|
Abstract:
|
It is common in Phase I clinical trials to administer a large battery of cognitive and psychomotor tests to assess safety or proof of concept, depending on the indication sought by the sponsor. As a statistician, one is faced with several interesting problems. First, these tests often vary from site to site due to the way in which subjects' responses are recorded and differing equipment. Second, there are often versions of the same test. Third, the battery is given often in the same order (i.e., no randomization to time). Finally, there are statistical issues such as correlation between various endpoints of the test battery, multiplicity, and the assumption of normality. We will discuss approaches to the above issues.
|
