0
$\begingroup$

I have a data set where I try to analyse a continuous variable according to 10 categorical variables and I would like to perform a ANOVA analysis. How should I proceed ? I'm able to interpret the results but I have no clue how to operate :

  • An anova with all the variables at once : summary(aov(Cont ~ . , df))
  • An anova for each categorical variable : summary(aov(Cont ~ Team , df)) and summary(aov(Cont ~ Materiel, df)), ect

Because some variables are significant alone, but not with all of the at once (because orders matters as I understood)

Also, what if I want to test interactions ? Should I test them all at once ?

$\endgroup$
0
$\begingroup$

first of all, good on you for getting into R. Its an amazing research tool.

As for your question, what you need to do is specify your model a bit. I do not know if you have looked at this page: http://www.statmethods.net/stats/anova.html

Their quick-r tutorials are very helpful.

And to point something out, the reason that some variables are significant in the model by themselves but not with other variables has to do with them sharing explanatory power. The order of your model matters because of how the explanatory power is distributed (think in a first come first serve manner rather than evenly distributing it among predictors)

$\endgroup$

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.