The paper is concerned with the methods of analysis of texts and lexical units such as collocations, colligations, chunks, clusters, hedging, etc. The aim of the research is to equip linguists, teachers and students with the skills of objective language analysis and adequate use of modern English language. The research is based upon statistical instruments of corpus manager. Cognitive interpretation of the data is presented. Conclusions are based on the results of compiling an original corpus for teaching British Studies, including 1 700 000 entries, and on analyzing professionally significant ones in a data-driven way. Implementation of corpus instruments such as concordances, collocation and colligation profiles, statistics, frequency characteristics, key-words search, word-clouds and others are under consideration. British National Corpus is used as a reference tool. Technically the research is fulfilled with the help of corpus manager WordSmith Tools 6.0.