A Listenability Measuring Method for an Adaptive Computer-assisted Language Learningand Teaching System

A Listenability Measuring Method for an Adaptive Computer-assisted Language Learning and Teaching System Katsunori Kotani Shota Ueda Kansai Gaidai University / Osaka, Japan Ryukoku University / Shiga, Japan kkotani@kansaigaidai.ac.jp Takehiko Yoshimi Hiroaki Nanjo Ryukoku University / Shiga, Japan Ryukoku University / Shiga, Japan Abstract In teaching and learning of English as a foreign language, the Internet serves as a source of authentic listening material, enabling learners to practice English in real contexts. An adaptive computer-assisted language learning and teaching system can pick up news clips as authentic materials from the Internet according to learner listening proficiency if it is equipped with a listenability measuring method that takes into both linguistic features of a news clip and the listening proficiency. Therefore, we developed a method for measuring listening proficiency-based listenability. With our method, listenability is measured through multiple regression analysis using both learner and linguistic features as independent variables. Learner features account for learner listening proficiency, and linguistic features explain lexical, syntactic, and phonological complexities of sentences. A cross validation test showed that listenability measured with our method exhibited higher correlation (r = 0.57) than listenability measured with other methods using either learner features (r = 0.43) or other linguistic features (r = 0.32, r = 0.36). A comparison of our method with other methods showed a statistically significant difference (p < 0.003 after Bonferroni correction). These results suggest the effectiveness of learner and linguistic features for measuring listening proficiency-based listenability. 1 Introduction Listening practice using authentic materials is necessary for learners of English as a foreign language (EFL) who have little or no chance to use English in their daily life because these materials let them immerse themselves in real-life settings. Since authentic materials are not usually constrained by the ease of listening comprehension or listenability (Chall & Dial 1948), teachers have to select materials according to learner listening proficiency; otherwise, too difficult or easy materials reduce the learning effect or spoil learner motivation (Hubbard 2004, Petrides 2006). An adaptive computer-assisted language learning and teaching system can pick up news clips as authentic materials from the Internet according to the listening proficiency if it is equipped with a listenability measuring method that takes into both linguistic features of a news clip and the listening proficiency. Therefore, we propose an automatic method that statistically measures listenability for EFL learners. This method is useful for learning and teaching English by showing listenability levels of authentic listening materials such as news clips. It also helps to create a computer-based self- learning environment because EFL learners can select appropriate materials with this method. PACLIC 28 387 Copyright 2014 by Katsunori Kotani, Shota Ueda, Takehiko Yoshimi, and Hiroaki Nanjo 28th Pacific Asia Conference on Language, Information and Computation pages 387–394 Although listenability of authentic listening materials can be measured with readability measuring methods using lexical, syntactic, and discourse features (Flesch 1950, Graesser et al. (2004), and Shen et al 2013), our listenability measuring method uses phonological features as well as lexical and syntactic features. Phonological …

Content truncated to 500 words