Centre for Mathematical Statistics
Centre for Mathematical Statistics has been set up to perform nation-wide tasks concerning the methodology and carrying out statistical surveys on different domains to the extent that this requires the use of advanced statistical and mathematical methods.
The Centre fulfills assignments on planning and implementation of statistical surveys of public statistics using mathematical statistics, sample selection and estimation theory and the econometric methods. The Centre undertakes these tasks within the Office specialization on the national level (small and medium-sized enterprise statistics and statistics on living conditions) and on request of other public statistics units conducting surveys on other statistical domains.
The tasks of the Centre include in particular:
Tasks directly related to the application of statistical and mathematical methods:
- designing and improvement of sampling schemes, estimation, calculation of precision of the results of sample surveys,
- designing of imputation and data editing algorithms and implementation of imputation process,
- methodological work in the field of small areas statistics,
- development and improvement of methodology and implementation of data processing elements, requiring the application of statistical and econometric methods, i.e. inter alia:
- seasonal and calendar adjustments,
- analyses of surveys results based on econometric models using taxonomic methods and other advanced analytical techniques,
- the use of so-called hedonic methods, e.g. to correct price measurement, assess statistical phenomena by means of measures resulting from econometric estimates (e.g. calculation of the so-called imputed rent),
- analytical work on improvement the efficiency of statistical surveys.
- development and management of a system for the measurement, evaluation and monitoring of the quality of public statistics surveys,
- improvement of statistical frames to adapt them to more efficient methods of surveys conducting and estimation,
- drawing up statistical forms and questionnaires.