Answers are grounded in this paper and cite their sources.
Purpose This study aimed to construct a large-scale developmental database of traditional Chinese characters, quantify the relative contributions of multilevel linguistic features to character difficulty, and examine selected feature effects across grade groups.Method A naming task involving 3,194 commonly used characters was administered to 851 Mandarin-speaking children in Grades 1–6 (52% female), including 260 lower-, 331 middle-, and 260 upper-grade students. Using lmg analysis and a generalized linear mixed model, we examined the contributions of 27 features across five categories to IRT-estimated character difficulty.Results The model accounted for 74.6% of the variance in character difficulty. Lexical-semantic features were the strongest predictors, followed by frequency, sublexical mapping features, and visual complexity. Grades 1–4 showed stronger effects of the number of words formed for high-frequency characters, whereas this pattern reversed in Grades 5–6; semantic-radical transparency facilitated lower-frequency characters in Grades 3–4; phonetic-radical sensitivity showed stronger effects for low-frequency characters from Grades 1–2 onward.Conclusion .Chinese orthographic learning is supported early by high-frequency exposure, with phonetic cues remaining influential across grade groups; higher-grade students show greater use of structural and morphological information, from semantic-radical cues to broader word-internal contexts. These findings may inform Chinese literacy instruction and facilitate quality education.
No takes yet. Share an insight, caveat, or question.
Tseng et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: