Data on disability are collected in many statistical surveys, but the largest ones are censuses. The survey of disability based on census data in Poland has a long history dating back to 1921. During this period both the methodology of the survey and the scope of the data have been changing. The biggest modifications took place in 2011. NCP 2011 was the first census in Poland to be carried out by combing administrative data sources and the survey sampling method. The aim of the present study is to analyze and assess changes in the disability survey in NCP 2011 compared to NCP 2002 in terms of methodology and the scope of information.
There are many sample surveys of populations that contain outliers (extreme values). This is especially true in business, agricultural, household and medicine surveys. Outliers can have a large distorting influence on classical statistical methods that are optimal under the assumption of normality or linearity. As a result, the presence of extreme observations may adversely affect estimation, especially when it is carried out at a low level of aggregation. To deal with this problem, several alternative techniques of estimation, less sensitive to outliers, have been proposed in the statistical literature. In this paper we attempt to apply and assess some robust regression methods (LTS, M-estimation, S-estimation, MM-estimation) in the business survey conducted within the framework of official statistics.
The article describes a hybrid approach to evaluating economic efficiency of medium-sized manufacturing enterprises (employing from 50 to 249 people) in districts of Wielkopolska province, using metric and interval-valued data. The hybrid approach combines multidimensional scaling with linear ordering. In the first step, multidimensional scaling is applied to obtain a visual representation of objects in a two-dimensional space. In the next step, a set of objects is ordered linearly based on the distance from the pattern (ideal) object. This approach provides new possibilities for interpreting linearly ordered results of a set of objects. Interval-valued variables characterise the objects of interests more accurately than metric data do. Metric data are atomic, i.e. an observation of each variable is expressed as a single real number. In contrast, an observation of each interval-valued variable is expressed as an interval. The analysis was based on data prepared in a two-stage process. First, a data set of observations was obtained for metric variables describing economic efficiency of medium-sized manufacturing enterprises. These unit-level data were aggregated at district level (LAU 1) and turned into two types of data: metric and interval-valued data. In the analysis of interval-valued data, two approaches are used: symbolic-to-classic, symbolic-to-symbolic. The article describes a comparative analysis of results of the assessment of economic efficiency based on metric and interval-valued data (the results of two approaches). The calculations were made with scripts prepared in the R environment.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.