I generally apply quantitative methods and would appreciate any advice for conducting a G* power analysis in SPSS 27 with single sample qualitative interview data, please?
"a G* power analysis in SPSS 27 with single sample qualitative interview data"
A few things confused me here. G*Power and SPSS are different programs. Do you mean conduct power analysis within SPSS? Second, if you have a SINGLE interview, is the power analysis to guide you on your choice of how long to have the interview? Why are you wanting to do a power analysis? Normally it is just a method to guide the researcher on how large a sample to have.
Hi Daniel,
Thank you for your thoughtful response. I referred back to my notes, and apparently, power is somewhat less risky here than in OLS regression or SEM. I am concocting approximately twenty structured interviews forming a collective sample, which I believe should approach the saturation level. Fortunately, my instructor successfully instilled the inclination to consider power at the start of my research rather than an afterthought when enduring insignificant findings. Yesterday was my first time posting here, this is an excellent asset, and I am sure I will have more questions as my project takes shape, exciting! All the best to you! My instructor's handouts advised the following programs if they may be helpful to you or anyone else.
1.ThePinT
https://www.stats.ox.ac.uk/~snijders/multilevel.htm#progPINT
2.Optimal Design
(Spybrook, Raudenbush, Congdon, & Martinez, 2011) https://sites.google.com/site/optimaldesignsoftware/home
Hi Chela, those (along with mlpowsim (http://www.bristol.ac.uk/cmm/software/mlpowsim/) are for multilevel power situations. Some more details of your design might be useful. Dan
Hi Daniel,
I appreciate your interest in the project. However, I am working on a minor component of a larger project with community partners. Accordingly, I am unsure about ethical obligations towards disclosing further details while the design is still in its nascent stages.
Others may not appreciate a student taking liberties with shared intellectual property, but this is an excellent question to ask. I noticed your broad interests include psychometrics; I am pleased to engage in a new dynamic forum. All the best! All the best to you, and thank you for your advice.
Be Well, Chela L
I am not sure what you mean by "structured interviews." If your goal is measure variables and the relationships between them, then that this unlikely to be accepted as qualitative research.
My understanding is that G*Power is used to calculate sample size. Chela, it appears you are using it to calculate minimum number of your audience (sample size) to interview.
- Badges
- Science topic
- Similar topics
- Psychology