This chapter examines the effects of computerized stroke order animation and visual chunking on character recognition and production. Through two experiments, the authors found that both computer-assisted presentation methods were effective, and their impact was comparable to or surpassed the traditional way of character learning through reading and writing. Specifically, animation was comparable to writing and more effective than reading in facilitating form recognition. Visual chunking produced better results in character production than writing when characters were presented in radical-based groups.
TopBackground
Due to the lack of systematic grapheme-phoneme correspondence in Chinese (e.g., DeFrancis, 1989), character learning is frequently identified as one of the greatest challenges for CFL learners (Everson, 1998). Learning how to write by rote repetition can be tedious (Allen, 2008), and it was suggested that character learning should be practiced individually by learners (Wang, 1998). Z. Zhang (1998) asserted that the computer is best suited for the time-consuming task of teaching students how to write, as the multimedia is not only instructional but also entertaining. The potential effect of multimedia learning is predicted by the Dual Coding Theory (DCT) (Clark & Paivio, 1991), which claims that the association of verbal and nonverbal representations enhances learning and memory retention. Nonverbal representations are nonlinguistic objects or events including shapes, sounds, and actions such as drawing lines. In the cognitive theory of multimedia learning, Mayer (1999, 2003) builds on the DCT and suggests that multimedia enables the combination of processing through the visual channel and verbal channel, which supports long-term memory. The impact of the DCT has been tested in previous character animation studies, although not always with positive findings (e.g., Jin, 2003; c.f., Zhu & Hong, 2005; Zhu et al., 2012). In the current project, we are interested in the roles of two particular visual encoding methods enabled by multimedia: stroke sequence animation, referred to as animation below, and character chunk presentation, or chunking. We selected these two encoding methods because animation and chunking are commonly-used features in character learning software (e.g., eStroke) and in on-line character learning platforms (Chen et al., 2011; Chen, Chien, & Chang, 2013). These two methods also respectively address two key characteristics of Chinese character composition, namely, stroke sequence and structural regularity.