Hi everyone,
My IV are four groups with uneven sample size: 5, 42, 55 and 87. I have 4 DVs and all of them violated both Box's M and Levene's Test. Can I still interpret the results of the MANOVA or would I have to use a nonparametric test like Kruskal-Wallis? Thank you so much for your time and guidance!
Why do you want to use MANOVA? Do you have one multivariate research question, or 4 univariate research questions? For more info about why I am asking those questions, please see the following articles:
Article Multivariate Analysis Versus Multiple Univariate Analyses
Article MANOVA: A Procedure Whose Time Has Passed?
See also Thom Baguley's blog post:
- http://psychologicalstatistics.blogspot.com/2021/08/i-will-not-ever-never-run-manova.html
Bruce Weaver Thank you for the articles! My DV are all related measures: four different types of parent-child conflict. As such, a MANOVA would help reduce the likelihood of Type One error.
El Bakkali Mohamed thank you for your thorough response! It's similar to what I found through ChatGBT as well, but the transparency reporting is very informative! I've read somewhere that Box's M is oversensitive to large sample size so that's been commonly criticized. When Levene is violated, some recommended using Welch or Brown-Forsythe ANOVA to see if it is significant. However, since there's no MANOVA version of Welch or Brown-Forsythe ANOVA, could I still use run the individual ANOVAs and report it transparently the measures I took to try to cover my basis? There's also the concern that my sample size across the IV Groups are uneven..
Thank you both so much!
El Bakkali Mohamed thank you so much for the step-by-step advice, it really helps make complicated issues so much simpler! I will also note these assumption violations as part of the limitation to my study. Thank you so much for being so kind and supportive!
Johnny Wong wrote (emphasis added):
El Bakkali Mohamed thank you for your thorough response! It's similar to what I found through ChatGBT as well, but the transparency reporting is very informative!
I think that was a very polite way of saying that El Bakkali Mohamed's first post in this thread appears to have been generated using AI. The QuillBot and CopyLeaks AI detectors both flag it as 100% AI-generated. I've attached an image showing the QuillBot report. As Johnny hinted, if people are going to post AI-generated content, they ought to say clearly that it is AI-generated content.
First, I write in French, and when I translate to English using AI tools, Quillbot detects plagiarism. But for me, the main thing is that I found the solution and translated it to try to answer the question, and I don't see what's wrong with that.
- Badges
- Science topic
- Similar topics
- Social Science
- Media
- Social Media