This project uses a longitudinal national educational database to answer both substantive and methodological questions. Models of academic achievement are built for female students in math and science courses. The main interest is in identifying and clustering individual educational projectories then to examine factors that are school-related, parental, environmental, and psychological to explain differential projectories. The process of identifying and clustering educational profiles is similar to that of finding a finite mixture of distributions. The mixed effects model is used to estimate random individual acceleration functions and to relate such functions with both time-dependent and nondependent factors that can be potentially intervened. The database is increasing every two years with new and additional information so that interesting questions on students' growth profiles can be addressed.
Sandy Erickson, Graduate Student Researcher
Ali Fahmy, Graduate Student Researcher
Qi Zhang, Graduate Student Researcher
The methodological portion involves producing robust estimates for the parameters of individual projectories and school-related factors in the hierarchical linear model. As the model complexity increases, so does the size of the data matrix. In order to circumvent this problem, reliance on data-resampling and iterative methods is used, in part, to generate robust estimators. It is important to determine limiting conditions for the model because there are typically many missing values in the data matrix. Hierarchical linear model software as it now exists does not effectively handle missing data though there are recent developments in handling of missing data. Also, the time-dependency in errors and the error structures are not fully incorporated in the existing program. Thus, interest lies in exploring the effects of misspecifications in the error structure using the mixed effects model.
As opposed to traditional research questions that focused on central tendencies or group average characteristics, recent research questions are moving toward focusing on individual differences in trajectories and their concominant factors. That is the focus of this project.
|
|
URL: http://www.msi.umn.edu/about/publications/annualreport/ar2000/depts/Education/edwards.html |
|
| This page last modified on Friday, 30-May-2008 16:14:02 CDT | ||
| Please direct questions or problems to help@msi.umn.edu | ||
|
Website related questions or problems should be directed to
webmaster@msi.umn.edu
The Supercomputing Institute does not collect personal information on visitors to our website. For the University of Minnesota policy, see www.privacy.umn.edu. © 2001 by the Regents of the University of Minnesota |
||