Schedule and Presentations 2008
This year’s theme will be “Using neuropsychological tests to detect effects of Alzheimer and cerebrovascular neuropathology. ” The data will come from the Religious Order Study from the Rush University Alzheimer ?s Disease Center. This is a longitudinal study that conducts annual clinical evaluations on more than 1,100 older Catholic clergy who were non-demented at the time of enrollment and agreed to annual evaluation and brain autopsy. The data are from fifteen waves of clinical evaluation, including a detailed neuropsychological test battery, with neuropathology results for 400-500 cases. The principal investigator of the study, David Bennett, M.D., and other investigators in the study will attend the meeting.
The specific training areas and learning goals of the workshop this year will be to 1) use Item Response Theory and Structural Equation Modeling based methods to evaluate dimensional structure of neuropsychological test results and to create composite scores with known psychometric properties, and to compare these composite scores with more traditional z-score composites; 2) to use mixed model regression analyses with z-score composites and IRT based composites to examine effects of Alzheimer and cerebrovascular neuropathology on cognitive change across multiple domains, and in addition, develop SEM based latent growth models that will simultaneously account for dimensional structure among tests and will examine effects of neuropathology independent variables on change in the latent dimensions; and 3) use explanatory IRT models that incorporate the neuropathology outcomes and compare ability estimates and item parameters from the explanatory IRT models with those from the more traditional IRT models.
Like in the previous conferences, part of the workshop will be organized around workgroups that will perform analyses related to scientific manuscripts. Our goal will be to substantially complete analyses during the week and then organize the workgroups for further interactions to complete the manuscripts. Conference participants will have a common interest in practical application of psychometric and statistical data analytic methods, but will have varied backgrounds in terms of specific areas of expertise and level of training and experience. Trainees including undergraduates, graduate students, and postdoctoral trainees are encouraged to attend.