3078-3348
CrossRef
OpenAIRE
Index Copenius
Scilit
Zenodo
Google Scholar
J-Gate

李蓉
江汉大学,中国
摘要
针对 HSK 书写部分合格率偏低及构念测量偏窄等问题,本研究基于构念效度理论与交际语言能力模型,从构念代表性、任务内在关联性及难度梯度适配性三个维度,系统对比 HSK2.0 与 HSK3.0,并辅以实证三角互证。研究发现: HSK2.0 的测查重心较多倾斜于低阶形式与认知技能,构念覆盖相对偏窄; HSK3.0 则实现了语篇、语用与策略构念的全维度覆盖,任务设计向高阶认知与深层功能关联跃迁,难度梯度更契合能力发展。同时,实证调查揭示 HSK3.0在实测中面临高阶计分信度存疑、四至五级字数跨度易致认知超载及学术写作缺位等潜在挑战。本文据此提出警惕应试反拨、平缓认知跨度及拓展测评边界的优化建议。
关键词
HSK2.0, HSK3.0, 书写部分, 构念效度, 交际语言能力
A Comparative Study of Task Characteristics and Potential Constructs in the Writing Sections of HSK 2.0 and HSK 3.0
Rong Li
Jianghan University, China
Abstract
Addressing the low pass rate and narrow construct coverage in the HSK writing section, this study systematically compares HSK 2.0 and 3.0 based on construct validity theory and the Communicative Language Ability model. The comparison focuses on three dimensions: construct representation, internal task relevance, and difficulty gradient alignment, complemented by empirical triangulation. Results show that HSK 2.0 tends to focus more heavily on lower-order forms and cognitive skills, with relatively narrow construct coverage. In contrast, HSK 3.0 achieves full-dimensional coverage encompassing discourse, pragmatics, and strategy constructs. Its task design shifts toward higher-order cognitive skills and deep functional relevance, featuring a difficulty gradient more consistent with capability development. Meanwhile, empirical findings reveal potential challenges for HSK 3.0: uncertain scoring reliability for high-level constructs, potential cognitive overload from the steep word-count increase between Levels 4 and 5, and the absence of academic writing. Consequently, this study proposes optimization suggestions: preventing negative washback, smoothing the cognitive transition, and expanding assessment boundaries.
Keywords
HSK 2.0, HSK 3.0, writing section, construct validity, communicative language ability