The meaning of the word representativeness. General and sample populations

The property of a sample, due to which the results of a sample study allow drawing conclusions about the general population and the empirical object as a whole, is called representativeness.

Representativeness (representativeness) of the sample- this is the ability of the sample to reproduce certain characteristics of the general population within the limits of permissible errors. A sample is called representative if the result of measuring a certain parameter for a given sample coincides, taking into account the permissible error, with the known result of measuring the general population. If a sample measurement deviates from a known parameter of the general population by more than the selected level of error, then such a sample is considered unrepresentative.

The proposed definition first establishes relationship between sample and population research. It is the general population that the sample represents, and only the general population can be extended to the trends identified in the sample study. Now it should be clear why earlier such attention was paid to the problems of correctly defining the general population and its description in research documentation and publications. The sample cannot represent a different population than the one from which the units for measurement were actually selected. If the researcher is mistaken about the actual boundaries of the general population, then his conclusions will be incorrect. If he mistakenly or intentionally expands or distorts the boundaries of the general population in reporting materials, publications, presentations on the results of the study, then this misleads users and may be considered as falsification of the results.

Representativeness is checked by comparing individual parameters of the sample and the general population. A common misconception is the existence of "generally" representative samples.

The representativeness or non-representativity of a sample can only be established in relation to individual variables. Moreover, the same sample may be representative in some respects and unrepresentative in others.

As a rule, in the professional discourse of sociologists, representativeness is presented as a dichotomous property - the sample is either representative or not. But this is not quite the correct approach. In fact, the sample may reproduce some parameters of the general population more accurately, while others less accurately. Therefore, it is more correct (although from a practical point of view and less convenient) to talk about representativeness specific sample for specific parameters.

As in the case of the sample as a whole, the key point in determining the representativeness of the sample is to justify the error within which the sample is considered representative for the purposes of the study. The opposite is also possible - fixing the size of actual errors and stating the fact that the sample represents the general population with certain errors. Again, the nature of the use of the results of the study plays a key role in this. Therefore, the same sample can be considered sufficiently representative for some purposes (for example, to predict voter turnout in the upcoming elections), but not representative enough for others (for example, to determine candidate ratings and forecast voting results).

What parameters should be used to check the representativeness of the sample? First, there are few such parameters in most research situations. After all, it is possible to compare the results of a sample measurement with data on the general population only if the latter are available. And research is being done because such data is just not enough. Therefore, even at the stage of object modeling and subsequent development of tools, it is advisable to provide for the measurement of one or more control parameters for which data are available that characterize the general population. This will provide the necessary empirical basis for testing representativeness.

Secondly, one should strive to check the representativeness of the sample in terms of parameters that are essential for the subject area of the study. In modern practice, the control of representativeness by the main demographic parameters - sex, age, education, etc. has become widespread. These data are usually available for any territorial object, as they are recorded during population censuses and subsequently recalculated by statistical agencies using reasonable mathematical models. . For this reason, the mandatory inclusion of several demographic variables in the "passport" has become an accepted professional norm. However, this practice can be classified as naive and subjected to justified criticism. The point is that the main and generally available for comparison demographic parameters do not always play the role of structuring factors in relation to the subjects of sociological research. Their nature in itself is not social, and the impact on the objects of research is often quite indirect. Therefore, demographically representative samples may actually hide significant problems in the form of system errors and uncontrolled biases. On the contrary, the demographic representativeness of samples that are effective in terms of the goals and objectives of the study may turn out to be low.

Here is an interesting example from practice. In 2009, one of the research companies operating in the Urals carried out a survey in the town of Kizel in the Perm Territory. In the course of field work, the researchers encountered serious obstacles to recruiting the sample provided for by the research plan - the lack of a sufficient number of available respondents, and worsening weather conditions. Apparently, the research company was not fully prepared to carry out work on such a large-scale project. Its production facilities were pushed to the limit to provide a survey of 6,000 respondents over a fairly wide area within a week. As a result, the actual sample in many survey sites was, according to the researchers themselves, filled with everyone who could be recruited to participate in the study. The demographic quotas established by the terms of reference were violated in most areas of the survey. In some areas, the distortion of the proportions of filling the sample in relation to the quota task reached 2.5 times for certain categories of the population, which actually cast doubt on the very fact of using the quota sample. It seemed that the customer of the study had every reason to make reasonable claims against the researchers.

