USING NATURALREADER: A FREE TEXT-TO-SPEECH ONLINE WITH AI-POWERED VOICES IN TEACHING LISTENING TOEFL

Abstract

This research describes the use of NaturalReader in teaching listening TOEFL. This research is descriptive qualitative research. The analysis shows that NaturalReader is an AI-powered free text-to-speech service online which helps teachers/lecturers to teach English. In teaching English, especially listening to TOEFL, an English lecturer in ITB AAS Indonesia uses face-to-face learning combined with Zoom and Live YouTube streaming. The lecturer uses NaturalReader Text-to-Speech (TTS) in teaching listening TOEFL including Part A (short dialogue), Part B (long dialogue), and Part C (monologue). The lecturer opens the NaturalReader application, then clicks the + icon to input the words, phrases, or sentences that she/he wants to listen to the audio with their pronunciation. NaturalReader offers female and male voices and variation accents for both free and paid users. This software is capable of converting text into sound in multiple languages, including English with diverse accents, such as British or American. In teaching listening TOEFL, pronunciation especially accent cannot be separated from listening. In listening, the presence of several accents, the speakers’ speed, and the speaker’s intonation from native speakers on listening tests may threaten test fairness. By using the NaturalReader app, the students can learn to listen to how the word/phrase/sentence is pronounced differently with different accents and voices as they want. It means that the students study listening also pronunciation. When students understand correct pronunciation, students will understand English listening easier because they will learn to identify the sounds directly from native speakers.