Upload
dylan-hallmark
View
215
Download
1
Embed Size (px)
Citation preview
SI0030 – Social Research Methods
Week 10Luke Sloan
Mixed Methods
Introduction
• Mixed Methods as Methodology
• Mixed Methods Designs
• Research Design Challenges
• Activities
Mixed Methods as Methodology I
“[Mixed methods] actively invites us to participate in dialogue about multiple ways of seeing and hearing, multiple ways of making sense of the social world, and multiple standpoints on what is important and to be value and cherished.”
Source: Greene (2007:20)
Mixed Methods as Methodology II
“As a method, it focuses on collecting, analysing, and mixing both quantitative and qualitative data in a single study or series of studies. Its central premise is that the use of quantitative and qualitative approaches, in combination, provides a better understanding of research problems than either approach alone.”
Source: Creswell and Plano Clark (2007:5)
Mixed Methods as Methodology III
• When little is known or written• Client specifications• Collection instrument feedback• Survey question design/selection• Interview theme development• Understanding a sample• Trouble-shooting non-response• Exploring quantitative phenomena• Generalising qualitative themes
Notation Example Application
What the Notation Indicates Key Citations
Shorthand: Quan, Qual
Quan strand Quantitative methods Morse (1991, 2003)
Uppercase: QUAN, QUAL
QUAL priority The qualitative methods are prioritised in the design Morse (1991, 2003)
Lowercase: quan, qual
qual supplement The qualitative methods have a lesser priority in the design Morse (1991, 2003)
Plus sign: + QUAN + QUAL The QUAN and QUAL methods occur concurrently Morse (1991, 2003)
Arrow: QUAN qual The methods occur in a sequence of QUAN followed by qual Morse (1991, 2003)
Parentheses: ( )
QUAN(qual) A method is embedded within a larger design/procedure or mixed within a theoretical/program-objective framework
Plano Clark (2005)
Double arrows:
QUAL QUAN The methods are implemented in a recursive process (QUAL QUAN QUAL QUAN etc.)
Nastasi et al. (2007)
Brackets: [ ] QUAL QUAN [QUAN + qual]
Mixed methods [QUAN + qual] is used within a single study or project within a series of studies
Morse & Niehaus (2009)
Equal sign: = QUAN qual = explain results
The purpose of mixing methods Morse & Neihaus (2009)
Source: Creswell and Plano Clark (2011:109)
Summary of Notations Used to Describe Mixed Methods Designs
Mixed Method Designs IDESIGN: DESCRIPTION:
Converge ResultsQUAN + QUAL
Collecting & analysing two independent strands of quant. and qual. data in a single phase. Looking for convergence, divergence, contradictions or relationships. Equal emphasis.
Explain ResultsQUAN qual
Two strands administered in sequence with quant. strand running first with greater emphasis in addressing purpose of study, qual. used to explain quant. results.
Generalise FindingsQUAL quan
Exploratory design with two strands in sequence. Qual. Running first with greater emphasis in addressing purpose of study, quant. to assess extent of generalisation to a population.
Enhance ExperimentQUAN (+qual)
Embedded design with secondary qual. strand within larger quant. experiment. Qual. occurred during experiment and enhanced conduct and understanding of experiment.
Source: Creswell and Plano Clark (2011:110)
Mixed Method Designs IIDESIGN: DESCRIPTION:
Converge ResultsQUAN + QUAL
Collecting & analysing two independent strands of quant. and qual. data in a single phase. Datasets are analysed separately and the two are then merged. Looking for convergence, divergence, contradictions or relationships. Equal emphasis.
CONSIDERATIONS: RECOMMENDATIONS:
Will the two samples include different or the same individuals?
Use same individuals of you intend to compare the quant and qual datasets
Will the sample be of the same size?
Equal size may be hard to achieve (low ‘n’ for quals) – limitation of study or inevitable?
Will the same concept be assessed qualitatively and quantitatively?
Parallel questions to measure the same social concept in different ways (for convergent design)
Will the data be collected from two independent sources or from a single source?
You may use the same sample, but how many data collection ‘events’ and what instrument?
Mixed Method Designs IIIDESIGN: DESCRIPTION:
Explain ResultsQUAN qual
Two strands administered in sequence with quant. strand running first with greater emphasis in addressing purpose of study, qual. used to explain quant. results.
CONSIDERATIONS: RECOMMENDATIONS:
Will the same/different individuals be used in both samples?
Due to explanatory nature qual phase individuals must be chosen from quant phase partcipants
Will the sample be of the same size?
Qualitative follow-up will be smaller, as is standard in qualitative projects
What quantitative results will be followed-up?
Depends on significant results, client specifications, gut feeling – research question?
How will follow-up participants be selected?
Depends – might be informed by interesting quantitative phenomena or random
Mixed Method Designs IVDESIGN: DESCRIPTION:
Generalise FindingsQUAL quan
Exploratory design with two strands in sequence. Qual. Running first with greater emphasis in addressing purpose of study, quant. to assess extent of generalisation to a population.
CONSIDERATIONS: RECOMMENDATIONS:
How many individuals should be in the quantitative follow-up phase?
Use a different and larger sample – largely determined by sample/population dynamic
What qualitative results will be used to inform quantitative data collection?
Depends on the study - key themes and phenomena, typological groups or demographics
How do you develop a good instrument for data collection?
Think about operationalising qualitative concepts and scale measurement, pilot your instrument
Mixed Method Designs VDESIGN: DESCRIPTION:
Enhance ExperimentQUAN (+qual)
Embedded design with secondary qual. strand within larger quant. experiment. Qual. occurred during experiment and enhanced conduct and understanding of experiment.
CONSIDERATIONS: RECOMMENDATIONS:
Why and when should the embedded data be used in the study?
Justify why embedded data is useful (to capture subtlety, as a passive auxiliary approach), is it simply an insurance? Is this acceptable?
Will embedding of a second dataset introduce bias?
When running experiments alternative data collection can be intrusive so think about method (e.g. use diaries during experiments)
How should the qual and quant strands be tied together?
Embedded method should have enhanced the study, not fundamentally altered it. Use qual to anecdotally support quants (e.g. case study?)
Research Design Challenges I• Collecting data is easy - but how organise and analyse it (SPSS? NVivo?
Excel?)
• Should variable be comparable? How are they measured?
• Matching qualitative and quantitative responses to individual respondents or comparing different cohorts?
• How to integrate both data types into the research report – anecdotal, supportive, quotes or themes?
• The key is to use one approach to support the other e.g. explain statistical trends with contextual ideas
Research Design Challenges II
• Methodological Tribalism
• Skills and Knowledge
• Time and Resources
• Convincing Others
Activity I
Widening participation in Higher Education: what are the barriers to HE and how can they be
overcome?
Understanding smoking: how do people consume tobacco products and why?
Activity II
• Could mixed methods enhance your project?
• Think about how a mixed methods research design could answer your research question
• Use the summary notation to describe your own research design
• Discuss this in your groups alongside how you could reconcile a mixed methods approach