|
Abstract:
|
In the 21st century two key developments in the mental health field have taken place, that constitute a fundamentally new research paradigm and require new perspective about uncertainty and the role of hypotheses testing, sample size and power. One is the major shift of focus from comparing treatments on their average performance in patient populations, to identifying which treatment is best for an individual patient. The second is the unprecedented diversity and amount of information that can be collected -- including the brain's structural and functional integrity; genetic representation; and deep behavioral and environmental characterization, in addition to traditional clinical features and family history. Despite those major changes, investigators continue to rely on the classic clinical trials design and analytic methodology and even though the amount of patient information has risen astronomically, the methodologies for analyzing and combining such information have not kept pace. I will discuss methodological challenges related to those changes and will consider directions of statistical inquiry necessary to address the new reality of research in mental health.
|
