关键词: 自我生成教学, 生成性学习, 社会临场感, 提取练习, 元认知加工

Abstract:

As a generative learning activity, learning by non-interactive teaching refers to learners play the role of teachers and teach what they have learned to others, and the activity is designed to help learners actively engage in knowledge building and improve their academic performance. For example, learners face a video camera to explain the learning material to imaginary, non-present peers in their minds (i.e., recording an instructional video). Given the vastly different ways in which learning by non-interactive teaching was implemented (e.g., video, audio, and text), the effectiveness of learning by non-interactive teaching in facilitating learning might be different. By summarizing the relevant studies, it was found that learning by non-interactive teaching in oral form with a tutor figure (e.g. video) was more effective in improving learner’s performance (d _{immediate comprehension} = 0.56, d _{delayed comprehension} = 0.63, d _{immediate transfer} = 0.35, and d _{delayed transfer} = 0.76) compared with simple learning activities such as restudy and retrieval practice, which was probably a better implementation. Learning by non-interactive teaching in oral form (e.g. audio only, d _{immediate comprehension} = 0.09 and d _{immediate transfer} = 0.02) or written form (e.g. text, d _{immediate comprehension} = -0.16, d _{delayed comprehension} = 0.39, d _{immediate transfer} = 0.08, and d _{delayed transfer} = 0.19) without a tutor figure had a smaller positive effect on learning outcomes. Learners with non-interactive teaching also experienced higher motivation (d = 0.44) and enjoyment (d = 0.76) and were willing to invest more mental effort (d = 0.47). The retrieval practice hypothesis and the generative learning hypothesis focused on different subcomponents of cognitive processing (e.g., retrieval, generation, or monitoring) to explain the positive effects of learning by non-interactive teaching on learning, respectively. The social presence hypothesis emphasized that social presence might facilitate whole cognitive processing and thus improved learning. Our results supported these three hypotheses to some extent. In addition, the cognitive theory of multimedia learning (CTML) may provide a supplementary explanation for differences in the effectiveness of different implementations of learning by non-interactive teaching. On the one hand, learning by non-interactive teaching (e.g., video) might successfully create teaching situation that stimulated a moderate sense of social presence and leaded learners to be more engaged and think more deeply about the material, i.e., increased their essential processing and generative processing, and thus facilitated learning. On the other hand, learning by non-interactive teaching (e.g., text) might distract learners from focusing too much on the typos, the standardization and rigorousness of written language, i.e., increased their extraneous processing. Due to the inherently high demands for processing capacity in generative activities, too much extraneous processing might cause learners' limited processing capacity being insufficient for adequate essential processing and generative processing, which in turn impaired learning. While learning by non-interactive teaching in the audio-only format might neither successfully facilitate learning with essential processing and generative processing because of the weaker teaching situation created, nor hinder learning with extraneous processing because of the automated spoken language. Research is needed to test and integrate theories, identify boundary conditions, and enhance the effectiveness of learning by non-interactive teaching in the future.

Key words: learning by non-interactive teaching, generative learning, social presence, retrieval practice, metacognitive processing