Thinking about Design Posted on February 29, 2008 The design selection process is thorough in epidemiology, notes Bruce G. Link (bio).

The advantage of epidemiology is the depth of the thinking about design. There's some classic, standard breakdowns of what the designs are. So case control studies or cases disease versus people who don't have the disease, and cohort studies where you follow people forward to see who develops the disease so you have time order on your side, and then of course randomized studies, clinical trials types of things.

So epidemiology sets forward those types of designs. But it also gives you this thinking about what you would need to do to be able to make a causal inference about the variable you're interested in or the risk factor you're interested in or the process that you're interested in, whether or not it is causal with respect to the disease of interest.

And it has a depth of thinking about that. Social science also uses it, but epidemiology has really used it well. The idea of a counterfactual, what would have happened if you hadn't been exposed to the thing that you were exposed to, is what you want to know and what designs best mimic that because if I smoke cigarettes, we'll never know what would have happened to me if I didn't smoke cigarettes.

So we need a design across people that can reflect and try to tell us whether or not that is a causal connection. So it's good at that. It's good at giving you the thinking that you need to be able to think through: Can I make a causal inference with these data? And how can I get closest to it given the nature of my question and the data available to me or the resources available to me?

« Back to Article