Effects of Robot-Based Multiple Low-Stakes Assessments on Students' Oral Presentation Performance, Collective Efficacy, and Learning Attitude

Low-stakes assessment has gained attention in recent years due to its link to enhancing learning effects and its essential role in learning evaluation. Unlike high-stakes assessments, low-stakes assessments have little or no consequences for learners' academic performance, and are designed to support the feedback-oriented learning process. Providing multiple low-stakes assessments to students yields significantly greater long-term retention of knowledge and skills. However, learners may not give their best efforts when taking low-stakes assessments, which could lead to poorer learning outcomes. Using emerging technologies such as social robots in the learning environment could foster interactive learning, engagement, and motivation for learning assessments. Therefore, integrating low-stakes assessments and robots might encourage students to exert greater effort while performing learning tasks. This study aimed to discover the impacts of robot-based multiple low-stakes assessments on students' oral presentation performance, collective efficacy, and learning attitude. A quasi-experiment was conducted in two sixth-grade classes of elementary students. The Robot-based Multiple Low-Stakes Assessment (Robot-MLSA) was randomly assigned to one class, while the Computer-based Multiple Low-Stakes Assessment (C-MLSA) was assigned to another class. The findings showed that the Robot-MLSA could enhance students' oral presentation performance, support their collective efficacy, and improve their learning attitude toward robots. Furthermore, an in-depth discussion of students' learning perceptions and experience is provided to explore the effectiveness of the Robot-MLSA.