Statistical Disclosure Control for Microdata by Matthias Templ

Author:Matthias Templ
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham

(5.5)

In the following code, a contingency table of (gender federal state) is computed for the anonymized data and original data. First, an object of class sdcMicroObj is created, and then we apply PRAM on federal state.

Let’s start from the beginning. First we create the sdcMicro object, then we apply PRAM and compare the original and table considering the prammed variable db040 (region).

We now compare the tables according to Eq. (5.5).

We see that the mean difference in the cell values of the tables is approximately .

However, contingency tables are of compositional nature (Egozcue et al. 2015) since no negative cell values are possible and not any cell can be larger than the sum of cells or the corresponding marginals in the table. To consider this, one can also compare the cells using Aitchison distances, i.e. since compositional data are represented only in the simplex sample space, we have to use a different distance measure, like the Aitchison distance. It is defined for two compositions and as

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.