Ecologic Study, Ecologic Fallacy

An ecologic study uses summary or aggregate information as the data points for analysis (i.e., units of observation).  For example, an investigator may use average alcohol intake and colorectal cancer rates of various geographic regions to examine whether there is an association between alcohol intake and risk of colorectal cancer.  One advantage of these studies is that they typically use data that is already available.  They can lead to interesting questions about relationships between an exposure and risk of disease.

Ecologic studies serve as a starting point for assessing exposure/disease relationships, since they are subject to the ecologic fallacy.  An issue with using summary statistics like averages and rates for analysis is that, essentially, there is no link in the data between which individuals are exposed and which have the disease.  Taking the example above, there may be a lot of drinking and a lot of colorectal cancer cases in a certain geographic area, but the people who drink a lot may not be the ones diagnosed with colorectal cancer.  This is called the “ecologic fallacy” because it identifies a relationship that isn’t there.