I am running ANOVA analysis. Two IVs and several DVs. The number of participants in the four groups are different. Thus, the estimated marginal means are different from descriptive means. My questions are (1) do I need to report the descriptive Mean and SD? (2) should I report estimated marginal means or descriptive means?
Report estimated marginal mean and standard error in writing results and making graph.
I may be off the mark a little but I don't see a mean value as something "descriptive". The (arithmetric) mean is the maximum likelihood value = least squares value of a model parameter (in the simplest case of an "intercept-only model"). Its interpretation includes or presupposes the idea of symmetric and independent expectations about deviations of observations from this value. Thus, reporting a mean value is reporting the result of a probabalistic model. If I had to describe/summarize the observations (without implicitly using some underlying model) I would report the modes or medians for the location and some percentile-ranges (e.g. the interquartile range) for the dispersion. Just my two pence, feel free to ignore this comment.
I know, exactly this. IMHO it is already kind of careless from the software designers to call this "descriptive statistics". But I am aware that this is a provocating view...
Report estimated marginal mean and standard error in writing results and making graph.
The Estimated Marginal Means in SPSS GLM tell you the mean response for each factor, adjusted for any other variables in the model. They are found in the Options button. (These are the same as the LSMeans in SAS GLM).
https://www.theanalysisfactor.com/why-report-estimated-marginal-means-in-spss-glm/
Basics of estimated marginal means
https://cran.r-project.org/web/packages/emmeans/vignettes/basics.html
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA