Friday, May 19, 2006

What I Study: Design & Analysis of Experiments

So I started to write about this class and it got all long and involved and more than most of you want to know. So let's simplify.

In a nutshell, Design of Experiments is similar to regression except instead of just observing data and trying to relate variables with one another, a researcher does a controlled experiment in which they assign people to certain "treatments" and then see what happens. For instance, in a drug trial people with a certain medical condition may be randomly assigned either a drug or a placebo. The results then are analyzed to determine if the drug has any effect. This is different than just observing a characteristic of a person (i.e. how much chocolate one eats daily).

Concepts, theory, variations, extensions, and complications about/regarding/on/of this basic idea pretty much kept us busy for the last semester.

Well, that and Stat Theory, which I'll brief (emphasis here) you on soon.

0 Comments:

Post a Comment

<< Home