Why are we taught that we can gain insight and the experience of beauty only through art, when this is but a limited and second-hand
representation of the infinitely deeper experience to be gained by direct observation of the world around us? For such observation
to become significant it must be made in the light of knowledge. The sense of wonder and excitement to be derived from watching
the way an insect's wing functions, or an amoeba divides, or a foetus is formed comes in its greatest intensity only to those
who have been given the opportunity to find out how these things happen.
— James Burke, Connections (Boston: Little, Brown & Co., 1978, p. 295)
The next four chapters address the methods used in epidemiology to discover how and why things occur. This is often referred
to as analytic epidemiology as distinguished from the preceding chapters that were concerned with descriptive epidemiology.
The most powerful research strategy in epidemiology or any science is the experimental method. An experiment provides
firm evidence for or against any causal hypothesis so tested. Experimental methods, however, cannot always be applied to every
study; therefore, quasiexperimental methods must often be substituted for the experimental method.
In epidemiology the two most commonly used quasiexperimental methods are the prospective cohort study and the case-control
study. The prospective cohort study can provide moderately firm evidence of causation. A case-control study can only provide
evidence which is strongly suggestive but not firm.
The fourth chapter discusses the application of analytic methods to the practical task of investigating an epidemic or
outbreak. This is often referred to as field epidemiology.