This paper stresses the importance of rigour, systematicity and transparency in collecting and interpreting data in research, with a particular focus on corpus linguistics and research syntheses. Following separate brief overviews of each area, the two are then brought together by looking at how a corpus approach has been used to complement other types of research synthesis in two case studies. The first specifically involves English for specific purposes through two recent HDR publications, both of which stop at the level of counting key items in journal articles. The second shows how the approach can be taken further in an adjacent field, data-driven learning (DDL), with a corpus analysis of research papers drawing on frequency lists of words and n-grams, concordances, and keywords / key n-grams in particular to address specific research questions.