IUJ Repository (International Univ. of Japan)
Not a member yet
645 research outputs found
Sort by
Islamic Revival and the Political Leadership of the 'Ulama': A Study on Muhammad Baqir al-Sadr's Political Guidance of the Religious Authority
Get PDFdepartmental bulletin pape
A Perspective on Endogenous Development in the Middle East: Economic Liberalization in Iran
Get PDFdepartmental bulletin pape
WHAT TEXT CHARACTERISTICS PREDICT HUMAN PERFORMANCE ON CLOZE TEST ITEMS?
Get PDFThis study explores the link between some of the linguistic characteristics of the text surroundding cloze test items and the corresponding item difficulty estimates. Fifty reading passages were randomly selecdted from an American public library andd made into thirty-item cloze tests by deleting every 12th word. The subjects were 2298 EFL students from 18 university level institutions in Japan. Each student took one of the 30-item cloze tests. The 50 cloze tests were randomly assingned across all of the subjects. Any differences between the cloze tests or the indeividual test items were therefore assumeddx to be due to other than sampling differences. The result was a set of 1500 item difficulty estimates (50 teats times 30 items). Each item was also analyzedd for linguistic characteristics, e.g., passge readability, number of words per sentence, frequencies of occurrence in passage(s), and others. Correlational, factor and multiple-regression analyses of the linguistic characteristics indicated that there were clear groupings among the variables and showed that linguistic characteristics can be used to account for up to 31 percent of the variance in item difficulties. These results are discussed in terms of their implications for cloze testing research.conference pape
Basic Considerations for the Socio-cultural Analysis of the Islamic World
Get PDFdepartmental bulletin pape
Measuring Editing Skills: Passage Correction and Peer Feedback
Get PDFThis paper reports on research currently in progress on the use of Passage Correction (PC) tests as a measure of editing ability. This is one aspect of an overall study on the value of peer versus teacher feedback for improving overall scores on a standard composition profile. Subjects are 138 first-year students at two junior colleges in Osaka. Subjects were administered pre- and post-test PCs and pre- and post-test compositions. No significant differences were found between the peer and teacher feedback groups. Findings indicate that PC tests may be an effective measure of proficiency in editing skills in comparison to NS ability. However, they appear to give a conservative estimate of learner progess and may not provide the information teachers need. As such, they appear to have limited value as achievement tests.conference pape
