Why do we need to conduct an experiment to determine if therapy is effective?
a. different people say different things in a survey.
b. the correlation is too close to zero.
c. some people prefer some therapies and other people prefer different therapies.
d. conducting an experiment is the only way to determine causality?