One of the objective empirical tool to draw conclusions in cognitive-linguistics research is corpus analysis. This article reports on data analysis applied in the study of language representation of sociocultural stereotypes. Application of this method is conditioned by the task to objectively describe categorial features of sociocultural phenomena and to verify the hypothesis of their stereotypic and symbolic nature. Sociocultural stereotypes are characterized by various explicit and implicit information comprising cultural schemata, emotional evaluation, allusions, and images, which can be revealed in the language. Since the appearance of language corpora, data-mining in investigation of represented in the language sociocultural phenomena has been simplified and methodology has become more sophisticated. Hybrid approach in this study is conditioned by the fact that cognitive status of the examined phenomena as fixed in culture typified images of social groups have been ascertained introspectively, while cognitive structure and changes in social perception of the stereotypes are revealed in the course of corpus analysis. Methodologically the work presents polyfactorial research that combines analysis of semantic and quantitative dimensions of verbalized sociocultural stereotypes. Based on distributive and statistical method, the research incorporates quantitative and qualitative techniques of verbal material. The results are formalized in frequency tables showing categorical features of the sociocultural stereotypes. As empirical method of study, corpus analysis avoids bias of introspectiveness. The acquired results and methods in study of linguistically represented stereotypes could assist the researchers interested in verbalized sociocultural phenomena. Keywords: sociocultural phenomena, empiric study, corpus analysis, categorical features, cognitive stricture.