published_or_final_version / Linguistics / Master / Master of Arts in Applied Linguistics

No description available.

The relationship between language proficiency and academic success in university studies is of major concern in educational institutions throughout the Western world. The particular focus of this study is the situation in the Republic of South Africa. This problem is of critical importance at this stage in the history of South Africa when universities have publicly stated their commitment to admit any students with merit or potential to succeed at university. In order for students to succeed at university they need to be communicatively competent in the language which is the medium of instruction. It has been assumed that this ability can be assessed by means of a formal test and it is this issue on which this dissertation focuses in order to establish how reliable such tests are as predictors of academic performance. The empirical research covers a six-year period from 1982 to 1987 and investigates two tests. One is a particular language test which was designed specifically for the selection of students for courses of academic study of English at university. The other is the senior certificate examination which provides the statutory admission requirement for university entrance in South Africa. An extensive review of relevant studies both within South Africa and overseas has been undertaken. In addition an unstructured questionnaire was sent to English departments throughout South Africa in order to establish the current practice with regard to the selection of students for first-year courses. The conclusion seems to be that in the context of a homogeneous population language proficiency as measured on a formal test is predictive of academic success in first-year courses in English. In heterogeneous student populations, like that of South Africa where the majority of prospective students may be described as being "disadvantaged", however, this is not the case. Academic success cannot be predicted with any degree of confidence on the basis of language proficiency. Extreme caution is necessary in the implementation of any language test for the selection of students for academic study at university in the present changing nature of university student populations in South Africa. / Thesis (M.Ed.)-University of Natal, Durban, 1988.

This study investigated the relationship between pragmatic ability and proficiency in 16 Japanese learners of English. Two measures of pragmatic ability were developed: a multiple-choice questionnaire designed to probe pragmatic awareness of various speech acts and a set of oral role-plays designed to elicit two requests, two apologies and one refusal. These measures were also administered to eight native speakers in order to establish a scoring system for the pragmatic awareness test and target norms for the role-plays. A background questionnaire was given to all participants while a proficiency test, Combined English Language Skills Assessment in a Reading Context, or CELSA was administered only to the Japanese participants. / The findings in this study demonstrated that there was not a strong relationship between proficiency and pragmatic ability, nor was there one between pragmatic awareness and production. The measures were moderately correlated but they also exhibited a great deal of variation from learner to learner. / Certain linguistic abilities were observed to be valuable regarding pragmatic ability, such as being able to make conventionally indirect requests. Learners were more direct in their speech acts than native speakers. They also used fewer and less varied strategies and lexical modification, with the exception of the politeness marker, please, which they over-used. / There was evidence both in terms of the pragmatic awareness measure and in the analysis of the production that over-directness decreased with increasing proficiency. The two assessment instruments produced different kinds of errors at different levels of proficiency, pointing to the possibility that they were tapping the abilities associated with pragmatic awareness and production at different stages of development. / This research has implications for both testing and teaching pragmatic ability.

Mode of access: World Wide Web. / Thesis (Ph. D.)--University of Hawaii at Manoa, 2005. / Includes bibliographical references (leaves 145-151). / Electronic reproduction. / Also available by subscription via World Wide Web / xi, 151 leaves, bound ill. 29 cm

Cloze procedure involves the skills of thinking, understanding , reading and writing based on the learners underlying knowledge of reading comprehension and writing subskills such as grammar and sentence construction. It is regarded as a very efficient test for measuring students integrative competence in English, and has been used in most of the important English tests in China. It is also used as a teaching instrument to help students to improve their reading competence. However, a majority of students perform poorly in cloze tests and they regard it as the most difficult and most unpopular part of the English test. The aim of this study was to find out the problem that Chinese students have with cloze tests and to determine whether they are associated with the inefficient use of reading strategies.

This thesis investigates the automatic assessment of essays written by Japanese low level learners of English as a second language. A number of essay features are investigated for their ability to predict human assessments of quality. These features include unique lexical signatures (Meara. Jacobs & Rodgers, 2002), distinctiveness, essay length, various measures of lexical diversity, mean sentence length and some properties of word distributions. Findings suggest that no one feature is sufficient to account for essay quality but essay length is a strong predictor for low level learners in time constrained tasks. Combinations of several features are much more powerful in predicting quality than single features. Some simple systems incorporating some of these features are also considered. One is a two-dimensional 'quantity/content' model based on essay length and lexical diversity. Various measures of lexical diversity are used for the content dimension. Another system considered is a clustering algorithm based on various lexical features. A third system is a Bayesian algorithm which classifies essays according to semantic content. Finally, an alternative process based on capture-recapture analysis is also considered for special cases of assessment. One interesting finding is that although many essay features only have moderate associations with quality, extreme values at both ends of the scale are often very reliable indicators of high quality' or poor quality essays. These easily identifiable high quality or low quality essays can act as training samples for classification algorithms such as Bayesian classifiers. The clustering algorithm used in this study correlated particularly strongly with human essay ratings. This suggests that multivariate statistical methods may help realise more accurate essay prediction.

No description available.

published_or_final_version / Education / Master / Master of Education

published_or_final_version / Linguistics / Master / Master of Arts in Applied Linguistics