So I am having a rather simple problem of how to create a table of means in R and post-hoc t-tests.


DV: Score (Continuous variable)

IV1: Country (Nominal, 3 levels)

IV2: Group (Nominal, 4 levels)

IV3: Condition (Nominal, 2 levels)

I have run an ANOVA with interactions and after removing non-significant interactions, I am left with a model that looks like this.

DV ~ IV1+ IV2+ IV3 + IV1*IV2

IV3 is non significant, but the rest are.

The question I have now is how do I create a table of means split by country and group?


            America         UK            France

Group 1       10.3         9.2              11.4

Group 2       7.8          8.6               2.4 

Group 3       4.5          10.7              8.6

Group 4       10.5        10.5               10.6

Secondly how do I then run comparisons to see where the significant interaction is?

I know the pairwise function but I seem to only get that to work for saying comparing the 4 groups, or the 3 countries

pairwise.t.test(Data$Score, Data$Country)


pairwise.t.test(Data$Score, Data$ Country, by = Data$Group)

Seems to do the exact same


1 Answer 1


First, you should not remove non-significant terms from a model simply based on p-values, that is not their purpose.

Second, to get means for each combination of groups (and comparisons) use a linear model.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.