3. Qualitative research Flashcards
Key functions of qualitative research
Contextual: describing ‘what’s there’ - experiences, attitudes; the nature of something.
Explanatory: exploring why something is happening.
Evaluation: assessing how well interventions work.
Generative: developing theories/strategies/actions/support.
Where can we obtain naturally occurring data from?
> Conversations and discourses
> Observations (participant/non-participant)
> Documents, texts, images, media
How can we generate [qualitative] data?
> Interviews
○ Structured/in-depth/narrative
○ Focus groups
> Questionnaires
List some sampling stratgies
> Convenience sampling
Random sampling
Purposive sampling
Pros and cons of convenience sampling
PROS:
○ Participants are easy to identify and contact
○ Efficient and straight forwards
CONS
○ Unlikely to represent wider population
○ May not include all the characteristics relevant to the research question
○ —> results may be difficult to generalise
Pros and cons of random sampling
PROS:
○ Every member of the population has an equal chance of selection
○ Reflects the characteristics of the wider population (if the sample is large enough) –> generalisable results.
○ Reduced risk of researcher bias.
CONS:
○ Large sample may be unrealistic for qualitative data collection and analysis.
○ May under represent groups/characteristics relevant to research questions.
Pros and cons of purposive sampling
PROS:
○ Participants are selected according to the characteristics important to the research question.
○ Requires expert knowledge of the population; aims to reflect diversity within the population.
○ Highly relevant data.
○ Sampling strategies can be adapted as research progresses.
CONS:
○ Risk of researcher bias.
○ Results may not be generalisable.
What needs to be considered with sample sizes?
> Sample sizes in qualitative research tend to be small – so they need to be carefully chosen.
> The aim is to reach ‘saturation’ (where themes and patterns in the data are repeating). Not looking to draw statistical inferences.
> Budgets and resources: qualitative analysis is time consuming.
List types of interviews
> Unstructured
Semi-structured
Structured
what are the characteristics of Unstructured interviews
- Not set lists of questions or topics
- Participants tell their story in their own way
- Narrative or biographical data
- Deeper understanding of personal and life experiences
what are the characteristics of semi-structured interviews
- Flexible use of topic guide
- Explore emergent themes are concepts
- Probe for clarification and further information
- Explore respondent’s perspectives and opinions in depth
what are the characteristics of structured interviews
- Standarised set of questions
- Concepts defined in advance
- Fact finding, hypothesis testing and large surveys
- Direct comparisons between responses
What are topic guides and what can they include?
> Aid the memory of the interviewer
> Topic headings and subheadings (and prompts if needed)
Topics related directly to the research question
Not a rigid list of questions; flexible order and flexible ‘conversational’ wording
How can we identify topics (for a topic guide)?
> Familiarisation with subject area e.g. literature review, discussions with stakeholders
> Brainstorming e.g. team discussion, pilot investigation
> Reflect on research beliefs and assumptions, identify and minimise research bias
How can we develop a topic guide?
Broad structure:
○ Intro -> opening topics -> core topics -> closing topics
Wording:
○ Single words or phrases rather than questions
- Easy to read at a glance
- Flexible
Natural language:
○ Minimise researcher bias
○ Consider potential misunderstandings
Follow up questions/ prompts/ other instructions to the interviewer may be helpful
Things to consider when starting an interview
- Ensure ideal physical conditions • Quiet private space • Seating arrangements • Recording equipment set up and tested - Initial introduction • Recap project aims and outline • Reconfirm consent - Reassure about confidentiality and anonymity
Things to consider during the interview
- Ask clear questions
- Give respondent time to reply
- Follow the topic guide flexibly
- Make notes of issues you would like to return to and explore further
- Probe until you think that the respondent has replied to the question as fully as possible
- Show interest in what respondent says
- Avoid leading questions
- Don’t express your opinion on responses
- Use neutral responses
Things to consider when ending the interview
- Try to end on a positive and complete note
- Use closing questions
- Thank the interviewee, re-affirm confidentiality
- Explain again how the information they have given will be used
List some useful interview probes
- Encouraging respondent to continue through the silence and nodding or min response
- Inviting the respondent to elaborate
- Stimulating further thought
- Giving permission to express a controversial view
What are focus groups?
A research technique that collects data through group interaction on a topic determined by a researcher.
Structured - formal/informal
What are the key features of focus groups?
> Small groups (7-9) who ‘represent’ the sample population.
> Group dynamics are used to explore the participants’ views.
> It’s run by a moderator.
