Online Tutorials

Tutorial 1:  Introduction

In tutorial 1, we introduce the problem and provide an overview of the maximum entropy imputation and spatial allocation procedure.  For users unfamiliar with U.S. Census geography, a brief primer on the geographic units used in the analysis is given.  This is followed by a review of the software necessary to complete the remaining tutorials.  Although the weights imputation is run in the R software environment, advanced knowledge of R is not a prerequisite, and the pre-processing and post-processing of the data can be carried out using any statistical or spreadsheet software.

Tutorial 1 can be downloaded here.

Tutorial 2:  Data Download and Preparation

Tutorial 2 covers the pre-processing of the data required for the weights imputation and spatial allocation.  This includes the retrieval, recoding, and restructuring of the summary data and the microdata, critical for success in the remaining tutorials.  The selection of constraining variables is also covered here.

Tutorial 2 can be downloaded here.

Tutorial 3:  Weights Imputation

In Tutorial 3, the weights imputation is run and the resulting tract-specific weights are output for use in the spatial allocation.

Tutorial 3 is currently under development.  Contact Matt Ruther for more information.

Tutorial 4:  Spatial Allocation

The spatial allocation of the microdata may be performed in a number of ways, depending on how the tract-specific sampling weights are interpreted.  These allocation possibilities are covered in Tutorial 4.  The final product from this Tutorial is the revised tract-level summary statistics.

Tutorial 4 is currently under development.  Contact Matt Ruther for more information.

Tutorial 5:  Evaluation

Once the revised tract-level summary statistics have been estimated, their accuracy should be assessed.  Of course, the rationale for the generation of these estimates is that they did not exist prior, so precise validation of the estimates is rarely possible.  There are, however, some methods by which evaluation can be carried out.  These methods are covered in Tutorial 5.

Tutorial 5 is currently under development.  Contact Matt Ruther for more information.