summarize: I would encourage you to look at the research already done on this as there may be some ex post facto data that already exists that you may use to answer your research question. However, if we are using a group comparison model there are some things we need to be aware of that might impact or skew our results. The first thing we need to try and manage is the sample, who are we using here, how are we getting our sample, is the sample a true random sample? Things to consider might be if we want to generalize then we need a true random sample. Random sampling means that anyone can be in either group and it is not controlled. Normally we would also want a double blind experiment to control for extraneous variables. In addition, might there be some ethical issues if you are withholding treatment from some of your participants? Is there any risk and what might that be/if any? How are you going to get your sample? In addition, how can you make it random if really it might be a snowball or convinence sample