Study of the use of AI for learning the subject of programming I: case study of information technology career

Abstract

The objective of this study was to describe the trends observed in academic performance during the use of Artificial Intelligence (AI) in the Programming 1 course in the Information Technology degree program. The methodology consisted of a quasi-experimental quantitative design that integrated two sources of data: academic performance and student perception, at a descriptive level, with non-equivalent groups. To collect data, an experimental group was used with Artificial Intelligence tools, and two comparison groups under traditional instruction; the three courses shared the same syllabus and evaluation periods. The quantitative results show the averages of the three courses in the “alert zone” below 7/10, with no evidence of improvement attributable to the intervention of these tools. 56.5% of students in the experimental group perceived a reduction in academic frustration, and 78.2% expressed satisfaction with learning with Artificial Intelligence. A dependency was evident: 82.6% of students reported low confidence in passing an independent practical exam without technological assistance. It was concluded that these tools functioned as a technical safety net for syntax correction but did not replace logical-algorithmic reasoning.