In 1980s, a national college English curriculum was established in China, which divided college English teaching into six bands. Band 1 to Band 4 is required, while Band 4 to Band 6 is optional. Each band corresponds to one semester; therefore, students are expected to reach Band 4 at the end of their second year in college. In order to motivate fulfilment of the national curriculum and evaluate college students' English ability, the College English Test (CET) was launched in 1987, including Band 4 (CET-4) and Band 6 (CET-6). Almost all the tertiary institutions in China require their students to take the CET-4, and those who have passed the CET-4 can choose to take the CET- 6. Traditionally the CET-4 consists of five sections: Listening (Dialogue and Short passage), Reading (Intensive reading and Fast reading), Vocabulary and structure, Cloze, and Writing. Except for the writing section, the test mainly uses multiple choice items. The CET has been generally stable despite some minor changes and adjustments. However, since 2006 the weight of the Listening section has been raised from 20% to 35%; the weight of the Reading section has being reduced from 40% to 35%; the Vocabulary and Structure section is removed; Cloze section still accounts for 10%; Writing section accounts for 15% as before; and the remaining 5% is Chinese to English Translation. In collaboration with the British Cultural Council, the CET Committee conducted a large-scale project to examine the validity and reliability of the test. They reached the conclusion that the CET can accurately measure college students' English ability and the reliability index is over .85 (Yang & Weir, 1998).
In the past decade, the CET has become the most influential English test in China and also the world's largest language test administrated nationwide (Jin & Yang, 2006). For example, over 11,000,000 test takers took the test in 2004 (Ministry of Education, 2005). The CET is now a highstakes test with massive educational and social impacts in China (Cheng, 2008; Yang, 2003). According to a national survey by Yu (2005), 81.7% of the Chinese universities regarded passing the CET-4 or getting certain grades on the CET-4 as the precondition for earning a bachelor's degree. Many employers also list the CET certificate or score report as one of their primary job requirements. The impact of the CET has thus become a hot issue not only in academia but also in the whole Chinese society. On the one hand, many assert that the CET has greatly propelled the enforcement of the national curriculum and made a great contribution to college English teaching (Gu, 2005; Yang, 2003). On the other hand, the CET is accused of leading teachers to teach to the test, which could narrow the curriculum and turn regular English classes into CET coaching (Cai, 2005; Han, Dai, & Yang, 2004). However, most of the arguments are based on personal reflections instead of well-designed empirical studies. Therefore, there is not much empirical evidence on what is truly happening in the classrooms as a result of the CET.
To further explore the influence of the CET based on empirical data, this study aims to find out whether teachers are truly teaching to the test and the potential reasons involved. As the CET targets at different language skills which may be covered in different classes, this study investigates only its writing section in order to have a more focused view of how teachers teach in the classrooms.
Theories have been proposed on the influence of testing on teaching and learning, a phenomenon commonly referred to as "washback". In the earlier stage, a major concern is that teachers may ignore the content not directly related to the test, and students may just practice exam techniques instead of learning real knowledge (Davies, 1968; Vernon, 1956; Wiseman, 1961). Alderson and Wall (1993) later hypothesized the general influence of tests on teaching as follows:
* A test will influence teaching
* A test will influence what teachers teach and how teachers teach
* A test will influence the rate, sequence, degree, and depth of teaching
* A test will influence attitudes to the content, method, etc. …