Scholarlink

The study “Generative Artificial Intelligence Applied to Student Assessment Activities in Undergraduate Programs” investigates the effectiveness of using ChatGPT in academic evaluations. Conducted at an University in Brasília (UnB), the project analyzes the impact of using ChatGPT on student assessment results in two classes of the Information Technology program. Two groups of students received six mobile programming challenges to solve in three hours. The control group (Group A) had access to books and course notes on smartphone programming languages, but no internet access. Group B students had access to ChatGPT and were encouraged to use it to help solve programming challenges. The challenge was carried out in a computer lab under faculty supervision. Student teams were scored based on their performance in solving challenges or problems. Results show that students who used ChatGPT had a scoring advantage; however, inconsistencies and inaccuracies in the programming codes presented consequently affected overall performance.

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

Generative Artificial Intelligence Applied to Student Assessment Activities in Undergraduate Programs

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Marcelo Carboni Gomes,
  • Silvana Rosa de Jesus Ramos,
  • Maristela Holanda,
  • Li Weigang

摘要

The study “Generative Artificial Intelligence Applied to Student Assessment Activities in Undergraduate Programs” investigates the effectiveness of using ChatGPT in academic evaluations. Conducted at an University in Brasília (UnB), the project analyzes the impact of using ChatGPT on student assessment results in two classes of the Information Technology program. Two groups of students received six mobile programming challenges to solve in three hours. The control group (Group A) had access to books and course notes on smartphone programming languages, but no internet access. Group B students had access to ChatGPT and were encouraged to use it to help solve programming challenges. The challenge was carried out in a computer lab under faculty supervision. Student teams were scored based on their performance in solving challenges or problems. Results show that students who used ChatGPT had a scoring advantage; however, inconsistencies and inaccuracies in the programming codes presented consequently affected overall performance.