Hi! I would like to evaluate the causal effect of a treatment, that i will indirectly measure through psychological questionnaires. Also, i have no control group (this is a natural experiment). I wonder if i can simply (randomly) ask some of the participants to answer before treatment, and others after, and compare the 2 pools of answers (knowing that my treatment already selects for very similar profile, and is conducted on a short period of time, so that i thought i could assume that my 2 pools of participants are comparable in any other way and would follow similar trend in time, so that if a difference appears in answers, it should be due to treatment only). does this seem correct?

Thanks in advance!

More Camille Perault's questions See All
Similar questions and discussions