Artificial intelligence and its approach to people’s daily lives have been rapid and sustained, from the use of virtual assistants that use natural language processing on commonly used devices such as smartphones to more complex artificial intelligence based on deep learning techniques, such as Generative Pre-trained Transformer (ChatGPT). In education, there is a lot of expectation about using this artificial intelligence correctly, taking advantage of its benefits, and minimizing the risks that its use can entail. Since many users with different interests can use this AI, it is interesting to identify perceived usability. Usability is an essential aspect of user experience; it considers the extent to which a product, system, or service can be used to complete specific tasks, considering the context of use. Usability in a tool used for education is critical since it will allow users to achieve their objectives, thus facilitating learning. In this work, we applied a survey to evaluate undergraduate students’ perceptions of ChatGPT usability and identify whether, from the point of view of usability, it is a good tool to use in education. The survey was applied to undergraduate students from two Chilean universities that belonged to two programs related to business and computer science. The survey was conducted online during October 2024. The questionnaire was divided into two sections. The first section was a characterization, and the second section contained the System Usability Scale questionnaire (SUS), widely used to evaluate the perceived general usability of different system interfaces. This work has a quantitative approach with an exploratory scope. The characterization information will be descriptively analyzed in percentages and frequencies. Data from the second part of the questionnaire were analyzed based on the usability categories defined and established in SUS. We had 117 participants, and the perceived usability score was 69.7, which is favorable for ChatGPT. In the analysis by gender and program on the perceived usability of ChatGPT (using the T-student test), we found that men have a better perception of usability than women; in turn, we found that students in computer science have a better perception of usability than those in business. This indicates that it may be a good tool to use in education. In future work, we want to expand the sample and focus the results on the positive use of this artificial intelligence tool in teaching and learning.

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

ChatGPT Usability. An Exploratory Study in Higher Education

  • Jenny Morales,
  • Fabián Silva-Aravena,
  • Héctor Cornide-Reyes,
  • Guisselle Muñoz

摘要

Artificial intelligence and its approach to people’s daily lives have been rapid and sustained, from the use of virtual assistants that use natural language processing on commonly used devices such as smartphones to more complex artificial intelligence based on deep learning techniques, such as Generative Pre-trained Transformer (ChatGPT). In education, there is a lot of expectation about using this artificial intelligence correctly, taking advantage of its benefits, and minimizing the risks that its use can entail. Since many users with different interests can use this AI, it is interesting to identify perceived usability. Usability is an essential aspect of user experience; it considers the extent to which a product, system, or service can be used to complete specific tasks, considering the context of use. Usability in a tool used for education is critical since it will allow users to achieve their objectives, thus facilitating learning. In this work, we applied a survey to evaluate undergraduate students’ perceptions of ChatGPT usability and identify whether, from the point of view of usability, it is a good tool to use in education. The survey was applied to undergraduate students from two Chilean universities that belonged to two programs related to business and computer science. The survey was conducted online during October 2024. The questionnaire was divided into two sections. The first section was a characterization, and the second section contained the System Usability Scale questionnaire (SUS), widely used to evaluate the perceived general usability of different system interfaces. This work has a quantitative approach with an exploratory scope. The characterization information will be descriptively analyzed in percentages and frequencies. Data from the second part of the questionnaire were analyzed based on the usability categories defined and established in SUS. We had 117 participants, and the perceived usability score was 69.7, which is favorable for ChatGPT. In the analysis by gender and program on the perceived usability of ChatGPT (using the T-student test), we found that men have a better perception of usability than women; in turn, we found that students in computer science have a better perception of usability than those in business. This indicates that it may be a good tool to use in education. In future work, we want to expand the sample and focus the results on the positive use of this artificial intelligence tool in teaching and learning.