We have a between subjects design, testing for the impact of aesthetics on occupant satisfaction. Intervention is carpets vs no carpets. The people participating in each leg (intervention vs control) are two separate groups.

However, due to constraints on the number of people we can accommodate on a given day, the study is spread over multiple days (for both control and intervention). While analysing the data, would it be appropriate to use ANOVA with blocking (days as the blocking factor)?

Edit (17 Feb) Sample size is 18+18 in the pilot study, and we will add another 60 to each leg during the full powered study. There is a control leg and an intervention leg. But, it is impossible to blind participants to the intervention (it kind of stares you in the face, literally) so, we need independent groups in each leg. While our design is for the intervention changing only a single parameter, there can be confounding effects, like the weather changing from day to day. How do we account for those? After some more digging, I feel ANCOVA may be the way to go for us.

  • $\begingroup$ Could you add some more information? Please do so as an edit to the post, as comments are often not seen by many and can be removed. Sample size? And, why did you use different people in each leg? That seems like a big, big problem! $\endgroup$ Feb 16 at 16:58


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.