Cohesion and Listening Comprehension Tests: An Analysis of Participant Identification and of its Perceived Difficulty in a sample of English Proficiency Tasks

The current work, entitled Cohesion and Listening Comprehension Tests: An Analysis of Participant Identification and of its Perceived Difficulty in a sample of English Proficiency Tasks, investigates the listening comprehension complexity of Participant Identification through a comparative analysis of different proficiency tests. The aim of this study is to examine the complexity of identifying and tracking participants in a sample of listening tests. It explores the usefulness of cohesion as a tool to track participants, particularly how referential distance, antecedent ambiguity and anaphor type vary across different proficiency tests and levels. Finally, it highlights which participants are harder to identify and which strategies facilitate or challenge participant identification in listening tests. The research is based on a mixed-method analytical approach combining quantitative and qualitative approaches. Firstly, a qualitative textual analysis of listening transcripts was carried out and template graphical representations were obtained to show the way and frequency of participant tracking. Secondly, a questionnaire was carried out by 20 test-takers to identify candidates’ perceptions regarding a set of listening comprehension tasks they performed. Then, an analysis of difficulty was conducted to investigate the relationships between test takers’ evaluations of the text difficulty and their actual performance. The results show how cohesive features used to track participants vary across proficiency levels in terms of their complexity and distribution. In addition, the findings explore the participants that proved harder to identify and the strategies mostly used to track them. It is important to highlight that this study is limited because of the small sample of tests and of the minimal number of test-takers. Further future research should investigate listening participant identification complexity among a larger range of listening tasks.