As human beings, learners bring along their emotions also to their English as a Foreign Language (EFL) classrooms. This makes emotions an inseparable part of learners’ lives, which plays an important role in their language achievement and performance. This especially applies in the Indonesian context, where EFL learners appear to experience anxiety when speaking English. Additionally, vocabulary and pronunciation learning and teaching have frequently been neglected in EFL classrooms. These two pivotal elements affect learners’ overall speaking competence.
We investigated the effects of two Automatic Speech Recognition (ASR)-based language learning systems, I Love Indonesia (ILI) and NovoLearning (NOVO), on students’ Foreign Language Speaking Anxiety (FLSA), Foreign Language Enjoyment (FLE), receptive and productive vocabulary knowledge, and word-level and sentence-level pronunciation.
The research reported on in this thesis confirms our expectations about the positive effects of practicing with ASR-based language learning systems on the four outcome variables investigated. Our systems successfully reduced learners’ FLSA and enhanced their FLE. We also discovered that our systems significantly improved learners’ vocabulary knowledge and pronunciation. Interestingly, the different types of feedback provided by ILI and NOVO had significant, distinct effects on all three pronunciation measures – phonetic distance, accentedness, and comprehensibility. NOVO outperformed ILI, but ILI remained effective for pronunciation learning.
In the Indonesian context, relatively few studies have discussed the potential of ASR-based practice for secondary school students’ affect and speaking skills. In this thesis, we have bridged this gap by providing new evidence and perspectives that may shed more light on how ASR-based language learning practice positively impacts EFL learning and teaching (speaking) in Indonesia.