Week 7 Notes
Vogt, W. Paul et al. When to Use What Research Design. New York: The Guilford Press, 2012.
Personal Connections and reflections:
The explicit patterns outlined in each chapter of research design add a level of consistency that I appreciate. For example, understanding that each design method comes with a breakdown of ethical issues allows for direct comparison between chapters and design methods. Several studies I am examining for my area of research utilize mixed methods (like observations and interviews, or surveys and archival research), so this module has allowed me to think about the diverse and adaptable ways research can be conducted and analyzed.
Chapter 6 – When to Use Combined Research Designs
Key Points, Quotations, and Terms:
The approach that involves conducting utilitarian research with practical significance has been termed pragmatism. Pragmatic can be taken to mean practical and useful, but not necessarily neat and straightforward.
When to use mixed methods:
· When you want to corroborate results obtained from other methods
· When you want to use one method to inform another method
· When you want to continuously look at a question from new angles and look for unexpected findings and/or potential contradictions
· When you want to elaborate, clarify, or build on findings from other methods
· When you want to tell the "full story" in an area of inquiry
· When you want to develop a theory about a phenomenon of interest.
Both quantitative and qualitative methods can be combined, and more than one researcher can be involved.
Triangulation—the use of several means to examine the same phenomenon—can occur within or between methods, and for several purposes.
· Triangulation can be built by using data sources, multiple methods, or multiple investigators.
Chapter 12 - Sampling and Recruiting for Combined Research Designs
Key Points and Quotes:
One distinction is very important: samples fall into two categories— either probability or purposive samples.
· In a combined research design, you will need to not only select sampling methods appropriate for each component of your design, but also to select them based on their compatibility/ appropriateness.
· Probability samples are needed when you want to generalize your findings to the larger population; purposive samples are selected when specific characteristics of your cases are important to your study.
· Use both probability and purposive samples when:
o One or more components of your design require probability sampling in order to make generalizations, and one or more components require in-depth study of specific cases of interest. If you do sequential sampling, when you can assure that the sampling process will not create or compound sampling error.
· Purposive samples should be used in your combined design study when you want to garner a variety of perspectives on an issue, or when your cases fit certain characteristics that are important to your study, such as being “typical,” or being “outliers,” or being part of a group receiving specific “treatment,” or representing a range of variation along a continuum or representing a particular combination of interesting characteristics.
Types of sampling include: identical samples (when you need to hold the participants constant while testing the effects of a variable (such as an intervention or treatment), parallel samples (when you can reasonably select two random samples from a larger population in order to administer different treatments to them and compare effects.) nested samples (when you seek more in-depth information from a subgroup of a larger sample), and multimodule population samples (when you need to hold the intervention or treatment constant in order to compare results across different populations.)
Chapter 18 - Ethical Considerations in Combined Research Designs
Key Points and Quotes:
For each method you employ, ethical considerations must be analyzed for each stage: design, sampling, data collection, and analysis and reporting. The more complex your design, the more complex your ethical issues may be.
Because combined designs utilize multiple research methods, they have the potential to compound consent considerations. These include adequately describing your research plan; fully assessing potential conflicts of interest in all components of your design; planning to maintain confidentiality or anonymity across all aspects of your data collection, analysis, and reporting. All of these considerations have the potential to add up or to multiply as a result of interactions among the various components of your combined research design. All must be included in the information provided to potential participants to obtain consent.
You must review all elements that could cause stress for participants and safeguard plans for conducting research on vulnerable populations.
Violation of privacy can also be considered harmful.
To preserve the privacy of research subjects, the researcher will usually use methods of concealment such as pseudonyms.