Main sources of qualitative data
- Observations: description of activities, behaviours, interactions, or any other observable aspect.
- Interviews. Open questions and other forms. Do not take anything for granted. (Focus groups, one-to-one, …)
- Documents. Written material from individuals or organisations.
Can be primary source of data (emails, books, social media…)
Sampling strategies in qualitative research
Purposeful sampling
The main goal is NOT representativeness, but rather generating the best understanding possible of the phenomenon under examination.
- > Selecting information-rich cases
- > Sample size is not known in advance
- > Sampling is finished when theoretical saturation is reached.
Types of purposeful sampling
- Outlier sampling: learning from unusual manifestations of phenomenon.
Intensity sampling: information rich-cases that experience phenomenon intensely (serial entrepreneurs).
Maximum variation sampling: pick wide range of cases.
Homogeneous sampling: reduce variation to simplify analysis. (freshmen only)
Typical case sampling: picking standard cases.
Critical case sampling: if “true” here, true everywhere.
Snowball sampling: ask to current case the next case.
Criteria sampling: sample base on criteria.
Theory-based sampling:
(Dis)confirming sampling: confirm to refine, disconfirm fro external validity.
Opportunistic or emergent sampling: follow new leads generated during fieldwork.
Purposeful random sampling: systematic way of generating samples in population of interest. To increase credibility not representativeness.
Convenience sampling: DO NOT USE.
Stratified random sampling: sample from subgroups.
- Mixed purposeful sampling: a combination of any of the above.
Risk of researcher’s role in qualitative research tools
Observation, subjectivity.
Interviews, requires mastery and technique.
Field notes, both facts and observations (separated).
Focus group, Researcher vs moderator.
Video Ethnography, Coding the data is crucial.
Virtual Ethnography, Ethical issues.
Well known non-ethical experiments
- The Milgram Experiment, Yale, 1963
Electric Shocks. - Stanford Prison Experiment, 1971
Result: The Belmont Report, Basic Ethical Principles in Research with Human Subjects.
What does the Belmont Report consist of?
- Respect of persons
- Beneficence
- Justice
Process phases in Qualitative Data Analysis with interviews
Transcribe interviews.
Coding, process of abstraction.
Approaches to Coding
- Thematic coding: Development of main themes and patterns, when comparing different cases.
Main steps:
- Summarise each case.
- Extract motto, demographic, key points
- Identify common patterns and categories - Theoretical Coding: generate a high level understanding by abstracting one layers after the other.
Main steps:
- First-order labels/categories.
- Axial Coding (Second-order labels)
- Selective coding (identify core categories)
- Provide possible explanation
Solutions to Validity dangers in qualitative research
- Conclusion validity
- Document the research project
- Standardise field notes, train for interviews, repeat data collection. - Internal validity
- Address rival explanations
- Explain negative evidence - Construct Validity
- Triangulation of data sources, investigators, theories, methods.
- Have key informants review the draft reports. - External Validity.
- Research is usually contextual, if generalising, provide convincing theoretical explanations.
