An Overview of Statistical Opportunities in Psychiatric Research

• Activity Number: 6
• Type: Invited
• Date/Time: Sunday, July 29, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Social Statistics Section
• Author(s): Douglas R Langbehn*+
• Companies: University of Iowa
• Keywords: Psychiatry ; Psychometrics ; Epidemiology ; Brain Imaging ; Microarray ; Clinical Measurement

Abstract

Among the medical specialties, Psychiatry enjoys an especially close relationship with Statistics. This is due to at least two historical influences. The first is dependence on sound epidemiologic and other observational analysis, given the paucity of directly measurable biological processes. The second is the close relationship with Psychology and psychometric methods such as scaling, reliability, latent variables, and causal path analysis. Currently, Psychiatry is heavily influenced by genetic microarray technology and its concomitant statistical analysis challenges. To this, we must add a dramatic influx of brain-imaging and image-processing modalities, which create additional specific statistical challenges. Finally, the recent emphasis on "functional" outcome measures in clinical trials has led to renewed emphasis upon the fundamental problem of assessing outcomes that have no natural unit of measurement-a decades-old problem in mental health research. Thus, Psychiatric research depends on an unusually wide range of applied statistical expertise. Similarly, it is a rich source of challenges that provide impetus to new developments in statistical methodology.