Survey Methodology

[pantouka]

---

Members do not see this ad.

deleted

 

[Stories]

Most epidemiology departments will have some sort of survey methodology course(s). Because of the recent divergence into molecular methods, some degrees will now focus on molecular methods as opposed to traditional questionnaire methods, but you can build your curriculum how you please at most places!

As for MD/PhD, it's a long and grueling process, but remember to be both a physician and a scientist 🙂 Some folks forget one or the other in the process of choosing their careers.

 

[MolBio]

Recently I've found myself interested in the field of survey methodology and the potential applications it has in public health and translational medical research. Does anyone here have a good background in this discipline? I am not sure I completely understand how it differs from a typical epidemiology degree, because I assume you could obtain a PhD in epidemiology with an emphasis on survey design, no?

The term "Survey design methodology" could potentially refer to at least a couple of different things:

1. The wording, ordering and design of the questions - i.e., "questionnaire design". For example, how to word specific questions to patients in order to reduce bias. I'm not sure there is a huge space for new cutting edge research here? Most of the preferred methods are known, and I'm not aware of many researchers who focus purely on this topic. There is a potential career for validating questionnaires or psychosocial scales, for example, with respect to specific diseases (e.g., do these questions really measure the true construct "depression"?), but there are many many such scales already validated to the point that they are conventionally used.

2. The study of survey sampling or "sampling design". This is more the statistical implications of different sampling designs (cluster randomization, stratification, weighting, etc.) either for planning or analysis. This is an extremely important field in public health and it would require a very strong understanding of statistics (ideally a PhD biostats) to be a cutting edge methods researcher.

I'm thinking of applying for an MD/PhD, and so the idea of being able to bring a technical, broad medical background to elements of survey design, which would allow me to be fully engaged in the process, sounds really appealing, even if I lose out on some of the data analysis training you might find in a biostatistics or epidemiology program.

Personally, I'm struggling to see the synergy between the clinical medical career path and either of the above. There could be some opportunity to using your clinical expertise to create, improve, or validate scales for your disease of specialty (i.e. the first option), perhaps for rare conditions or diseases that have not already been well established. You could do that with just an MD/MPH.