Dear colleagues,
I am tryng to understand what the option "planned comparisons of LS means" in Statistica is doing?
Let's say I have a factorial ANOVA with two categorical explanatory variables: sex (2 levels) and treatment (4 levels). So sex*tretment includes 8 subgroups.
"Traditional PostHoc" do all possible paired comparisons between the 8 groups, and includes a correction taking into account that there are 8 groups.
BUT with "planned comparisons of LS means" I can ask to do a specific comparison between, for examples, males and females with treatment 1. How is doing it? I suspect that is not comparing among all categories and show only the output for males and females with treatment 1, right?
If I extract the data for males and females with treatment 1 and do a simple ANOVA between this two group, result is the same?
And, overall, what is the correspondence in R?
Thanks, I am an R user and I cannot find how statistica work regarding this issue
Planned comparisons means you chose those comparisons prior to collecting data. They are from your pre- data collection hypotheses not based on information from your data. Best,, David Booth PS social science directed packages make a bigger deal out of it than say R because most R users are familiar with conditional probability
I do know nothing about Statistica but since you are familiar with R and asked for the correspondence in R you should definitely look into the R-package emmeans. The documentation covers the concept of LS means and how to apply it.
I think results from your hypothetical ANOVA will be similar to LS means but keep in mind that with more comparisons there is always more alpha error.
Best,
Sven
Planned comparisons means you chose those comparisons prior to collecting data. They are from your pre- data collection hypotheses not based on information from your data. Best,, David Booth PS social science directed packages make a bigger deal out of it than say R because most R users are familiar with conditional probability
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA