A study of writing development: from the perspective of orthographic units of different grain sizes

The properties of phonetic radical and semantic radical of a Chinese phonetic-semantic compound character would produce different effects on the process of writing output (Meng et al., 2000; Meng, Shu, Zhou, 2000; Luen et al., 2001). The regular relationship between the pronunciations of phonetic radical and that of the Chinese character shortens the response time required for the process. Studies have been conducted to investigate the effects of phonetic radicals and configuration of Chinese characters on writing output, focusing on specific error made. However, the relations between different writing errors types and the children’s writing performance across grades have yet been studied. The present study attempts to categorize children’s writing error types from the perspectives of orthographic units to investigate the effects of various error types on the writing of Chinese characters and its developmental change.
A total of 90 subjects of Grade2 and 3 students, 30 from each of three primary schools randomly selected in Shenzhen (a city in the South of China) to participated in the a writing-to-dictation task. Based on the database established on Chinese primary textbooks (People’s Education Press (PEP) version), 24 phonograms were selected as target characters, evenly distributed across high, mid and low frequencies. Each target stimulus character was presented as one of the component characters of a two-character word in order to minimize the effect of homophone. For example, 马，白马的马 (ma, baima de ma; horse, horse in white-horse).
Statistic analyses have been conducted. The following significant results were found. First, the percentage of semantic-radical errors decreases but that of phonetic-radical error increases from Grade2 to 3. Second, homophone substitution is the most frequent error, and its rate reduces from Grades2 to 3. Third, deletion is the major errors type. Younger children tend to use the whole character when they are learning to master sub-character structure and the rate of logograheme errors increases from Grade2 to 3.
All the error patterns will be further analyzed from the developmental perspective. Their theoretical implications and its implications on the experiment tasks for writing investigation will be discussed. Copyright © 2014 15th International Clinical Phonetics and Linguistics Association Conference.