On the Dimensionality of Reading Comprehension Tests Composed of Text Comprehension Items and Cloze Test Items

Tabaran Institute of Higher Education. Shariati 60, Shariati Blvd, Ghasem Abad, Mashhad, Khorasan Razavi, Iran. Tel: +98 (51) 35227215; e-mail: [email protected]; Web site: http://www.ijlt.ir/

Y

15

Reading Comprehension; Reading Tests; Test Items; Test Format; Test Bias; Test Validity; Error of Measurement; Factor Analysis; Cloze Procedure; Multiple Choice Tests; College Entrance Examinations; High School Graduates; English (Second Language); Scores; Foreign Countries

2476-5880

Test method is deemed to affect test scores along with examinee ability (Bachman, 1996). In this research the role of method facet in reading comprehension tests is studied. Bachman divided method facet into five categories, one category is the nature of input and the nature of expected response. This study examined the role of method effect in the Iranian University Entrance Exam (UEE). Research showed the current test method of assessing reading can highly impact test takers? performance and thereby examinees? test scores. Hence, different test formats may extract various traits and therefore, contaminate test scores and threaten validity. The researchers using exploratory factor analysis (EFA) extracted a second method factor in the reading comprehension section of the UEE which is composed of a multiple-choice cloze passage and multiple choice questions based on sustained texts. Results of the analysis showed that cloze test can create construct-irrelevant method variance, contaminating test scores, and led to violation of the unidimensionality assumption of the test.

As Provided

2021

EJ1299336