Methodology

• Almost always retrospective, case-control studies look back on a sample containing diseased individuals, or "cases," and healthy individuals, known as "controls," to determine the presence or absence of a risk factor, also referred to as "exposure." Cohort studies, in contrast, look forward in time by choosing a "cohort," or group of individuals, and following it throughout their lives. Usually, the group is chosen so that some members have a certain risk factor and others don't. Later, researchers record whether individuals in the cohort developed the disease.
  
  

Causal Links

• Case-control studies can show an association between a risk factor and a disease, also known as a "correlation." However, as statisticians are fond of saying, "Correlation does not imply causation." It's possible, for example, that the disease and the risk factor arose from some other, unstudied variable, and no direct link exists between the two. Cohort studies -- because they select exposure first and later look for disease -- are positioned in time so that they sit directly on the causal link. This advantageous viewpoint allows researchers to draw stronger and more applicable conclusions from the results.
  
  

Cost Factors

• Choosing a study methodology often comes down to cost. Case-control studies -- often conducted by simply looking over compiled public health records -- prove cost- and time-effective. Cohort studies, however, take longer and cost more, and often require direct contact between the study's subjects and researchers. The ease of conducting a case-control study makes it a good choice for preliminary work that explores a new hypothetical risk factor-disease connection, and to possibly secure funding for a more extensive cohort study later.
  
  

Examples

• As an example, suppose Dr. Joe wants to know if growing up within five miles of a chemical factory causes cancer. For a case-control study, he finds a sample containing cancer patients and cancer-free people in the same age range and with a roughly equal mix of genders, races and other characteristics. After finding out whether these people grew up near the chemical factory, he performs statistical computations to see if the two variables are related.
  
  To do a cohort study, Dr. Joe chooses a sample of young people -- some living near the chemical factory and some not. As the years elapse, he keeps track of them, and records who develops cancer and who doesn't. At some future point, the data is collated and analyzed.