THE IMPLICATIONS OF ITS RESEARCH ON CALL SOFTWARE

Abstract

Current research into Computer Assisted Language Learning (CALL) software demonstrates a existence of cognitive, constructivist and behaviorist learning styles. Within these systems exists a tendency towards a behaviorist learning style of content creation and user interaction and evaluation. This tendency often leads to a structuring of tasks in a drill and practice format, and the interaction between the user and the system involves the administration of rewards and punishments for the solidification of user knowledge. However, research into Intelligent Tutoring System (ITS) reveals that effective ITSs utilize domain, learner and pedagogical knowledge in their interaction with the user. The implementations of these knowledge forms in ITSs are greater user retention of content. Subsequently, the implementation of learner and pedagogical knowledge provides better individual-focused instruction and a more accurate system model of user knowledge. Furthermore, the evaluation of ITSs based on the existence of these knowledge forms provides consistency for developers and content creators in the implementation of updates and innovations in current ITSs. In contrast, the evaluation and construction of current Computer Assisted Language Learning (CALL) systems frequently center more around tasks creation in the system and less on the interaction of the system with the user. A pure ITS system is costly and time consuming to develop, thus this paper argues for a restructuring of CALL systems with the distinction between these knowledge forms within ITSs. Domain, learner and pedagogical knowledge are often perceived differently from a linguistic perspective. This paper explores the definitions of these knowledge forms in relation to the linguistic concepts of native language (L1) and target language (L2) interdependency and influence on system error analysis. In this paper we investigates some of the current CALL software including Duolingo, Transparent Languages, Rosetta Stone and Memrise. In our analysis we identify domain, learner and pedagogical knowledge in these systems. We found that domain knowledge was easily identifiable, but the manifestation of learner and pedagogical knowledge was rudimentary at best. Moreover, this paper examines task construction with learner and pedagogical knowledge and gives recommendations for improvements within these CALL systems. These improvements are related to system feedback, content construction, content sequentialization and user interface− including the implementation of semantic tagging, L1 and L2 error analysis within domain knowledge and learning path creation based on linguistic domain knowledge. In conclusion, the evaluation of software in relation to the employment of domain, learner and pedagogical knowledge can lead to better consistency in CALL system creation and evaluation. Furthermore, the implementation of the stated recommendations would further increase user retention of the content in these systems.