;Digital game-based learning has been prevalent and applied in many different subjects as tools for assisting learning. There are many studies use digital game-based learning strategy in English learning, especially in vocabulary, sentence and reading. However, there are few studies focus on oral practices in digital game-based learning environments. To this end, the current study integrated automatic speech recognition (ASR) into a digital game-based learning environment to develop an English oral system, aiming to provide leaner′s with an independent and flexible environment for English oral practice learning. Two types of oral practice feedback were provided in this system, including oral practice with feedback and without feedback. For the type of oral practice with feedback, it provides correct answers, scores, and recognized error inputs (vocabulary/sentence); for the type of oral practice without feedback, it only provides icons indicating correct or wrong responses. Therefore, immediate feedback was offered in learners’ oral practices, enabling for timely self-correction of English oral performance based on the feedback provided by the system. A total of 61 fifth and sixth graders students were recruited. This study aims to examine the effects of respective oral practice feedback types and prior knowledge on learning performance, game performance and speaking anxiety in digital game-based language learning. Learning performance includes oral, listening, vocabulary, sentence, and overall learning performance. Game performance contains learning type and gaming type. Learning type includes number of times of entering Challenge stage, completed times of sentence dragging practicing, number of times of correct in oral practicing, and number of times of listening; gaming type includes number of coin, number of completed task, and number of purchased items. Data analysis included descriptive analysis, independent samples t-test, ANCOVA, and Pearson product-moment correlation coefficient.
The results showed that significant differences were not found between the group with oral feedback and the group without oral feedback in learning performance, game performance and speaking anxiety in digital game-based language learning. Potential reasons for this might be that the oral practice feedback was not detailed enough which resulted in no significant differences were found. The results also showed that low prior knowledge learners significantly enhanced oral and overall learning performance than those high prior knowledge learners. Learners with low prior knowledge had significantly high speaking anxiety than high prior knowledge learners prior to the experiment. However, after the experiment, there was no significant difference in speaking anxiety between the two groups. Such results revealed that digital game-based English learning environment effectively reduced low prior knowledge learners’ speaking anxiety. In addition, the results demonstrated that the speaking anxiety and learning performance were negatively correlated at a significant level but the significance was not found in oral performance; there was a positive correlation between prior knowledge and learning performance at a significant level; the results also suggested that the negative correlation between speaking anxiety and prior knowledge.