Scientific environment and publication activity: risks of bibliometric estimates

The analysis of approaches to evaluating science and scientific activity based on: a) quantitative estimates of publications and their citations - bibliometric indicators; b) estimates of research and development costs, the number of scientific personnel, etc.; C) technological results of research and development. Bibliometric indicators have become dominant in decision-making of a managerial nature. The article presents decisions at the state level on the use of bibliometric indicators in Russia. The state of the research environment in Russia was studied, which showed that there is a decrease in the number of research organizations and researchers with an increase in the number of publications. The features of working with bibliometric data are analyzed on the example of WoS CC (Clarivate Analytics), Scopus (Elsevier) and RSCI (Scientific electronic library LLC). It is shown that when evaluating publication activity, it is necessary to take into account the thematic area and the scale of the assessment. The choice of a specific research methodology should take into account the following levels: macro-level, meso-level, micro-level. The main types of inaccuracies are shown. Confirmed restrictions when working with bibliometric tools.