The effect of test type on reading comprehension in English as a Foreign Language: The case of recall protocol and multiple choice.

摘要

The present study investigated the influence of different testing methods on students' reading comprehension of English texts. Specifically, it explored how the readers' ability in the second language (L2), their reading ability in the first language (L1), and the language of the test affected performance in reading comprehension when different testing methods were used.;Two types of testing reading comprehension were used in this study, namely, recall protocol and multiple choice. Two hundred twenty-six high school students in the Hebron district in Palestine participated in the study. Half of them (n = 113) were tested in reading comprehension by recall protocol, and the other half were tested by multiple choice. To study the contribution of the language of the test, one half of each group were asked to respond in their native language (Arabic) and the other half were asked to respond in the target language (English). Two authentic texts with different content were used to test students reading comprehension in English as a foreign language (EFL). To measure the students' reading ability in L1 (Arabic), they were asked to read two authentic texts in Arabic and to respond to recall protocol questions. The students' scholastic GPA in EFL was considered their language ability in English.;The relationship among the variables were determined through correlations. Multiple regression analyses were used to determine the contributions of the independent variables, the language ability in English (LAE), the reading ability in Arabic (RAA), and the language of the test (LT) to the dependent variable, reading comprehension when recall protocol (RP) is used as a testing method and when multiple choice (MC) is used as a testing method.;The results of the present study showed that the contributions of the independent variables, LAE, RAA, and LT differ significantly when different testing methods were used.;When RP was used as a testing method, the three independent variables accounted for 65% of the variance in reading comprehension. However, the individual contributions of the different variables were 49% by LAE, 10% by RAA and 6% by LT. When MC was used as a testing method, the three variables collectively accounted for only 21.5% of the variance in reading comprehension (15% by LAE, 6% by RAA and.004% by LT). This study provides a strong quantitative evidence that the testing method of reading comprehension in EFL has a significant effect on the readers' performance. Implications for research, theory and pedagogy are presented and discussed.