The paper addresses the methodological problems of evaluating the processes and outcomes of the international integration of national science, using the case of Russian research in the field of media and communication. Within the study, several questions emerge that, while looking seemingly simple and straightforward under formal approaches, in fact call for detailed and critical examination. In particular, the working definitions for the concepts “international integration,” “article on media,” “international journal,” “Russian author,” and “Russian article” had to be developed to ensure accurate data interpretation. To select integration parameters, quantitative data from the Web of Science (WoS) were used. However, the analysis went beyond standard indicators of publication activity and citation. Combinations of different indicators were applied, e. g. differences in citation between all journals and foreign journals indexed in the core WoS databases, the ratio of publications in formally versus genuinely international journals, the pool of sources cited by Russian authors, and others. The authors propose solutions to the key methodological challenges: refining publication selection criteria through combined search strategies and manual filtering; developing journal classification system to distinguish between formally and actually international publications; and creating typology of formally Russian authors based on their actual geographical affiliation. Several additional methodological issues were also addressed to obtain tangible results in the analysis of international ties, including authors chronological clusterization and countries grouping. This part of the article focuses on the key methodological problems mentioned above, excluding the differentiation between formally and factually Russian authors and articles. The findings demonstrate that without in-depth analysis of these methodological issues, relying on the indicators of international databases may lead to significantly distorted representation of reality. The proposed approaches can be adapted to refine scientometric assessments in other academic fields, particularly in the social sciences and humanities.