Qualitative analysis of data sets in the study of social and economic processes

Abstract

Introduction. Ssocio-economic phenomena research provides monitoring and gathering of information according to various indicators that characterize them. Analysis of socio-economic phenomena which is associated with handling large amounts of data information has been carried out. A set of evaluation data is characterized by heterogeneity and inaccuracy. Information is formed by many experts, so it can  contain questionable and unprocessed data. For efficient and reliable evaluation of socio-economic phenomena it is necessary to carry out a qualitative analysis of data sets. This will ensure the formation of a homogeneous set of data, extracting of atypical data with abnormally large positive or negative value, improving of the reliability of the results of the study.

Purpose. The aim of the article is to consider and develop a method of qualitative analysis of data sets in the process of evaluation of  the socio-economic phenomena.

Results. The overview of the most common methods for assessing general data set has been done. Attention is drawn to imperfect methods of qualitative analysis of data sets in the process of evaluation of the socio-economic phenomena. Qualitative analysis of data sets containing an analysis of lost and uncertain information has been carried out. Methods for lost and uncertain information treatment are recommended. If there is a loss of information then it should be used "reduce to zero", conditional coding, method of maximum likelihood of  principal components. If there is questionable information its identification and evaluation should be held. Identification is carried out according to the criteria F.Hrabbs, G. Moore, N. Smirnov, G. Tityen. On the basis of assessment of data the questionable information should be eliminated from further processing or shoul be modified. The variants of information modification using robust statistics methods are proposed. The latter includes methods of Ch. Vinzor, A. Poincare, Jh. Tukey, P. Huber. Qualitative analysis of the data sets in the process of evaluation of the socio-economic phenomena involves the detection and treatment of atypical information. We have represented a generalized scheme of qualitative analysis of data sets. The technique can be applied in the study of various economic activities. Further research will involve testing methods for evaluation of the efficiency of industrial production in Ukraine.