0
$\begingroup$

I have two samples of the same size, one being the 'treated' group, and the other used as a 'control' group.

Each group have a different univariate variable (stress level), repeatedly observed over 100 periods.

Suppose that after period 75 the traetment begins.

I want to test the significance of the hipothesis that the treatment has an impact in the stress level in the individuals within the 'treated' group.

Is there a statistical test that checks signifficant differences accross groups and accross time?

$\endgroup$
1
$\begingroup$

Since treatment begins at period 75, it seems you only have 25 periods. In any case, you could use a paired test that uses the change in stress level from the time when the treatment was administered to the end of the study for each individual, and check whether this quantity varies between the treated and untreated groups. To do this you could use either a t test or (my preference) a non-parametric alternative such as the Wilcoxon rank sum test to determine if stress levels were stochastically larger in the untreated group.

$\endgroup$
0
$\begingroup$

Julian -- I would take advantage of the 75 periods of 'baseline' and go the ancova route.

An alternative (if your groups are randomized and indeed similar during that baseline period) would be to calculate the difference in their pre-post change.

$\endgroup$
  • $\begingroup$ I've assigned the individuals in order to assure that both groups are similar prior to the treatment. $\endgroup$ – Julian Lopez Baasch Jul 3 '15 at 15:45

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.