The aim of this study is to analyze the use of English vocabulary in English listening tests for high school students. This study analyzed single words, lexical bundles, and delexical verb collocations in the listening test scripts of CSAT, GSAT and GLT from the year of 2011 to 2020. The corpora were constructed with 210 listening test scripts. The investigation of single words first showed that test items under 2009 Revised National Curriculum used almost 20% of words that were not in the basic word lists, which means test items used much more difficult vocabulary than those National Curriculum guided. However, under the 2015 Revised National Curriculum, all test items used 10% of words not in the lists. Second, the word ‘go’, ‘good’, ‘look’ have appeared most frequently in all listening tests regardless of the school grade. The investigation of lexical bundles reveals that stance function bundles were used most frequently in GSAT, and GLT, but in CSAT, special function bundles (i.e., inquiry function) were used more than stance function bundles. In GSAT, the frequency of stance bundles increases as the school grade of test-takers rises. Overall, lots of lexical bundles were used in common among tests irrespective of test takers’ school grades. Lastly, the analysis of delexical verb collocations shows that delexical verb+noun collocations in GSAT were markedly less used than those in other listening tests. Especially in the case of the verb ‘have’, have+noun collocations were used only about 9% of the verb’s overall occurrence in the whole texts. The findings show there is unbalance in the use of delexical verbs in listening tests and there is no hierarchy in the use of delexical verbs in relation to the school grade of the test takers. These results provide meaningful implications to teachers in teaching English listening skills and developing listening test items.
Figures & Tables
Fig. 1. Analysis of lexical bundle’s functions by tests