I am trying to determine if the treatment effect on different groups are statistically different using a statistical test. More specifically, my data is divided into 6 strata. and in each stratum there is a treatment as well as a control group. For each group in each stratum, I then measure three outcome variables post-treatment. However, I do not know which statistical test is the most appropriate to use when I try to determine if the different in each outcomes between the two groups in each stratum differs significantly with respect to the chosen stratum. Also, if there is a suitable test, please also point out the corresponding function in R (if there is one).

I will give a more specific example. Say in stratum one group C1 has 1000 data points with an average outcome value of 0.06, while in group T1, the treated group in the same stratum, the outcome value has a mean of 0.11. We then have a difference of 0.05. Say in the 2nd stratum with 800 data points, C2's mean outcome value is measured to be 0.10, and T2 0.24. The difference is 0.14. We are then interested in finding out if the two differences 0.14 and 0.05 are statistically different.

Greatly appreciated!

  • 1
    $\begingroup$ Questions about which statistical test to use with your data are not appropriate for Stack Overflow. You should ask such questions on Cross Validated where questions about statistics are on-topic. Once you know what test you want to perform, then you can later ask how to do that in R here if you can't find the answer with google first. $\endgroup$ – MrFlick Mar 13 '19 at 18:12

Assuming your data meets the criteria for parametric tests:

  • if you have both, several treatment groups and several measurements: 2-way ANOVA for repeated measurements.
  • if several treatment groups and one outcome: 1-way ANOVA
  • you can also perform pairwise comparisons with Students t-test
| cite | improve this answer | |
  • $\begingroup$ Thank you for the suggestions. I think the two-way anova test is applicable. I will just test if the two factors are independent. $\endgroup$ – Peiyao Li Mar 14 '19 at 20:58
  • $\begingroup$ Great! Which test you are planning to use for that? $\endgroup$ – Oka Mar 19 '19 at 9:31

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.