However, the examination carried out on behalf of the arbitration court found that such significant distortions in quotas and, accordingly, the obvious non-representativeness of the sample obtained in terms of the main demographic parameters practically did not lead to distortion of the research data! By re-weighting the data array, the experts obtained the effect of a sample that is representative of the controlled parameters. Almost all frequency distributions of data checked by experts showed statistically insignificant differences between the results of processing the actual and reweighted arrays. De facto, this means that, despite the gross violations of the survey technology and the practical disregard for quota tasks, the researchers provided the customer with the same data that he could count on if the sampling procedures were fully followed and demographic representativeness was ensured.

How could this happen? The answer is simple - the demographic parameters used to control the representativeness had practically no (and this was confirmed by correlation analysis) influence on the subject variables of the study - the population's assessment of the socio-economic situation and the parameters of its socio-political activity. In addition, the sample size was very large relative to the general population (in fact, the study covered a quarter of the adult population of the municipal district), which, as a result of the law of large numbers, led to the stabilization of the observed distributions long before the required number of respondents was interviewed.

The practical implication of this cautionary tale is that efforts and resources should be devoted to ensuring and controlling representativeness with respect to those sampling parameters that the researcher believes have a significant impact on the subject matter. This means that the parameters for control of representativeness should be selected specifically for each research project according to its subject specificity. For example, assessments of the socio-economic situation are always strongly related to the real well-being of the respondent's family, his position in the labor market and in the business sphere. Accordingly, it is these parameters that should be used to control representativeness. Another thing is that it can be difficult to obtain objective data characterizing the general population based on them. This requires creativity and, perhaps, a compromise. For example, the level of well-being can be monitored by the presence of a car in the respondent's family, because the statistics of registered cars in the region can be available.

Interestingly, research reports and publications almost always refer to representative samples. Are non-representative samples really that rare? Of course not. There are enough samples in research practice that are problematic in terms of representativeness for one or another parameter. Rather, there are even more of them than samples, the representativeness of which can be assessed not formally (by demographic parameters), but in essence. However, their public mention in the professional sociological environment, unfortunately, is taboo. And none of the researchers is ready to admit that the representativeness of his sample in terms of parameters that are essential for the subject area of measurement is problematic or unverifiable.

In fact, the discovery of signs of non-representativeness of the sample is not a disaster. First, the existing technologies for “repairing” (re-weighting) the sample in many cases make it possible to completely eliminate the effect of non-representation with respect to the parameter that disturbs the sociologist or his client. The essence of the re-weighting method is to assign certain categories of observations (in the case of a survey - respondents) to certain weight coefficients, compensating for insufficient or excessive actual representation of these categories in the sample. In the future, these weights are taken into account when performing all computational operations with the data array, which makes it possible to obtain distributions that are fully consistent with the balanced (corresponding to the calculated quotas) data array. Modern statistical programs, such as BRvv, allow you to make calculations taking into account weighting factors in automatic mode, which makes this procedure fairly easy to perform.

Second, even if a “good” representative sample is not obtained, “moderate” representativeness may be sufficient for many research problems. Recall that representativeness is more of a fit measure than a dichotomous marker. And only certain research tasks - mainly related to the accurate prediction of certain events - require really high (statistically confirmed) representativeness from the samples.

For example, in order to predict the market share of a new product in a marketing study, a sample is required that encompasses and represents potential customers. However, most often, marketers do not have sufficient data about who actually forms the circle of their customers, especially potential ones. In this situation, it is generally impossible to check the representativeness of the sample - after all, it is not known what parameters it should reproduce. Nevertheless, many marketing tasks are successfully solved, since statistically representative samples are not needed to identify customer preferences, respond to advertising materials, and analyze reviews of a new product - it is enough to cover a typical clientele that is easy to find right in stores. Non-representative samples are quite suitable for solving search problems, identifying strongly pronounced trends, analyzing the specifics of individual categories (represented by small independent subsamples), comparing such categories with each other (two-dimensional analysis), analyzing relationships between variables, and other tasks in which the accuracy of the obtained statistical distributions has secondary importance.

Let's get acquainted with three concepts that anyone who in one way or another comes into contact with sociological research needs to know: the general population, the sample population (sample), representativeness.

General population - these are all units of the object of study defined by the program. If we are talking about an all-Russian public opinion poll, it will be the entire adult population of Russia. Or all Moscow students, if we undertake to conduct a survey among them. Or all the homeless children of Kaluga, if we are going to undertake a sociological study on this topic.

Sample population (sample) - this is the part of the general population that we will directly investigate, that is, these are the people to whom we will turn with interview questions or questionnaires; those materials that we will study by the method of content analysis, etc.

