8th International Conference on Spoken Language Processing

Jeju Island, Korea
October 4-8, 2004

Theory and Data in Spoken Language Assessment

Jared Bernstein (1), Isabella Barbier (2), Elizabeth Rosenfeld (2), John H.A.L. de Jong (3)

(1) Stanford University, USA
(2) Ordinate Corporation, USA
(3) Velp, Netherlands

Spoken language performance depends on both psycholinguistic processing in the individual and on communicative uses of language in dialogue. A social-communication view of language emphasizes that the spoken form of a language is used in social settings to accomplish explicit or implicit tasks. In testing, the communicative tradition is associated with oral proficiency interviews. A psycholinguistic view of spoken language emphasizes that the development of component skills forms the basis of real-time performance in a language. The paper reviews the development and validation of psycholinguistic testing task types (e.g. elicited imitation) that measure performance in spoken language based on empirical models of processes internal to the individual speaker-listener. Data from a recent set of experiments indicates that the two theoretical approaches to testing do not produce different patterns of proficiency scores for populations of second language speakers. This data obviates the need to posit a communicative basis for language test design.

Full Paper

Bibliographic reference.  Bernstein, Jared / Barbier, Isabella / Rosenfeld, Elizabeth / Jong, John H.A.L. de (2004): "Theory and data in spoken language assessment", In INTERSPEECH-2004, 1685-1688.