Optimizing AI ChatGPT: Prompt Engineering in the Learning Assessment Process for High School Students

Along with the development of artificial intelligence (AI) technology, the use of AI in education is increasingly widespread, including in the learning assessment process. This study aims to explore the optimization of the use of ChatGPT, a text-based AI model, in creating learning assessments for high school students. ChatGPT offers efficiency in designing various questions, ranging from multiple choice to short essays, which can be used as a tool for teachers to speed up the assessment creation process. With its ability to generate questions quickly, ChatGPT allows teachers to focus more on other aspects of learning, such as monitoring student progress and improving teaching methods. In addition, this AI can also help teachers create different variations of questions for students with varying abilities, ensuring that students' individual needs can be better accommodated. However, there are limitations in its application, especially in producing questions that test students' in-depth understanding, critical thinking skills, and analytical skills. This study uses a qualitative approach by analyzing various studies and literature related to the use of AI in learning and testing several assessments produced by ChatGPT. The study results show that ChatGPT can improve teachers' time efficiency and flexibility in creating questions, especially for questions with a medium level of difficulty. However, more complex assessments require manual adjustments by teachers to ensure the questions' relevance and depth align with the curriculum. The combination of AI and teacher involvement in evaluation can improve the overall quality of learning.