ‘method will emerge from the data’ or that researchers’ tacit knowledge will lead them to ‘make sense’ of the data, account for generally poor analytical practices and opaque reporting.

Contribution of this course

This course introduces some of the key concepts and stragegies to plan, conduct and report QDA in a transparent, traceable, and auditable manner. You will learn the best fit between the research questions you ask, your study purpose, the data you intend to collect, and the approach to analyse your qualitative data. Whether you intend to conduct interviews or focus groups, collect policy papers or press articles, use Internet data, or conduct open-ended survey questions and wonder how you should analyse your data, this course is for you.

Objectives

The course’s learning objectives are to:

1. Review the problems and prejudices associated with QDA
2. Illustrate the relationship between research design and QDA
3. Learn different approaches to code qualitative data
4. Seek patterns across themes, cases and contexts
5. Generate graphic displays to present findings

Expected outcomes

At the end of this course, participants will be able to:

1. Recognise the ‘black box’ problem in many qualitative studies
2. Design a QDA plan congruent with one’s research design
3. Choose the right strategy to code qualitative data
4. Apply the right technique to seek patterns across the data
5. Communicate qualitative findings effectively

Course schedule

Day 1 – Foundations. The course opens with a lecture on the foundations of qualitative data analysis: definitions, history, problems and challenges ahead. The qualitative analysis cycle is introduced as a device to understand that qualitative analysis often, if not always, occurs iteratively between the phases of data coding, patterns seeking, and results display. We then look at the function that data analysis plays in a qualitative study by connecting the analysis phase with the rest of the research design: that is, with a well-formulated research question, a clearly-defined research purpose, a congruent data collection, and an effective presentation of findings. Our attention then turns to a number of approaches to analyse qualitative data. In turn, the aim, specificities, and requirements of thematic analysis, qualitative content analysis, grounded theory, cross-case analysis, and analytic induction are presented. The second half of the class is a workshop where students develop the analytical plan of their study.

Day 2 – Data coding. Concepts and practice associated with coding qualitative data are introduced on day 2. In qualitative research, coding is the process by which data are segmented in coding units and assigned a code. A code is a word, or short phrase, that captures what the data is about. A code can mean to a concept, a construct, an event, a place, or an organisation. Coding is the first phase of qualitative analysis, so knowing how to code one’s data and the outcome coding should yield is of central importance. Central notions to coding such as meaning unit, coding unit, codes and codebook, coding scheme, and coding outcomes are discussed. Students then have the opportunity to work on the development of their coding scheme and the coding protocol of their study, first using pen and paper and then in NVivo. The second half of the class is a workshop where students code interview data (or their own if they have any). NVivo’s coding functions are taught alongside features that enhance the development of an audit trail for transparency and auditability.

Day 3 – Patterns seeking. Day 3 addresses inductive and deductive approaches to seeking patterns across the data and to create second and third level constructs as the analysis reaches its end and formal conclusions are formulated. Depending on the study design, the tasks of seeking patterns and that of identifying relationships may take place while coding inductively or, by means of retrieval techniques where coding co-occurrence in cross-tabulated tables indicates an association between codes which should be investigated further for possible emerging relationships. The use of other patterns seeking techniques such as frequency counting, code clustering, code subsuming and code aggregation will be demonstrated in the light of the instructor own and other people’s research.  The second half of the class is a workshop where participants try the different techniques of seeking patterns using their own or sample data, so that relationships between codes may be uncovered, comparisons built, meaning generated, and interpretation supported.

Day 4 – Reporting findings. Day 4 covers best practices when reporting qualitative analysis and communicating research findings. In qualitative inquiry, reporting findings involves the presentation of data into a coherent structure, often supported with a narrative, so that a clear, unambiguous message is conveyed to the target audience. The types of analytical outputs in qualitative research - descriptive vs explanatory accounts – are introduced. Our attention then turns to the purpose of graphic displays for qualitative data: we see that models are best to illustrate conceptual integration, matrices are appropriate for cross-tabulated information, tables are well-suited to present typology, and diagrams work well to depict structure. We conclude the first half of the class by looking at the quality criteria when reporting qualitative research generally, and those concerned with communicating qualitative results in particular. The second half of the class is a workshop on the visualization displays available in NVivo and their application in students’ research context.

Day 5 - Master Class. The Master Class consists of participant presentations about how they intend to analyse their data in their current research. A short powerpoint must highlight the congruence between their research questions, the study purpose, data collection method, strategies for coding and analysis, and reporting strategy. Those at the start of their research are equally welcomed to present what they prospectively intend to do. Presentations are voluntary, are limited to 20 minutes, and do not form part of the formal evaluation system of the Summer School.

Assignment for 2 ECTS credits

Students can earn 2 ECTS credits upon the production and satisfactory marking of an academic essay. Deadline for the essay is August 19th 2016.

Basic knowledge of qualitative research required. No requirement of qualitative analysis.

NVivo 11 Pro for Windows will be used by the instructor to demo the application of the QDA process in a qualitative software environment. If you want to do the demo in NVivo, you must possess basic knowledge of the software as NVivo will not be taught in class. The refresher course Introduction to NVivo for Qualitative Data Analysis offers a comprehensive entry to NVivo for those who wish to use it. Participants may also use other qualitative software they’re familiar with, including Word or Excel, for the demo sessions.