Content №4 от 2021
On Approaches to Ensuring the Representativeness of Non-Random Sampling
The article discusses the theoretical, methodological, and applied aspects of ensuring that the results of discontinuous surveys at the national, regional, and municipal levels are representative. It reviews the principles, methods, forms, and technologies of sample observations implemented by state and public institutions when various mass sample surveys are arranged and conducted. We also define the problem of obtaining non-random samples in the survey process and assess the practice of spreading their results to the general population. Without the basis for a general population (i.e., a unit listing), researchers must solve the inverse problem of the sampling method; this circumstance does not allow obtaining random samples, and besides, often there is no reasoning for the representativeness of the data obtained.
Viewing these problems with adequate statistical inference allows us to find an acceptable solution. We have developed and tested an algorithm to generate representative non-random samples on a wide range of surveys. The proposed methodology is used in arranging and conducting a sociological survey on enforcing the critical provisions of several federal laws. Ensuring the representativeness of the obtained non-random sample has made it possible to assess the level and degree of differentiation of citizens’ satisfaction with the measures taken to regulate public relations in the studied laws across federal districts and constituent entities of the Russian Federation.
Glinskiy V. V. [email protected]