Sometimes the sample is equal to the general population (for example, in the case when we interview all first-year students of the Faculty of Journalism of Moscow State University). But usually it is less, sometimes by several tens and hundreds of times. At the same time, the practice of sociological research has proved that in nationwide surveys it is enough to select 1.5-2 thousand people for surveys. If the sample is well, correctly, representatively formed, then it can provide objective information about the opinion of all Russians.

So, the main thing is to form the right sample. The sample size depends on the objectives of the study, the specifics and degree of homogeneity of the object of study, the fragmentation of the groups to be studied, and the planned degree of its representativeness. What does this magical and most important concept in empirical sociology - "representativeness" mean?

Representativeness- this is the correspondence, adequacy of the sampling population (sample) according to the main characteristics of the general population. If in the structure of the population 55% of women and 45%; men, then the sample should have the same ratio. The same can be said about age, profession, type of settlement, and so on. In short, the sample configuration must match the configuration of the population. This can be depicted in such a figure (Fig. 8).

The most important thing in a sociological study is the representativeness of the sample, because it is precisely with this that the accuracy and objectivity of the results obtained are connected.

The sample can be formed in different ways. But the main types are two representative and non-representative samples.

Representative samples

Probabilistic, or random, sampling It is based on the fact that any of the objects in the general population has an equal probability of being included in the sample. There are several types of probability sampling.

1.systematic selection. It is very popular and often used in social research. This means that, depending on the size of the sample, each individual is selected from the general population. n-th (6, 20, 45, etc.) object. For example, we are interviewing the adult population of one of the polling stations. We take electoral lists. Let's say they have 10,000 people. And we need a sample of 500 people. We divide the number 10,000 of the general population by the number 500 of the sample, we get 20. This means that we will select every twentieth voter from the lists.

Suppose that we need to interview Muscovites by phone and find out from them what program they are currently watching on TV. We take a reference phone book, count how many numbers are in it, divide this number by the number that we need to interrogate, and we get a step at which we will conduct a systematic selection of numbers.

The same can be done with houses on the streets if we interview our recipients at home. For example, on the even side of the street we go into every fifth house. Etc.

2.Selection by lottery or draw. This method is familiar to you when you throw in a hat, a vase, a box, for example, all the streets of Moscow and choose 20 on which you will conduct research. Regions, settlements, post offices, etc. can also be selected.

3. Selection by random number method. To do this, special mathematical tables of random numbers are compiled according to the number of samples and an object is selected that is previously marked with this number.

Quota sampling is formed in accordance with quotas (that is, objects that have a certain attribute by sex, age, place of residence, etc.), which in percentage terms correspond to the general population. Suppose we study the population of a small town and know what percentage of it is young, middle-aged and old, men and women, employed and retired. We must select for the survey people with these characteristics in the same percentage. This sample is close to probabilistic in terms of the degree of representativeness.

Stratified sampling differs from the quota in that artificially, in connection with the objectives of the study, layers, strata are formed, which are subject to study and, as a rule, they are equal in quantitative terms. The strata should be more homogeneous than the whole population. For example, we study the readers of different publications: AiF, Izvestia, Truda, Komsomolskaya Pravda, MK and form equal strata of readers of different publications, let's say 200 people each.

Zoned sampling usually used in the study of areas, often using a geographical map, a scheme of settlements, etc., from which certain units are selected for study. For example, regions are selected from different geographical zones of Russia, or Moscow districts. Sometimes the so-called geographical cross technique is used, when points are selected on the horizontal and vertical of this geographical cross. This is how the sample was formed in public opinion studies in the 60s at the Institute of Public Opinion at Komsomolskaya Pravda.

Serial, nested, clustered sampling does not work with units, but with nests, homogeneous groups (family, production team, student group, football match fans, TV viewers who watch TV in the same room, urban areas, etc.). Usually, in this case, a continuous survey is conducted.

The concept of representativeness in sociological research

In other words, representativeness is the quality of the sample. The sample may be representative or non-representative. If a large group of people was used in a sociological study, then the sample will be representative.

Definition 2

A sample is a selected number of elements of the general population. A representative sample is characterized by the fact that all elements of the general population are presented in the same proportion.

The representativeness of a sample of a sociological study is determined by two random components: errors that were made during registration and random errors.

Example 1

For example: if the object of sociological research is complex and has several elements, then more interviewers will be required. Not all interviewers are always well qualified, which can lead to registration errors. In contrast, conducting a sample study by interviewers who are more prepared and instructed leads to a decrease in the number of errors, that is, to random errors.

