R - Mixed Design ANOVA post hoc test

Question

I have the following data structure (with example values):

id     var1     var2   value
1      true     tr     1.34
2      true     ct     4.89
3      false    mm     2.38
4      true     tr     1.28

The data is saved in 'longData'. So 'var1' is between subject variable that can be true or false, 'var2' is a within subject factor with 3 levels (tr, ct, mm) and 'value' is a numeric value.

I've made a mixed design ANOVA like this:

anovaResult = ezANOVA(data=longData, 
                                dv=.("value"), 
                                wid=.("id"),
                                within=.("var2"),
                                between=.("var1"),
                                type=3)

The result showed signifikant interaction between var1 and var2. Now I would like to examine this interaction further, but I don't know how. I've heard about the emmeans package (estimated marginal means seems to be the statistic of choice here, since I am new to statistics, feel free to advise me otherwise) but could not get the command to work. This is probably because I am new to R and do not understand the syntax fully.

Can anyone provide me with a working example of how to test the interaction between the two factors? I would not say no to an explanation of how to interpret the results as well.

I know this is much I am asking for, but I cannot figure it out for myself and have to present results soon without much time to learn statistics and R. Thank you.

Answer 1

It would help to provide an example dataset.

However, you can run a Tukey test:

mod1<-aov(value~Factor1*Factor2, df)
TukeyHSD(mod1)

Or to run emmeans on an Anova with an interaction:

mod1<-aov(value~Factor1*Factor2, df)
library(emmeans)
emmeans(mod1, pairwise~Treatment*Time)

Or to do a mixed model, which is what you seem to be doing:

mod1<-lmer(value~Factor1*Factor2+(1|subject), df)
Anova(mod1)
summary(mod1)
emmeans(mod1, pairwise~Factor1*Factor2)