Pre-test/post-test quantitative designs have been the subject of criticism among methodologists. What practical steps do you think a researcher can take to address the limitations of such a design if it is the only one available?