EVALUATION OF THE PRACTICAL EFFECTIVENESS OF STATISTICAL TOOLS IN SOCIOLOGICAL RESEARCH ON THE EXAMPLE OF IDENTIFYING STRESSFUL SITUATIONS IN LIFE OF UKRAINIANS

Abstract

An example of studying of indicators of stress due to loss of faith in people and own strength shows the problems of evaluation of the obtained sociological data. These problems are caused by the lack of effective interaction between theory and practice in sociological studies. The historical stages of the development of statistical methods of processing the results of sociological surveys are considered as the main tool for measuring public opinion on many issues. Analysis of recent publications has shown that the majority of studies is devoted to analytical methods of clinical trials, epidemiological studies, statistical genetics, econometrics, image analysis and marketing. The methodological issues of quantitative analysis of sociological data are less reflected and need to be developed in accordance with current digital trends, as well as criticism and distrust in the media to the latest official census data. The comparative characteristics of the methodology of conducting a sociological survey of two research centers are presented: Kyiv International Institute of Sociology and Pew Research Center (USA). It has been shown that the following errors occur in the processing of sociological surveys: small particle sampling error, sub sampling error and general poll error. As a rule, in the survey of sociological data the sample of respondents is divided into subgroups by weight indicators: age, gender, place of residence, level of financial security and more. The margin of error reported relates to estimatesthat take the entire sample into account, and since the number of subgroup surveys is smaller, the error in their estimation is greater. Also, an example of the analysis of respondents' distribution on two stressful situations (to lose faith in people and to lose faith in own strength) shows, in the context of subgroups, the significance of the difference between the two shares. It’s important to take it into consideration while visually analyzing the results of the study, since the visual difference of the subgroups can be offset by the margin of error of a subgroup and the lack of significance at a given level of trust.