the statist. observation, with which to a study subject
not all elements of the studied totality (called in this case
"general"), but only some, in a specific manner their selected part.
The selected part of the elements of totality (sample) will
represent entire totality with the acceptable accuracy with two
conditions: it must be sufficient numerous so that in it the
regularities, which exist in the general population, could appear;
selection elements must be selected objectively, independent of
the will of researcher, so as to each of them would have identical
chances to be selected or so that these chances would be known to
researcher. These conditions are established matematich. by the
theory of sampling. It is based
on a number of the most important theorems of the probability theory,
which compose Vol. n. law of large numbers (see the
large numbers law). Only with the observance of
these conditions the objective possibility to estimate accuracy v. n.
on the basis of themselves selected data appears. Accuracy v. n.
is measured with the aid of the mean error of the sample, whose the
value is directly proportional to the degree of a variation in those
studied it is indicative and it is inversely proportional to the
sample size. However, v. n. it is possible to produce more rapid
than the continuous, with the smaller expenditures and to obtain
results, on the accuracy little being inferior to the results of
continuous observation, and taking into account the possibility of
more thorough observation - even frequently exceeding them. With
social- economic studies for the selection in the majority of the
cases is required the basis of sample, i.e. list or the enumeration of
ones, from which will be conducted the selection. Objects in the
locality, for example, houses, populated areas, sections of territory,
are conveniently selected on the map. Is useful also the some
preliminary information about the nature of the studied totality for
the correct calculation of the sample size. Impressiveness, or
representativeness, samples are ensured not only by its volume, but
also by strict observance of the scientifically substantiated
selection rules, which guarantee its objectivity. The methods of
selection are very diverse. In the social- economic inspections
is extended sistematich. (mechanical.) selection, i.e. the
selection of ones on their list through the established interval.
More rarely adapts the simple random selection, with which ones
are selected on the lot, according to the table of random numbers or
by another analogous method. If preliminarily there is
information about the totality being subject to study, then it they
divide on more or less uniform, typical groups produce selection from
each such group separately, obtaining the typical or stratified
sample. To select it is possible both the sections elements (for
example, people) and group of such elements (for example, family).
In the latter case the selection is NAZ jack, or series.
With the inspections of large scale the sample is produced
usually in several steps, i.e. first select the larger units (for
example, the populated areas), and then in them - smaller units
(family). The different methods of selection in practice usually
combine.
V. n. widely practiced by Russ. pre-revolutionary by
zemstvo statistics. Some methods, is in particular the highly
evaluated By v.I.Leninym polyphase selection, did not lose value,
also, to real, time. TSSU OF THE USSR regularly carries out
inspection ok. 62 thousand budgets of the families of workers,
employees and collective farmers, and also conduct yedinovrem. of
inspection in the different regions of social- economic statistics.
Selectively obtained the part of information with the All-Union
population census 1970. Sample surveys widely practice by
scientific establishments, in particular with sociological studies.
It is developed even tyuey samostoyat, region v. n. - quality
control of industr. production.
Lit.: Kowalevski a. g., the bases of
the theory of sampling, Saratov. 1924; Boyars' A.
4., Staroyeskiy v. n. [ and of other ], the theory mathematical
statistics, M., 1930 and M., 1931; Yu l George. 3. and
Kendel M. dzh., the theory of statistics, translated from English, 14
publ., revision, and rasshir., M., 1960; And against and t with
f., sampling in the censuses and the inspections, the translated from
English, M., 1965; Selective observation in the statistics OF
THE USSR. Coll. of articles edited by A. 4. boyars' [ and
other ], M., 1966; Guards and nN. K., sampling and its
application in social and economic studies, M., 1970. A. g. volkov.
