Faculty of Education - Theses
Permanent URI for this collection
Search Results
1 - 2 of 2
-
ItemConversational or instructional discourse : the opportunity for production of language by high school age students in English conversation classes in Japan( 1999)Some recent views on English language teaching suggest that one of the goals should be to use natural or genuine conversation in the classroom in preference to more traditional classroom communication. This study set out to find if naturalistic conversation used by the teacher was able to promote more active production of language by learners in a classroom of Japanese high school learners of English in Japan than in comparison to instructional classroom discourse. The length of responses to both referential and display questions were measured in terms of the mean length of utterance of the responses of the students to teacher questions in order to gauge the relationship between the type of language used and the potential for production of language. It was assumed that referential questions were characteristic of more naturalistic language and that display were characteristic of more 'traditional' classroom language. Opportunities for the negotiation of meaning were also sought by measuring the frequency of occurrence of echoic questions, associated with negotiation of meaning for both conversational discourse and for instructional discourse. The results for this particular study initially showed that in terms of the length of response and opportunities for the negotiation of meaning, both display questions and referential questions produced longer responses in instructional discourse than in comparison to conversational discourse. This suggests that in the case of the students studied, traditional classroom language or instructional discourse showed more potential for production of language than did naturalistic conversation, regardless of the question type. The results also showed that display questions promoted longer segments of classroom talk and negotiation of meaning than did referential questions suggesting that classroom conversation, as characterised by display questions, holds more potential for the negotiation of meaning than natural conversation. The results also showed, however, that there may be difficulties in producing conversational discourse and instructional discourse in a lesson in this context suggesting that the research questions and design of this particular study may be in need of reconceptualisation.
-
ItemAssesing the spoken skills of Japanese secondary students( 1999)This thesis has evaluated a test of spoken English for secondary school students in Japan in relation to its validity, reliability and practicality. Although the Japanese Ministry of Education has put a strong emphasis on the improvement of communicative competence, little research and development work has been carried out on speaking tests in particular those tests administered to secondary school students in Japan. Recent research on Item Response Theory in language testing has primarily investigated test validity and reliability for adults and immigrants in a second language context (e.g. Brown 1995, McNamara 1990 a, b, O'Loughlin 1997). The present study has applied Item Response Theory to analyses of a speaking test where English is learned as a foreign language in the classroom. The various types of data used in this study were gathered from a test administered to 109 students in Tokyo in December 1998. The test was conducted by five interviewers, and seven raters rated each student's performance independently. Analyses of candidate ability, item difficulty and rater severity were carried out with the Item Response Modelling Program 'ConQuest' (Wu, Adams and Wilson 1998). Questionnaire responses from test takers and interviews were also analysed. Overall, the results showed that the test possessed high validity, reliability and practicality in the context for which it was developed. In particular, item analysis and feedback were found of crucial value in providing evidence of validity. It was also found that there was variability in the relationship between individual items and raters in terms of rater severity. These findings have demonstrated the need to conduct further research on how task difficulty affects student performance and the relationship between the washback effect and student speaking ability in Japan.