Scholarlink

<p>Large language models (LLMs) have a wide range of current and potential applications across the higher education sector. Most research to date has focused on the benefits and challenges of LLMs and other generative artificial intelligence tools in teaching, learning, and research. However, how these technologies might influence prospective students’ choices of where to study by shaping the way study opportunities are presented has yet to be explored. This study investigates responses generated by four LLMs regarding international study opportunities. The responses are analysed using qualitative and quantitative content analysis. The findings reveal framing mechanisms through inclusion and exclusion, ordering, and qualitative propositions. Variations across prompts, tools, and countries are discussed, along with methodological considerations for working with LLM-generated data. The article concludes by reflecting on the potential implications of the growing reliance on LLMs for sourcing information related to higher education choices.</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

Framing of higher education opportunities by large language models

  • 期刊论文

  • 发表日期：

    2026年4月25日

查看全文
  • Pii-Tuulia Nikula

摘要

Large language models (LLMs) have a wide range of current and potential applications across the higher education sector. Most research to date has focused on the benefits and challenges of LLMs and other generative artificial intelligence tools in teaching, learning, and research. However, how these technologies might influence prospective students’ choices of where to study by shaping the way study opportunities are presented has yet to be explored. This study investigates responses generated by four LLMs regarding international study opportunities. The responses are analysed using qualitative and quantitative content analysis. The findings reveal framing mechanisms through inclusion and exclusion, ordering, and qualitative propositions. Variations across prompts, tools, and countries are discussed, along with methodological considerations for working with LLM-generated data. The article concludes by reflecting on the potential implications of the growing reliance on LLMs for sourcing information related to higher education choices.