Picking a good research question is perhaps the most important step in the research process.  When you think you may have a good idea, apply the "FINER" criteria to it.

F- Feasible.  Is the question answerable?  Do you have access to all the materials you will need to do the study?  Do you have access to enough subjects?  Will you have enough time and money?  Do you have the expertise to do this study or can you collaborate with someone who does?

I- Interesting.  The question has to be interesting to the investigator, but should also be interesting to others.

N- Novel.  Has this study been done before?  Does it add to the current body of medical knowledge?

E- Ethical.  Can the study be done in a way that does not subject subjects to excess risks?  Will an IRB approve the study?

R- Relevant.  Will it further medical science?  Will the results change clinical practice, health policy or point towards further avenues of research?

If your question fails on any one of these 5 criteria, it is probably not worth putting much effort into.