A Comparative Study Of The Difficulty Of Error Correction In CET-6 And TEM-8

CET and TEM are two influential nationwide test batteries in China. Error Correction (EC) is an important component of CET-6 and an indispensable component of TEM-8. Difficulty is an important consideration in establishing a test's validity and reliability. In theory, EC in TEM-8 should be more difficult than that in CET-6. However, no research has ever been done to confirm or refute this idea. The present study is intended to compare the difficulty of EC in CET-6 and TEM-8 and probe into the possible contributing factors accounting for the results of the comparison.There are two research objectives in this study. The first objective is to find out whether all the ECs in TEM-8 are more difficult than those in CET-6 or vice versa and the second objective is to find out how vocabulary, sentence length, Error Correction item (ECI) type, and the proportion of excessively difficult item (EDI) contribute to this situation based on the above comparison result.In order to achieve the research objectives, two research designs were employed: a major design and a minor one. The subjects involved in this study were 193 English major juniors, 163 from School of Foreign Languages, Yangzhou University and 30 volunteers from School of Foreign Languages, Yancheng Teachers College. They were required to do some ECs employed in CET-6 and TEM-8.The major findings of the present study are summarized as follows:Firstly, all the ECs in TEM-8 are more difficult than those in CET-6. However, not every ECI in TEM-8 is more difficult than that in CET-6.Secondly, the most contributing factors in making the ECs in TEM-8 more difficult than those in CET-6 are sentence length and the proportion of EDI.Thirdly, sentence length, the proportion of EDI, ECI type and vocabulary work together to determine the difficulty of EC. Besides, the four factors have inter-effect on each other.The above mentioned findings provide implications for English learning, English teaching and language testing.First, accuracy should still be emphasized in English teaching and learning.Second, students'EC ability should be improved.Third, test designers should attach great importance to the difficulty of each item when designing test papers. If the proportion of EDI in EC is too high, its validity and reliability will be called in doubt.Fourth, the proportion of beyond-sentence-level (BSL) ECIs should be raised so as to meet the requirements of national syllabus for the English majors, to improve EC's validity and sharpen students'discoursal awareness in their English learning.However, there are three major limitations to the present study. Firstly, the number of original ECs in TEM-8 and CET-6 selected for the present study is not big enough for the findings to be more generalizable. Secondly, the number of sample subjects doing all those 12 ECs is too small. Thirdly, to the limited capacity of the researcher, this study was only conducted quantitatively.