The main purpose of the present study was to investigate the four types of cloze test by comparing the original C-Test with the New Constructed C-Test (the NC-Test, and the original Modified C-Test (the MC-Test) with the New Modified C-Test (the NMC-Test) in measuring the English language proficiency of first-year undergraduate students studying in the science program at Mahidol University. The 110 students in high- and low-language-ability groups were divided into four sub-groups taking four types of the cloze test. Moreover, this study aimed to examine the test-taking strategies these target students used to restore deleted items in the C-Test, the NC-Test, the MC-Test, and the NMC-Test in measuring English language proficiency. The results of the study may provide an alternative way for language teachers to assess the English language proficiency of non-native-speaking university students.
The statistical devices used in this study were the mean and standard deviation reporting the average scores for the four types of the cloze test in measuring the students' English language proficiency. The independent sample t-test was employed to analyze whether there were any significant differences between the students' test scores on the original C-Test and the NC-Test, and between the original MC-Test and the NMC-Test. In addition, item analysis was also used in this study to see whether the third-word deletion in the NC-Test and the NMC-Test affect the test discrimination power. For the test-taking strategies, the students' responses obtained from group interview were reported by calculating the frequencies of what test-taking strategies these volunteer students used while taking the four cloze tests.
Get your grade
or your money back
using our Essay Writing Service!
The mean scores on the cloze tests were calculated by the independent sample t-test to find whether the original C-Test yielded any different results from the NC-Test in measuring the students' English language proficiency within high- and low-language-ability groups. These findings show that there are no statistically significant differences at the .05 level of significance (p < .05) between the mean scores on the C-Test and the NC-Test within the high- and low-language-ability groups. This indicates that the NC-Test can be a substitute for the original C-Test in both high- and low-language-ability groups. In addition, the test scores on the original MC-Test and the NMC-Test, calculated by the independent sample t-test, showed that there were statistically significant differences between the mean scores on the MC-Test and the NMC-Test in the high-language-ability group at the .05 level of significance (p < .05), whereas these findings found no statistically significant differences at p < .05 between the mean scores on these two language tests in the low-language-ability students. However, the high- and low-language-ability students made higher scores on the NMC-Test than on the original MC-Test.
In this study, third-word deletion was employed in the NC-Test and the NMC-Test to provide more clues for the EFL university test-takers. These findings reveal that using third-word deletion has an influence on the discrimination power of the test, such as high or low discrimination power. When comparing the original C-Test with the NC-Test, third-word deletion gave lower discrimination power to the NC-Test than to the original C-Test. The discrimination indices of the NMC-Test were slightly higher than those of the original MC-Test. Another factor that affected the four types of cloze test in this study was the type of deleted words. For example, deleted functional words could be restored by using only linguistic or grammatical competence. Consequently, language teachers should be aware that different types of deleted words and different deletion techniques can have an influence on these four types of cloze test in measuring the English language proficiency of EFL tertiary students.
The interview data in this study reveal that the original C-Test and the NC-Test can measure language ability at the sentence level, so these two language tests are suitable for measuring specific knowledge of English, such as grammar or vocabulary. The students' responses also show that these volunteer students used the Across Sentences, Within Paragraph strategy, indicating that these two type of cloze test can measure the English language proficiency of the target students within an EFL context.
6.2 Recommendations for Further Study
This comparative study was limited to the first-year undergraduate science students studying in the Faculty of Science at Mahidol University. Further studies should be conducted to examine the four types of the cloze tests: the original C-Test, the NC-Test, the original MC-Test, and the NMC-Test taken by university students in different academic fields. Moreover, there should be a further study using the different types of the cloze test in different testing situations, such as reading proficiency. Boonsathorn (2000) also proposed the latest form of the cloze test, the S-Test (Semantic/Syntactic Test) by deleting half or part of every second content words. Example 22 provides the S1-Test in which the first half or the first part of every second content word is deleted, and the S2-Test in which the second half or the second part of every content word is deleted.
Always on Time
Marked to Standard
In addition, an investigation of the test-taking strategies for cloze tests, including the S-Test, should be carried out by using a greater number of students to better establish the validity of the research findings. For English language teachers, these findings provide several choices in measuring English language proficiency. Language teachers may use the C-Test and the NC-Test in measuring the specific knowledge of English language ability, such as vocabulary and grammar, and use the MC-Test and the NMC-Test as a part of testing English language proficiency of EFL university students.