Mixed Methods Research
Created by Axel Burch
| Term | Definition |
|---|---|
descriptive research | seeks an overall summary of study variables |
correlational research | seeks to investigate relationships between study variables |
experimental research | seeks to systematically examine cause-and-effect relationships between study variables |
quantitative research | descriptive, correlational or experimental research
|
descriptive statistics | summary of data with measures of averages and variability |
inferential statistics | data with which one can make predictions or generalisations |
strengths of quantitative research | >replicable
>direct comparisons
>large samples
>uses hypothesis testing |
weaknesses of quantitative research | >complex
>superficial
>narrow focus
>lack of context |
qualitative research | gains deeper understanding of a subject, and can be used to develop ideas or hypothesis |
strengths of qualitative research | >flexible
>natural settings
>meaningful insights
>idea generation |
weaknesses of qualitative research | >unreliable
>subjective
>limited generalisability
>labour intensive |
mixed methods research | uses both quantitative and qualitative research methods |
when to choose MMR | >one data source is insufficient
>result need explaining or generalising
>secondary method can enhance a primary method
>employing a theoretical stance
>objective is best addressed with multiple projects |
fixed MMR | predetermined quantitative and qualitative methods with research and procedures implemented as planned |
emergent MMR | additional methods implemented due to issues arising while conducting research so a second approach is added after the initial study began |
convergent parallel design | qualitative and quantitative data is collected simultaneously and compared for similarities/differences, and the results are used to understand a research problem |
explanatory sequential design | qualitative results used to expand on quantitative findings |
exploratory sequential design | explores a phenomenon using qualitative data followed up by quantitative data (may be used to develop and test new instruments) |
embedded design | quantitative and qualitative data is collected simultaneously or sequentially, with one type supporting the other |
transformative design | using one of the four basic designs and encasing it within a transformative framework/lens |
multiphase design | builds upon the four basic designs to examine problems/topics through a series of phases or separate studies (single method study -> single method study -> MMR) |
MMR steps | >(objectively determined for each study)
>determine feasibility of MMR
>identify rationale
>identify data collection strategy and design
>develop research questions
>collect data and analyse |
strengths of the convergent parallel design | >intuitive sense
>efficient design
>data can be analysed separately and independently |
weaknesses of the convergent parallel design | >effort and expertise needed
>consequences with merging data
>conflicting findings possible |
strengths of the explanatory sequential design | >easy to implement
>lends itself to emergent approaches |
weaknesses of the explanatory sequential design | >lengthy
>need to be clear which quantitative results need addressing
>need to be clear with sampling and criteria of second phase |
strengths of the exploratory sequential design | >straightforward to describe. implement and report
>produce new instrument |
weaknesses of the exploratory sequential design | >lengthy
>sample size considerations
>ensuring scores of the instrument are valid and reliable |
strengths of the embedded design | >used if sufficient time/resources
>fits team approach
>focus on different questions=different publications |
weaknesses of the embedded design | >need expertise in design and in MMR
>must decide when is best to collect qualitative data
>difficult to integrate results when two methods and two research questions used |
strengths of MMR | >easy to describe and report
>useful for unexpected results
>helps generalising qualitative data
>useful in designing and validating instruments |
weaknesses of MMR | >time consuming
>possible discrepancies between data types
>difficulties deciding when to proceed
>little guidance with some models |
naïve realism | reality exists independently of human constructions and can be known directly |