Sampling comes down to three main problems:

determine the sample size (that is, build a certain procedure in order for the sample to be representative);
determine the sample size (the number to be surveyed);
assessment of the quality of the sample (analysis of the accuracy of the results).

Remark 1

It is important to remember that the sample and population rates should not exceed 5%. If such a proportion is violated, then the conclusions of such a sociological study will not correspond to reality.

Sample types

Samples are divided into: random and purposeful.

Random sampling is the most accurate and representative. The essence of this sample is that due to random selection, all units of the general population have the same chance of being included in the sample. This type of sampling is usually used before elections, referendums and other public events. In addition to the fact that this sample gives us accuracy, it has difficulties in applying. In order to conduct a random sample, a sociologist must have a list of the elements of the general population, which is not always easy. Random selection requires a large sample size to get accurate results.

Varieties of random sampling are serial, regionalized, mechanical and others.

Serial or nested sampling has the form of series. It consists in the selection of individual elements (family, group, school, team, etc.), which are subjected to continuous research.
Zoned sampling is used when the entire data array needs to be divided into homogeneous parts. Such parts can be districts of the city.
The principle of mechanical sampling is that all elements of the general population are assigned to one list and the required number of respondents is selected from it through equal integrals. Mechanical sampling has the ratio of the general population to the sample. For example: If the general population is 2000 people, and the sample is 200, then this means that every tenth person is selected from the general list.

Purposeful sampling is a type of sampling where selection is carried out according to the criteria of accessibility, typicality, equality, etc. Purposeful sampling is divided into spontaneous, snowball and quota sampling.

A spontaneous sample is a sample of the first comer. The disadvantage of this sample is that it is impossible to pre-determine the population.
The snowball method is to build up information. Each interviewed respondent gives contacts of colleagues, friends, acquaintances who can take part in the study, etc.
Quote sampling. In this sample, all data is a quota. When using quota sampling, respondents are selected purposefully, adhering to the parameters of quotas. Characteristics that are selected according to quotas are gender, education, age, skill level, or others, which are determined by the goals and objectives of the sociological research itself.

Generally. Representativeness determines how much it is possible to generalize the results of the study with the involvement of a certain sample to the entire population from which it was collected.

Representativeness can also be defined as the property of a sample to represent the parameters of the general population that are significant from the point of view of the objectives of the study.

Example

Suppose the population is all the students of the school (600 people from 20 classes, 30 people in each class). The subject of study is the attitude to smoking. A sample of 60 high school students represents the population much worse than a sample of the same 60 people, which will include 3 students from each class. The main reason for this is the unequal age distribution in the classes. Therefore, in the first case, the representativeness of the sample is low, and in the second case, the representativeness is high (ceteris paribus).

Literature

Ilyasov F. N. Representativeness of survey results in marketing research. Sotsiologicheskie issledovaniya. 2011. №3. pp. 112-116.

Wikimedia Foundation. 2010 .

Synonyms:

See what "Representativeness" is in other dictionaries:

- (French representatif indicative, characteristic), representativeness, a measure of the ability to restore, reproduce the idea of the whole in its part or a measure of the ability to extend the idea of the part to include this part ... ... Philosophical Encyclopedia

The property of a sample to reflect the characteristics of the general population being studied. In English: Representativeness Synonyms: Representativeness See also: Sample populations Financial Dictionary Finam ... Financial vocabulary

Representativeness, characterization, exponentiality Dictionary of Russian synonyms. representativeness n., number of synonyms: 3 indicative (8) ... Synonym dictionary

- (from French representatif indicative) representativeness of a sample of economic indicators (most often statistical) used to analyze economic processes and phenomena. Representativeness depends both on the reliability of the available ... ... Economic dictionary

representativeness- and, well. representatifadj. Representation, display. NS 2. The light, graceful stylization of the urban landscape as an old engraving conveys the unique flavor of the era. The panel is characterized by solemnity and representativeness. bringing it closer to... Historical Dictionary of Gallicisms of the Russian Language

Validity of transferring the results obtained in the analysis of the sample population to the general population. Dictionary of business terms. Akademik.ru. 2001 ... Glossary of business terms

- (from the French representatif representing), in statistics, methods for determining the parameters of a sample population (parts of an object, a set), the study of which allows one to represent the state of the general with sufficient validity ... ... Modern Encyclopedia

- (from French representatif indicative) in statistics, the correspondence of the characteristics obtained as a result of selective observation to the indicators characterizing the entire population. The discrepancy between these indicators represents ... ... Big Encyclopedic Dictionary

