What is the criteria for a study to be mixed methods?
Morse & Niehaus, 2009
- Contain a core study
- An incomplete, supplementary study which expands the scope of the projects
What are the positives about a mixed methods design?
Morse & Niehaus, 2009
- Completion of a single research project more expeditiously and efficiently than conducting multiple methods designs through a series of related projects.
- To comprehensively address the research question sometimes a mixed methods design is needed. eg. in a qualitative study in order to measure an aspect of the phenomenon.
Name some issues with mixed-methods designs?
- Greene, 2006 - Quantitative & Qualitative research belong to different and incompatible paradigms.
- Morse & Niehaus, 2009 - problems with mixed-methods designs may lie with different ‘academic turf’ and the devaluation of the qualitative component.
- Morse & Niehaus, 2009 - No consensus about hot to evaluate different mixed-methods designs, some researchers suggest evaluating quant and qual using their separate criteria, others argue the study is more than just the sum of its parts.
What are the points of interface for a mixed-methods study?
Morse & Niehaus, 2009.
- In the analysis of the core component.
- In the results narrative of the core component.
When considering Sampling, what should be considered?
Mertens, 1998
- Who has access to the information.
- The characteristics of the people who have experienced the phenomenon - age, disability, etc - any elements which may prevent meaningful participation.
- Type of information needed - eg. fact based information a carer/organisation may be able to give you the information. However, if wanting attitudinal information a carer may not be able to share this accurately.
- What are they hoping to find - therefore homogeneity or variability of sample.
Explain probability Sampling
Random Sampling with a lottery procedure
Explain systematic Sampling
Choose every (eg.10th) participant on a population list
Explain Stratified Sampling
Added through sub-groups eg. male and female. Disproportionate or proportionate sampling from subgroups.
Explain Cluster Sampling
eg. choose a random city block, classroom and sample all within this group.
Explain Purposeful Sampling
Choosing particular people for reasons eg. increasing or decreasing variability of the sample.
Explain Snowball Sampling
Also known as chain sampling - starts with key informants/experts and they refer to other participants.
What is a Coverage error?
Participants should be in a sampling frame but are not there / people are ineligible for the sampling frame.
What is a nonresponse error?
When someone refuses to be interviewed or doesn’t complete the questionnaire or cannot be reached.
What is a Sampling error?
Each sample drawn from the population is somewhat different from another sample.
Why is it useful to use a larger sample size for a qualitative IPA study?
It will reduce the risk of the sample being biased by atypical experiences or viewpoints. Forrester, 2010.
Henry, ___
‘Non probability samples add uncertainty when the sample is used to __________
Henry, 1990
‘Non-probability samples add uncertainty when the sample is used to represent the population as a whole’
Stake, ___
‘The first criterion should be to maximise what we can ____. Given our purposes, which cases are likely to lead us to ____, _____, perhaps even to modifying _____?’
Stake, 1995
‘The first criterion should be to maximise what we can learn. Given our purposes, which cases are likely to lead us to understandings, assertions, perhaps even modifying generalizations?’
What 2 things does Stake suggest we need to consider about sampling? Stake, ____.
- Accessibility is important - time and access are limited.
- Consider how contexts may aid or restrict our learnings.
Who talks about saturations points?
Yow, 1994 ‘ When you have heard the same story for the 20th time and you are not getting any new information, you probably have heard enough.’
What partly determines our ability to detect statistically significant results?
The amount of variability in the sample
Less variability from larger sample size = greater sensitivity.
More variability from smaller sample size = less sensitivity.
