Grantee Research Project Results
2000 Progress Report: Using Multilevel Statistical Models to Address Representativeness and Data at Different Spatial and Temporal Scales
EPA Grant Number: R826763Title: Using Multilevel Statistical Models to Address Representativeness and Data at Different Spatial and Temporal Scales
Investigators: Berk, Richard , Ambrose, Richard
Institution: University of California - Los Angeles
EPA Project Officer: Hahn, Intaek
Project Period: October 1, 1998 through September 30, 2000
Project Period Covered by this Report: October 1, 1999 through September 30, 2000
Project Amount: $414,149
RFA: Regional Scale Analysis and Assessment (1998) RFA Text | Recipients Lists
Research Category: Aquatic Ecosystems , Ecological Indicators/Assessment/Restoration
Objective:
The main objective of this project is to extend existing multilevel statistical models beyond to the generalized linear model, thus allowing for various kinds of dependence in the error structures. With the theory completed, portable software would be written implementing the procedures. Finally, one of the key applications would be to consider how best to generalize from a set of case studies.Progress Summary:
There already exist packages that can do some of what we are trying to do. The MLWin package, by Goldstein and others, for instance, can do generalized multilevel models with independence, and by using its macro facilities it can be extended to handle forms of dependence.
We have been working on a complete and user friendly set of procedures under XLISP-STAT. We began by extending existing modules to allow for the case of the generalized linear model within a multilevel framework and to handle time series-like dependence. Both extensions run and are documented, and we have used both in real applications. However, it has become apparent that it would be difficult to merge the two extensions into a self-contained and more general module, let alone add more features.
Therefore, we revised our strategy to a bottoms-up approach. We start with the XLISP-STAT weighted least squares module, and extend it in such a way that it can handle non-diagonal but fixed weight matrices. We then take this module, and extend it further with block relaxation techniques so it can handle parametric covariance structures. This allows us to fit additive structures, random coefficient models, hierarchical linear models, and mixed linear models. The theoretical work on this has been completed and some initial new code has been written.
Future Activities:
The next step in the project will be to construct a module that inherits from the XLISP-STAT generalized linear model module, but also from the non-diagonal weighted least squares module constructed in the previous step. The new code will be incorporated into Arc, an elegant package on top of XLISP-STAT. A key feature is a wide variety of modern regression diagnostics that can be applied to our enhancements. Finally, we will continue with our work on applications to demonstrate the usefulness of our software.Journal Articles:
No journal articles submitted with this report: View all 1 publications for this projectSupplemental Keywords:
modeling, analytical, statistical inference, external validity, hierarchical models., RFA, Economic, Social, & Behavioral Science Research Program, Ecosystem Protection/Environmental Exposure & Risk, Regional/Scaling, Environmental Statistics, data synthesis, regional environmental data, risk assessment, non-linear functional forms, ecosystem assessment, representativeness studies, multiple response variables, survey data, environmental risks, multilevel statistical model, hierarchical statistical inference, satellite data, modeling, external validity, statistical models, regional scale impacts, data analysis, spatial-temporal methods, spatial and temporal scales, representativeness, multiple response variable, data models, hierarchical statistical analysis, innovative statistical models, regional survey data, remotely sensed data, statistical methodsProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.