> Venue and time: aim to be accessible, and minimise distractions.
> Capturing the data: audio equipment, video equipment, field notes, 2nd researcher.
What is the role of a moderator in a focus group?
> To ensure that all participants contribute fully.
> Give guidance and explanation.
> Controlling the balance between individual contributors.
> Questioning and probing.
> Observing non-verbal behaviour.
What are some ethical issues with qualitative analysis?
> Informed consent > Anonymity/confidentiality > Risk of harm ○ Potential to cause distress/embarrassment ○ Vulnerability of participants
What are some approaches to qualitative analysis?
> Thematic coding approach ○ Grounded theory ○ Framework analysis > Conversational analysis > Discourse analysis
- Managing the data - organise the data, it will often be rich and detailed.
- Making sense of the data through descriptive or explanatory accounts.
What is ‘Framework’ analysis?
A tool for qualitative data analysis
> Matrix based method for ordering and synthesising the data.
> Derived from ‘thematic framework’ - used to classify data according to key themes, concepts and emergent categories.
Key requirements when analysing data
> Organise your material - need a ‘system’ to help you organise the data.
> Systematic and comprehensive - need to cover all the material.
> Within and between case searches - need to compare different participants’ stories, but also look within one participant’s story.
> Remain ‘true’ to actual data - ideas, patterns, emerging concepts to come from the data, rather than imposing abstract theories onto the data.
What are the 6 stages of ‘framework’ analysis?
- Familiarisation
- Construct thematic framework
- Indexing
- Charting
- Sort and synthesis labelled data within the charts
- Develop descriptive and explanatory accounts
Framework analysis: what’s involved in the ‘familiarisation’ stage?
Identify initial themes through familiarisation with the data.
> Gain an overview of all the material. > Immersion in the data ○ Read transcripts ○ Listen to recordings ○ Study observational notes > List key ideas and recurrent themes ○ Begin to conceptualise data
Framework analysis: what’s involved in the ‘construct thematic framework’ stage?
> Develop an index of key issues, concepts and themes.
> Organise index into
• Main categories
• Subcategories
> Index will provide mechanism for coding data
May need to modify index as analysis proceeds
Framework analysis: what’s involved in the ‘indexing’ stage?
> Apply the thematic index (tag/label) to the raw data.
• Reread transcripts/observations
• For every line or passage:
○ Infer the respondent’s meaning
○ Annotate margin with index codes
• One passage may have more than one code
> Show emergent associations within data.
Framework analysis: what’s involved in the ‘charting’ stage?
> Aim is to synthesize and summarise data.
> Write distilled summary of each passage of coded data onto correct column of chart
• Don’t copy long quotes verbatim
• Summarise meaning (using pps’ lang)
• Annotate with page/line numbers (for ref)
> Don’t impose interpretations on data at this stage
Framework analysis: what’s involved in the ‘sort and synthesis labelled data within the charts’ stage?
> Sometimes known as ‘mapping’
> Interpretation of whole data set after all interviews have been coded and charted
* Review the charts and research notes * Compare and contrast accounts and perceptions of different interviewees * Search for patterns or associations within data * Weigh up salience and dynamics of patterns
> Build theory to describe and explain the data
[see later for using computers to assist with data management]
Framework analysis: what’s involved in the ‘develop descriptive and explanatory accounts’ stage?
DESCRIPTIVE
> Identifying key dimensions
> Mapping the range of each phenomenon
> Includes the actual, verbatim language of interviewees
EXPLANATORY
> Finding patterns of association in the data and accounting for why they occur
> Rarely cite a single cause/reason; clarify nature and inter-relationship of contributory factors and influences
Why computer assisted data management in framework analysis?
> CAQDAS: computer assisted qualitative data analysis software
- Very useful for managing large data sets
- Speeds up some stages of analysis and allows flexibility e.g. to modify coding categories as you analyse
- Provides powerful tools to explore and visualise data
How can we assure validity and reliability in qualitative research?
Respondent validation: Taking findings back to the participants to see if your interpretation is consistent with their experience.
Quality of the work : Appropriate design and conduct of the research.
> Audit trail
Reflective journal
Triangulation
What is triangulation and what types are there?
Triangulation is often used to indicate that two methods are used in a study in order to check the results of one and the same subject.
○ Data triangulation ○ Investigator triangulation ○ Theory triangulation ○ Methodological triangulation
Benefits of triangulation
○ Provides richer description and deeper understanding
○ Improves reliability and validity
○ Reduces bias
○ Allows cross checking of data and of interpretation
