Hello Dreamscat-users.

I study psychology and phisiology, and my first idea is conducting a research study measuring the relationship between food preferences and some basic psychological characteristics like prevailing psychological defense mechanisms (PDMs).
This survey could be distributed online in a random sample of users from all over the world. It would be a quantitative research, with as many extraneous variables taken into consideration as possible. The first limitation I see is that perhaps the online wouldn't be that representative, because it's likely that users filling my questionnaire online have certain kind of PDMs peculiar..

do you see more limitations of such a study?