Using Multilevel Statistical Models to Address Representativeness and Data at Different Spatial and Temporal Scales

EPA Grant Number: R826763
Title: Using Multilevel Statistical Models to Address Representativeness and Data at Different Spatial and Temporal Scales
Investigators: Berk, Richard , Ambrose, Richard , DeLeeuw, Jan , Gould, Robert , Turco, Richard
Current 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 Amount: $414,149
RFA: Regional Scale Analysis and Assessment (1998) RFA Text |  Recipients Lists
Research Category: Ecosystems , Ecological Indicators/Assessment/Restoration


We will consider "representativeness" when probability sampling cannot be employed. From this examination, we will then extend multilevel statistical models to provide new techniques for working scientists. The extensions include: 1) multiple response variables, 2) nom-linear functional forms, 3) missing data, 4) disturbance covariance matrices allowing for temporal and spatial dependencies, and 5) latent variables. In so doing, we will not only provide better tools to determine "representativeness" but also a convenient means to properly analyze data at different spatial and temporal scales (e.g. satellite data and survey data). We will write software for the extended multilevel statistical models. Finally, we will illustrate the use of these models with three very different data sets, two of which were collected as part of an EPA-funded project to study the Los Angeles basin watershed.


The proposed research builds on a number of rich traditions in statistics. While we will need to do some original statistical work, we will primarily be assembling and integrating a number of existing techniques into the multilevel statistical framework and then writing the necessary software. The empirical examples will exploit data that are already available in machine readable form.

Expected Results:

We expect to provide new statistical procedures working scientists can use to better generalize their results. With better statistical procedures for generalizing, the risk assessment generalizations will be on more sound footing.

Publications and Presentations:

Publications have been submitted on this project: View all 1 publications for this project

Supplemental Keywords:

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 methods

Progress and Final Reports:

  • 1999 Progress Report
  • Final Report