Academic Text Logic and Language Quality Test
This test evaluates how well different models revise academic text. It focuses on logical completeness, theoretical reasoning, linguistic precision, coherence, and readability. Models must preserve the original meaning and citations while correcting weak reasoning, conceptual ambiguity, and redundant expression.
Prompt
中文版： 请将下段改写为120–150词的学术英语。要求：避免用“研究较少”直接倒推出研究价值；说明SDT-based AI discourse strategies为何会影响大学生的student engagement；解释perceived autonomy support与student engagement的区别及其链式作用；保留原引用，不新增文献；语言严谨、连贯、易读，避免长句和夸大结论。只输出改写后的段落。 原文： AI can improve learning outcomes, but few studies have examined AI discourse strategies (Wang & Xue, 2024; Li & Chiu, 2025). Therefore, this study compares SDT-based AI discourse strategies with answer-oriented discourse. Perceived autonomy support and student engagement may explain their effects on conceptual understanding. English version: Rewrite the paragraph below in 120–150 words of academic English. Avoid using limited prior research as the main justification. Explain why SDT-based AI discourse strategies may affect university students’ student engagement. Clarify the distinction and sequential relationship between perceived autonomy support and student engagement. Retain all citations and add no new sources. Improve logical completeness, rigor, cohesion, and readability. Avoid long sentences and exaggerated claims. Output only the revised paragraph. Original paragraph: AI can improve learning outcomes, but few studies have examined AI discourse strategies (Wang & Xue, 2024; Li & Chiu, 2025). Therefore, this study compares SDT-based AI discourse strategies with answer-oriented discourse. Perceived autonomy support and student engagement may explain their effects on conceptual understanding.
