Hi everyone! I conducted a-priori analysis in G power before collecting data, and it said I need to have 269 responses. Now I finished data collection with 289 responses in total, but the complete, valid responses is only 172. I want to do MANOVA, how can I know if the data I have now is enough? in other words, should I continue to collect more data? if so, how can I know how many more to collect? Thank you!
@Yuanyuan Zhu Refer to the YouTube below. It talks about power analysis for MANOVA using Gpower.
https://youtu.be/YFWINBJWSOw
Yuanyuan Zhu try to go through the following paper to understand a step by step process to use Gpower for minimum sample size calculation, as I feel 269 is a very high minimum sample size requirement canculated through Gpower.
Ringle, C. M., Da Silva, D., & Bido, D. d. S. (2014). Structural Equation Modeling with the SmartPLS. Brazilian Journal of Marketing, 13(2). doi:10.5585/remark.v13i2.2717.
Hello Yuanyuan,
If your a prior power analysis was suitable, then it looks as if you need 289 - 172 = 117 additional cases for your analysis to have at least the target power to detect the target ES (should it exist in the population).
If you are unable to collect additional data, then obviously you'll have to deal with the data you have in hand. Be sure to note that as a limitation when you summarize your findings.
You can possibly cut down a bit on the impact of incomplete data cases by applying a suitable imputation method. (Either way, it's probably worth your while to compare complete to incomplete cases on any relevant variables to be sure that self-selection of your nominal respondents would not bias your results.)
Good luck with your work.
- Badges
- Science topic
- Similar topics
- Psychology
- Mental Processes
- Learning