 |
The other type of statistical test has been called “interval approach”
or “scan test.” This type of test focuses on creating a series of time
intervals (again predetermined by the investigator) with the null
hypothesis of equal distribution. Again, if the “a” cell (i.e. most
recent time) is statistically the largest, then there is a disease
cluster. In other words, there is a statistical relationship in time
between the cases. There are several of these statistical tests; some
of the most famous are listed by name.
There are also references with a more detailed discussion of these
tests, as well as a statistical software package available through the
US Agency for Toxic Substances and Diseases Registry (ATSDR) (Atlanta,
Georgia).
|
