The influence of Vibe Coding on programming education in universities and its experimental evaluation

Abstract

The purpose of this article is to identify the causes of the phenomenon of “vibe coding,” which is often encountered in the teaching of programming disciplines in higher education institutions, and its impact on learning outcomes, as well as to evaluate the effectiveness of methods aimed at preventing it. Currently, the decline in the search process and the ability of students to solve problems using artificial intelligence is becoming a serious problem in science. In the context of teaching programming disciplines, the implementation of tasks designed for algorithmization and the creation of a training program using artificial intelligence (AI) is described by the term “vibe coding.” In this regard, to determine the extent of the active spread of “vibe coding” tools among students, the level of their use, and the problems of preventing them, a survey was conducted with students of the “Informatics” program at the K. Zhubanov Aktobe Regional University. As a result of the survey, 96% of students reported regular use of AI tools. The study identified the main factors that prompted students to use AI-generated answers, ignoring the requirements of academic integrity. In addition, a methodological tool was developed aimed at developing algorithmic thinking, determining the authorship of a solution, and providing a step-by-step explanation of tasks. The effectiveness of the developed methodological tool was tested using a pilot experiment. Forty-seven students participated in the experiment: 23 of them were assigned to the experimental group and 24 to the control group. The experiment revealed a statistically significant difference between the two groups. Cohen's d statistical effect size was 0.671, confirming a medium-high effect size, i.e., the influence of the experimental methodology used was reliable.