Introduction to Epidemiology 6th Edition by Ray M Merrill

Exercise 13

What are two statistical challenges in cluster investigation?

Step 1 of 3

The primary challenge in cluster investigations occurs because most of the cluster analysis tends to involve post hoc (posteriori) over a priori hypotheses.

Posteriori is the hypotheses formulation upon the observation of certain events that have already occurred for example excess of cancer cases. This type of hypotheses is problematic as these are established with a lack of prior knowledge about the levels of health issues in a specific population and the P-value is interpretable only with a priori hypotheses.

There is an increase in the probability of finding significant results because a suspect cluster is being selectively chosen for statistical testing and thus this testing become equivalent to multiple testing.

Another challenge occurs in the cluster investigations such that the rates can be overestimated because of boundary-shrinkage of the population where we presume the cluster to exist.