Presentation of Census Results by Interactive Statistical Models

2001 Czech Census Model

For the sake of estimating the statistical model of the 2001 Czech Census we have chosen 24 categorial variables (questions). To illustrate a general possibility of information fusion from different sources we have combined two originally separate databases of individuals and households. In particular, the first ten variables from the database of individuals have been merged with fourteen variables of the corresponding household.

Basic data properties

Number of respondents10230060
Number of questions24
Answers per question2-14
Mean absolute error407
Mean relative error4 %