Just as Nicholas Grimshaw's airy interior design for the LSE's New Academic Building reveals an academic institution in transition during a summer vacation period, our newly acquired Qualitative Analysis Methods training should give us the tools to clearly see the themes, content, narratives and discourses in the data we will analyse in our own research projects.
As my two weeks at the LSE and regular lectures and seminars here draw to a close today, I'm wondering how much of the learning I've actually absorbed. If asked what I've learned, I'd struggle to string it into a coherent narrative, but as I'll be tested shortly on my ability to apply my new-found knowledge to some simple theoretical challenges - I'll give it a go...
Qualitative Research Design includes methods suited for:
In our first week here, we were introduced to a number of qualitative research data collection methods which are more or less suited to particular types of enquiry:
Qualitative Research Design includes methods suited for:
- phenomenological studies, understanding context
- ethnographic - structure and culture of communities, situational contexts
- representing the actors' voices
- subjective, lived experiences
- life histories, stories
- Case studies - proof of theory, etc.
- Grounded Theory building
- Discourse Analysis (esoteric!)
- Action Research (Participatory methods)
In our first week here, we were introduced to a number of qualitative research data collection methods which are more or less suited to particular types of enquiry:
Collection Method |
Useful for: |
Qualities: |
Issues: |
Structured Interviews |
eliciting participants' knowledge, experiences, perspective on a topic, or key informants |
private space to 'air' thoughts / feelings, (1:1 i-views) include non-verbal cues |
The Interviewer may be 'present' in data Reflexivity needed Affected by situation |
Unstructured Interviews |
Life history / narrative / episodic stories |
More conversational than structured i-views |
Harder to align with existing theory? Better for 'Grounded'? |
Focus Groups |
insights / attitudes / values / beliefs / opinions / norms / language |
the group's composition and interactions can also be analysed |
'in-group' can lead to consensus, can also stifle dissent. beware of comparing! |
Case Studies |
may reveal features of a wider class of phenomena; generates & tests theory; |
Limited number of cases; depth vs breadth; multiple DC qual/quant methods; |
case selection needs to be rigorous and mapped to the study's purpose for the research to be valid |
Document Analysis |
As Resource or Topic? A tool of authority or power, content, themes, narratives, discourse; of doc's origins, purposes & uses; contestation? |
Supports other methods: Critical Discourse, case studies, embodiment of practice |
Relatively thin depth, may lack context; some docs difficult to access, though easier with www |
Ethnography |
observation; bounded communities and settings; |
Holistic analysis possible, can include interviews |
requires time to become immersed |
Visual Materials |
Photo-voice; photo-elicitation; participants story-telling in drawings; |
articles similar to documents; embodiment of (artistic) practices; fits ethnography |
Are photos or the text narrative more useful? Adults unconfident at making drawings; may require training for ptp |
Sampling:
Start by defining inclusion & exclusion criteria - my study's participants will have experience of cycling to or from some kind of public transport (PT) on several occasions within the past 18 months. People who don't cycle are outside the scope of my study, as are people who don't use public transport, although they may be included where my research explores whole household experiences of one individual's cycling with PT. The boundary of relevant experience may be permeable where people with more limited experience are able to comment critically on a one-off experiment with combining modes.
In Qual research, 'purposive' sampling is typically used where the research is interested in particular characteristics (random sampling is often used in large-scale statistical methods).
'convenience' sampling acknowledges the constraints on purposive S. due to limitations of resources, etc.
Snowballing used within population groups to increase numbers of participants, but shares some of characteristics of convenience sampling.
The population sample needs to be matched to the Research Questions, according to the particular characteristics of interest to the study and the relevance of their experiences.
Writing this both before and during the session leads me to divert my attention from some of Simi's discourse, and answer a different question to the one she asked me ('"I'd need the taxi fare..." D'oh!). Conclusion - I'm not great at multi-tasking!