finding p value for pre/post and treatment/control groups (using r)

Question

I need to find p value for a set of variables in a 2 groups (treatment and control) 2-times scenario (same subjects, pre and post). groups have the same n, with different subjects that have been randomly assigned.

My data looks like this:

> str(t_wide)
Classes ‘tbl_df’, ‘tbl’ and 'data.frame':   32 obs. of  31 variables:
 $ Subject    : num  1 1 2 2 3 3 4 4 5 5 ...
 $ Time       : num  0 1 0 1 0 1 0 1 0 1 ...
 $ Group      : chr  "Treat" "Treat" "Treat" "Treat" ...
 $ VAR1         : num  7.6 7.7 9.4 9.5 8.3 8.2 7.5 7.6 7.4 7.4 ...
 $ VAR2       : num  8400 8350 8200 8220 8300 8400 7200 7380 7200 7250 ...
 ...

my first try was to get p value like this:

t.test( VAR1 ~ Group, data=t_wide )

but i guess this leave Time variable out of the model (and probably just duplicates the cases in each group). how should i proceed? should i calculate some sort of variance of pre/post for each subject?

many thanks!

Answer 1

I think we first need to clarify whether the same subjects exist in pre and post because your original post is not quite clear on this. Assuming the pre and post contain the same subjects:

You will probably need to use ANCOVA or a linear mixed model. You may need to control for the individual differences if you have more than 1 repeated measure for each subject that is pre/post. You will definitely need to examine the pre/post differences relative to condition.

Refer to here for a full discussion of tradeoffs in this situation: Best practice when analysing pre-post treatment-control designs