A dissertation should be 1) scholarly 2) significant and 3) original.
Be wary of horse-race research--pitting two methods or paradigms against each other, pulling the trigger, and seeing who wins. This is great for physics, biology and other types of research but in the sociology/educational realm positivists dont' like them because it is hard to control variables, qualitative folks don't like it because the situations aren't that clear cut to the point that you could assign a couple of factors to the study and get the information you need. Plus you miss a lot of important things from their perspective.
One important factor that you should analyze in doing research is your background--values, beliefs, theoretical grounding, etc. Then look at how that impacts your research question and the methods you will use.
Don't throw methods together udner minimal constraint.