It is very interesting to see the contents of the longest cohesive chains in the corpus. What was common in RAs and MA theses is that the eight longest chains (that were already studied in detail for their comparison with the contents of the abstracts) always contained the following referred entities: the participants of the study, the main research aim or topic, and the research tool used. However, the two sets of research papers differed in that referents in long cohesive chains in RAs contained the method of analysis (e.g., working memory assessment) besides the specific research tool (e.g., questionnaire), and they (RAs) also contained the main theories and concepts the study uses as a background. Instead of reference to the main concepts and theories that serve as a background to the study, MA theses had various sets of participants or sets of research tools in the longer cohesive chains, with a particularly high frequency of ties in the Methods section.