The message focuses on the methodological aspects of improving reliability of the results of the content analysis of the texts of mass communications, made in different techniques of this method: automated and manual. Statistical analyses the results of the matches of the measuring scales the information content of newspaper reports concerning their purpose, obtained in automatic and manual modes. Correlation of these scales to certain categories of content analysis is discussed: the text reflected in the sphere of social life, type of author, timeliness, genre material. On the basis of proven mathematical tools of data analysis of major sociological projects, the use of automated methods of content analysis as the basic categories of modality, sign, orientation judgments, the subject and forms of activity in the text, subjects and type of views and at different levels of text analysis: a means, a channel of mass communication, a single message, its elements, entities and forms of activities within the boundaries of various localities, the acts of reception of specific messages. Relationship between the basic semantic features of informative text is revealed with some parameters of the production media and the prospect of practical use of obtained data in the development of indicators of the analysis of the media content.