Qualitative data analysis Principles of qualitative data analysis

Qualitative data analysis

Principles of qualitative data analysis • Important for researchers to recognise and account for own perspective – Respondent validation – Seek alternative explanations – Work closely with same-language key informant familiar with the languages and perspectives of both researchers and participants

Principles of qualitative data analysis • Context is critical i. e. physical, historical, social, political, organisational, individual context Dependence/interdependence – Identify convergence / divergence of views and how contextual factors may influence the differences

Principles of qualitative data analysis • Role of theory guides approach to analysis – Established conceptual framework – predetermined categories according to research questions – Grounded theory – interrogate the data for emergent themes

Principles of qualitative data analysis • Pay attention to deviant cases / exceptions – Gives a voice to minorities – Yield new insights – Lead to further inquiry

Principles of qualitative data analysis • Data analysis is a non-linear / iterative process – Numerous rounds of questioning, reflecting, rephrasing, analysing, theorising, verifying after each observation, interview, or Focus Group Discussion

Stages in qualitative data analysis • Interrelated rather than sequential • During data collection – Reading – data immersion – reading and rereading – Coding – listen to the data for emerging themes and begin to attach labels or codes to the texts that represent themes

Stages in qualitative data analysis • After data collection – Displaying – themes (all information) – Developing hypotheses, questioning and verification – Reducing – from the displayed data identify the main points

Stages in qualitative data analysis • Interpretation (2 levels) – At all stages – searching for core meanings of thoughts, feelings, and behaviours described – Overall interpretation • Identify how themes relate to each other • Explain how study questions are answered • Explain what the findings mean beyond the context of your study

Processes in qualitative data analysis 1. Reading / Data immersion – Read for content • • • Are you obtaining the types of information you intended to collect Identify emergent themes and develop tentative explanations Note (new / surprising) topics that need to be explored in further fieldwork

Processes in qualitative data analysis Reading / Data immersion – Read noting the quality of the data • • Have you obtained superficial or rich and deep responses How vivid and detailed are the descriptions of observations Is there sufficient contextual detail Problems in the quality of the data require a review of: – – – • How you are asking questions (neutral or leading) The venue The composition of the groups The style and characteristics of the interviewer How soon after the field activity are notes recorded Develop a system to identify problems in the data (audit trail)

Processes in qualitative data analysis Reading / Data immersion – Read identifying patterns • After identifying themes, examine how these are patterned – – Do themes occur in all or some of the data Are their relationships between themes Are there contradictory responses Are there gaps in understanding – these require further exploration

Processes in qualitative data analysis 2. Coding – Identifying emerging themes – Code themes that you have identified • • • No standard rules of how to code Researchers differ on how to derive codes, when to start and stop, and on the level of detail required Record coding decisions Usually - insert codes / labels into the margins Use words or parts of words to flag ideas you find in the transcript Identify sub-themes and explore them in greater depth

Processes in qualitative data analysis Coding – Identifying emerging themes – Codes / labels • Emergent codes – Closely match the language and ideas in the textual data • ‘Borrowed’ codes – Represent more abstract concepts in the field of study – Understood by a wider audience • Insert notes during the coding process – Explanatory notes, questions • Give consideration to the words that you will use as codes / labels – must capture meaning and lead to explanations • Flexible coding scheme – record codes, definitions, and revisions

Processes in qualitative data analysis Coding – Identifying emerging themes – Code continuously as data collection proceeds • Imposes a systematic approach • Helps to identify gaps or questions while it is possible to return for more data • Reveals early biases • Helps to re-define concepts

Processes in qualitative data analysis Coding – Identifying emerging themes – Building theme related files • Conduct a coding sort – Cut and paste together into one file similarly coded blocks of text – NB identifiers that help you to identify the original source See example on Clandestine Microbicide Use

Processes in qualitative data analysis 3. Displaying data i. e. laying out or taking an inventory of what data you have related to a theme – – – Conduct quantitative and qualitative analysis Capture the variation or richness of each theme Note differences between individuals and sub-groups Organise into sub-themes Return to the data and examine evidence that supports each sub-theme – Note intensity/emphasis; first- or second-hand experiences; identify different contexts within which the phenomenon occurs

Processes in qualitative data analysis 4. Developing hypotheses, questioning and verification – Extract meaning from the data – Do the categories developed make sense? – What pieces of information contradict my emerging ideas? – What pieces of information are missing or underdeveloped? – What other opinions should be taken into account? – How do my own biases influence the data collection and analysis process?

Processes in qualitative data analysis 5. Data reduction i. e. distill the information to make visible the most essential concepts and relationships – Get an overall sense of the data – Distinguish primary/main and secondary/subthemes – Separate essential from non-essential data – Use visual devices – e. g. matrices, diagrams

Processes in qualitative data analysis 6. Interpretation i. e. identifying the core meaning of the data, remaining faithful to to the perspectives of the study participants but with wider social and theoretical relevance – Credibility of attributed meaning • • Consistent with data collected Verified with respondents Present multiple perspectives (convergent and divergent views) Did you go beyond what you expected to find?

Processes in qualitative data analysis Interpretation – Dependability • • Can findings be replicated? Multiple analysts – Confirmability • Audit trail – Permits external review of analysis decisions – Transferability • Apply lessons learned in one context to another – Support, refine, limit the generalisability of, or propose an alternative model or theory