Scholarlink

<p>Since its launch in November 2022, Chat Generative Pretrained Transformer (ChatGPT) has rapidly become an attractive artificial intelligence (AI) tool for students around the globe. This chatbot is becoming more and more popular because of its ability to generate science-related texts practically indistinguishable from human-produced texts. While a major shortcoming of texts generated by ChatGPT is that these can contain false and/or inaccurate scientific information, very little is known about how to promote critical reading of science-related texts produced by this AI tool. We aimed at showing that the construction of scientific argument maps—visual representation of scientific argument structure—can be used to foster critical reading of these texts. The data were drawn from the argument diagrams constructed by 44 undergraduates (27 females and 17 males, 17–23&#xa0;years old) during an introductory science course and audio recordings of student discussions as part of the process of co-creation of argument maps. The findings suggest that argument mapping effectively supported participants’ critical reading of ChatGPT‐generated texts while engaged them in the construction of arguments, the anticipation of counterarguments, and the production of rebuttals. This study contributes to the literature on scientific AI literacy—the combination of scientific literacy with AI literacy —by providing insights into how to engage students in the critical reading of science-related texts produced by generative AI technology.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Critically Reading Science-Related Texts Produced by ChatGPT

  • 期刊论文

  • 发表日期：

    2025年5月23日

查看全文
  • Pablo Antonio Archila,
  • Brigithe Tatiana Ortiz,
  • Anne-Marie Truscott de Mejía,
  • Jorge Molina

摘要

Since its launch in November 2022, Chat Generative Pretrained Transformer (ChatGPT) has rapidly become an attractive artificial intelligence (AI) tool for students around the globe. This chatbot is becoming more and more popular because of its ability to generate science-related texts practically indistinguishable from human-produced texts. While a major shortcoming of texts generated by ChatGPT is that these can contain false and/or inaccurate scientific information, very little is known about how to promote critical reading of science-related texts produced by this AI tool. We aimed at showing that the construction of scientific argument maps—visual representation of scientific argument structure—can be used to foster critical reading of these texts. The data were drawn from the argument diagrams constructed by 44 undergraduates (27 females and 17 males, 17–23 years old) during an introductory science course and audio recordings of student discussions as part of the process of co-creation of argument maps. The findings suggest that argument mapping effectively supported participants’ critical reading of ChatGPT‐generated texts while engaged them in the construction of arguments, the anticipation of counterarguments, and the production of rebuttals. This study contributes to the literature on scientific AI literacy—the combination of scientific literacy with AI literacy —by providing insights into how to engage students in the critical reading of science-related texts produced by generative AI technology.