Summary: This project will develop elicited imitation, sometimes referred to as "sentence repetition," as a standardized measurement instrument of foreign language linguistic proficiency. A completed test will be developed for Vietnamese and made available in a ready-to-use computerized format, along with guidelines on administration, scoring, and interpretation. In addition, a handbook on elicited imitation will be developed containing guidelines for the development of comparable measures in any language.
This project proposes to develop direct measures of second language performance by the use of elicited imitation (EI), sometimes referred to as "sentence repetition." Elicited imitation has been used to a considerable extent in the assessment of first language disorders in children, in first language acquisition research, and in a large number of research studies in second language acquisition. To date, it has been used only in limited fashion in language assessment batteries, though some uses in English and Japanese as a second language, and bilingual Urdu and Pashto speakers of Hindko, for instance, have aimed at broad proficiency assessments. Its current incorporation in telephone-based data collection for assessments of English as a second language performance has shown its viability as a standardized measurement instrument. Research has shown not only that EI attains high reliability and validity measured against alternative assessments, but that some data can be used for diagnostic procedures, and, like cloze testing, it lends itself to relatively easy-to-construct items and batteries.
The project proposes to do the following:
Elicited imitation tests will be developed for one language and made available in a ready-to-use computerized format, along with guidelines on administration, scoring, and interpretation. In addition, a handbook on elicited imitation will be developed containing guidelines for the development of comparable measures in any given language. The demonstration language for test development will be chosen from one of the following languages: Indonesian, Vietnamese, Filipino (Tagalog), or one of the languages of East Asia. This is a two year project and will not begin until 2004, Year 3 of the grant cycle. Selection of the demonstration language depends on the availability of graduate students who are competent in the construction of computerized speech and data collection methods and are also native speakers of the language selected, which cannot be predicted with precision this far in advance.