- (from fr. representatif indicative) eng. representativeness; German Reprasen tativitat. 1. Representativeness. 2. Index k.l. observations in statistics and other sciences. 3. The property of the sample to reflect the characteristics of the studied general ... ... Encyclopedia of Sociology

- (from French representatif representing something) the most important property of a particular sample of information, consisting in the reflection (representation) of the features of the entire general population (for example, the entire cenopopulation). About the representativeness of the sample ... ... Ecological dictionary

representativeness- The property of the sample population to reproduce the parameters and significant elements of the structure of the general population. The term "representative sample" was introduced for the first time in relation to socio-economic research by the Norwegian statistician A. Kiaer in ... Technical Translator's Handbook

Books

Medieval Intellectual Culture, A. M. Shishkov. This textbook is a reference book on the history of the intellectual culture of the Middle Ages, as it was reflected in the works of philosophers, theologians, natural scientists and ...
The role of soil in the formation and conservation of biological diversity,. The collective monograph presents the materials and results of many years of research on the role and importance of soils in the formation and conservation of biological diversity, carried out by the…

The concept of R. is widely used in the practice of empiricism. research within so-called. sampling method, when the survey of the whole (general population) reduced to examining part (sample or sample population) followed by dissemination of the survey results of the sample to the general population. Here R. acts as a formal one and is usually expressed through the magnitude and boundaries of the interval (called confidence interval), in which with a given degree of confidence (or, as they say, with a given confidence level) may be determined. numerical characteristic of the general population. The value and boundaries will entrust. intervals are calculable and depend on the size of the general population, the sample size, the method of selection, the degree of reliability that you specify (confidence probabilities), the method of calculating the surveyed characteristic and its value for the sample population. The study of such dependences is carried out by formal R., which is developed within the framework of one of the sections of mathematical. statistics - the theory of the sampling method.

Wentzel E. S., Theory of Probability, M., I9604; Cockren, W., Sampling Methods, per. with English, M., 1976.

Philosophical encyclopedic dictionary. - M.: Soviet Encyclopedia. Ch. editors: L. F. Ilyichev, P. N. Fedoseev, S. M. Kovalev, V. G. Panov. 1983 .

REPRESENTATIVENESS

(from . representatif - representing something) - a sample population, which allows with a certain error - pre-specified or calculated on the materials of the actual. sampling - to identify the distribution of the studied trait in the sample with the distribution of this trait in the general population. R. is the representativeness of the sample in relation to the general population according to some given specific characteristics. R. depends on subjective and objective factors. Thus, the conformity of the selection method with the objectives of the study, the observance of the selection procedure, the correctness of the method of collecting information that affect R., are not achieved by any formal means and completely depend on the conscientiousness and preparedness of the researchers. Especially great is the danger of subjective errors in sociological. researches, in to-rykh selective data turn out by means of interrogation. Objective errors are associated with the selective nature of the study (sampling errors). These errors are amenable to formal calculation and depend on the fluctuation of the trait under study, the sample size, the method of selection, and the nature of the general population. Methods for calculating errors are developed by the corresponding sections of mathematical. statistics.

Sampling errors can either be specified before the survey is conducted to determine the sample size that guarantees errors that do not exceed those specified, or can be calculated from the survey results to determine whether they are reasonable to extend to the population.

Lit.: Romanovsky V.I., Elementary course of mathematics. statistics, 2nd ed., M.–L., 1939; Kramer G., Mathematic. methods of statistics, trans. from English, M., 1948; Mills F., Statistical. methods, trans. from English, M., 1958; Van der Waerden B. L., Mathematic. , per. from German, M., 1960; Yul J. E., Kendal M. J., Theory of Statistics, trans. from English, 14th ed., M., 1960.

Philosophical Encyclopedia. In 5 volumes - M .: Soviet Encyclopedia. Edited by F. V. Konstantinov. 1960-1970 .

Synonyms:

See what "REPRESENTATIVE" is in other dictionaries:

Representativeness, characterization, exponentiality Dictionary of Russian synonyms. representativeness n., number of synonyms: 3 indicative (8) ... Synonym dictionary

Validity of transferring the results obtained in the analysis of the sample population to the general population. Dictionary of business terms. Akademik.ru. 2001 ... Glossary of business terms

Books

Medieval Intellectual Culture, A. M. Shishkov. This textbook is a reference book on the history of the intellectual culture of the Middle Ages, as it was reflected in the works of philosophers, theologians, natural scientists and ...

Portal for the student. Self-training

Representative samples

Sample types

Example

Literature

See what "Representativeness" is in other dictionaries:

Books

See what "REPRESENTATIVE" is in other dictionaries:

Books

RELATED ARTICLES