This report is a follow-up to a study by Shizuka (1999a; 1999b), in which a new scheme for classifying reading test items was proposed and tried out. In this scheme, an item was to be placed in one of nine categories based on its position on two dimensions, one concerning the size of the relevant text portion and the other relating to the depth of cognitive processing required to respond correctly to the item. In the trial phase, four testing experts were asked to apply the taxonomy to Cambridge KET, PET and FCE. While acceptable agreement rate was observed for the two lower-level tests (KET & PET), rather low inter-classifier reliability was revealed for FCE. The present paper focuses on detailed qualitative and introspective examination of the FCE items over which classifiers' perception differed to a considerable extent in Shizuka's (1999a; 199b) trial. The analyses revealed several different ways language testers view specific items, highlighting limitations of the adopted scheme.
The authors have conducted several studies concerning reading and listening comprehension processes of Japanese learners of English. Assuming that working memory is the crucial factor in these two types of processing, established reading and listening span tests are needed. However, so far, span tests have traditionally been conducted in a 'face to face' situation. It has been impossible to gather a large number of subjects or to adapt these tests for pedagogical use. This paper is intended as an experiment to explore the possibility of modifying span tests for a large number of subjects, i.e. for group testing. A Reading Span Test (RST) was made on a video tape, using a character generator, VTR and controller. A Listening Span Test (LST) was recorded onto a cassette MD tape, using an MD recorder. These tests were administered to students in a Language Laboratory room. The results are presented and some suggestions are made about how the results of this research can contribute to the improvement of current English teaching.
The purposes of this research are (1) to examine the appropriateness of using Mochizuki's (1998) Vocabulary Size Test as an indicator of estimates of general English proficiency, and (2) to investigate which vocabulary size test (VST) is better for that indicator, the written VST or the listening VST. 401 senior high school students, who were divided into four groups, from two schools participated in this experiment. The subjects took either Mochizuki's written or listening VSTs and general English proficiency tests (Kawai-juku mogi-shiken or TOEIC IP). Middle to high correlation coefficients between VSTs and general English proficiency tests were found. The correlation coefficients between the written VST and general English proficiency tests are higher than those between the listening VST and general English proficiency tests. The item analyses were also conducted for VSTs using FACETS. High IRT reliability coefficients were confirmed for both VSTs, and no major differences in the means of standard errors and item discrimination powers could be found between the written VST and the listening VST. The results show (1) Mochizuki's VST is suitable as a quick and rough indicator of general English proficiency and that (2) the written VST may be more efficient than the listening VST.
This paper firstly reviews what are regarded as the preferable ways to measure and analyse conceptual variables such as language proficiency, motivation, attitude, strategy use and so on. Since it is difficult to measure such concepts using only a few indexes, many observed variables are used for that purpose. The Exploratory Factor Analysis (EFA), which is a kind of multivariate analysis, can be used to assume latent variables (factors) behind observed ones. Here, attention should be paid as each solution offered by EFA is not absolute and is subject to variation according to the way the EFA is conducted. Secondly, this paper investigated five nation-wide journals (ARELE, JACET Bulletin, JALT Journal, JLTA Journal, Language Laboratory), which have been published in Japan and have broadly focused on language teaching and learning for these five years (from 1995 to 1999). All 15 EFAs conducted in them are analysed and described according to the method set out above. Finally, points to be aware of for further study are noted on the basis of the description of these five years' research tendency.
The purpose of this study is to examine the effects of error feedback on writing tasks of Japanese EFL learners. The fifty one junior high school students are divided into two groups, depending on the types of feedback; direct feedback such as undelining a word and providing a written tip and indirect feedback such as placing a question mark alongside a confusing phrase or structure. Those settings are adopted from Hendrickson(1980). The pre- and post- tests are analyzed by comparing the number of errors in each feedback type. We prepared mainly three research questions in this study: 1) Whether or not the group given direct feedback becomes more accuracy-oriented than the group given indirect feedback. 2) Whether or not direct feedback becomes more effective to the lower level students than the upper level students. 3) Whether or not indirect feedback becomes more effective to the upper level students than the lower level students. The results indicate that there was significant difference between direct feedback group and indirect feedback group in the revision. The efficacy of the error feedback also showed significant difference between the upper level students and the lower level students. These results lead to the conclusion that error feedback from their teachers has influence on students' own revising.
The present study examines the validity of Hamilton hypothesis in English relative clause (RC) production. Hamilton (1995) argues that English as a second language (ESL) learners may be sensitive to the difference of configuration rather than that of grammatical relation in English relativization. Twenty item RC test was administered to Japanese learners of English (N=100). The test consists of four parts: (1) argument/ object of preposition; (2) adjunct/ object of preposition; (3) argument/ indirect object; and (4) argument/ object of preposition. The difference between (1) and (2) is configuration (argument vs. adjunct) and (3) and (4) grammatical relation (indirect object vs. object of preposition). The results indicate that Japanese learners of English are more sensitive to configuration than grammatical relation. The criticism of recent treatment of Noun Phrase Accessibility Hierarchy and implications of the research findings for language testing and second language acquisition research methodology are also discussed.
The purpose of this research was to consider whether senior high school English tests should be conducted with instructions in English or in Japanese. The students of a senior high school who took a test with the instructions in English showed more positive attitude to English instructions than those who took the same test with the instructions in Japanese.. High school English teachers who prefer test instructions in Japanese are concerned about the complexity of some test items and English proficiency level of students.