Abstract:

Drawing on Shaw and Weir’s theoretical framework for validating writing tests (2007), this paper highlights the issues of the writing constructs measured in English writing tests in university entrance exams, and recommends improvements. The paper analysed the writing response formats of 66 English tests used by Japanese universities and one English test of National Centre Exams (NCE) for 2013 entry. It was found that translation was the most commonly used skill in the writing tests, and accounts for 45% of the total. The most common writing response format used by the state universities was translation, whereas word-reordering was commonly in use at the private universities and NCE. Because word-reordering and translation tasks can assess very limited English grammatical and lexical discrete writing skills, there is no conclusive proof that the task can assess writing skills needed by the applicants to write cohesive texts in English. However, there are potential reasons why indirect writing assessments have remained a key method for Japanese university admission in the system of designing the English tests: the number of applicants and time constraints. Taking these factors into account, alternative English tests should be introduced to Japanese university entrance examinations.

Author: Yumiko Moore