Spoken grammar practice and feedback in an ASR-based CALL system

Abstract

Speaking practice is important for learners of a second language. Computer assisted language learning (CALL) systems can provide attractive opportunities for speaking practice when combined with automatic speech recognition (ASR) technology. In this paper, we present a CALL system that offers spoken practice of word order, an important aspect of Dutch grammar. The system uses ASR technology to process the learner's responses and to detect errors so that immediate corrective feedback (CF) can be provided on learner errors. We evaluate the system as a learning environment by analyzing proficiency gains in pre- and post-tests, the logs of the practice sessions, and the learners’ appreciation of the system. In this paper, we present two learning conditions: (1) the learners received oral practice and immediate CF on spoken performance and (2) learners received oral practice and NO CF on spoken performance. We found that our system was successful in providing L2 speaking practice. Results show that both groups improve their proficiency on the target feature as a result of treatment. Between the groups there is no significant difference in learning, but the groups proceeded differently through the sessions, and the learners in the group that received automatic CF evaluated the system more positively than the NO CF group. We discuss the performance of the system as an environment for language learning and the obtained proficiency test results, and relate them to current views on second language acquisition.