Contents:
Course with employment: "The profession of a Methodologist from scratch to PRO"
Find out moreHow the study was conducted
Forty-one students with C++ programming skills took part in the project. They were tasked with solving the classic "Eight Queens" puzzle using programming. The essence of the problem is to place eight chess queens on a 64-square chessboard so that none of them can attack another. This means that the queens cannot be in the same row, column, or diagonal. The students had to develop an algorithm that would calculate all possible coordinates for their placement, ensuring that these conditions are met. Solving this problem requires not only programming knowledge but also logical thinking and algorithmic analysis skills.
All experiment participants had access to additional materials, including three video lectures. These lectures examined the Eight Queens problem in detail, explaining the algorithms used to solve it, as well as solution approaches based on artificial intelligence and programming concepts. These materials helped participants better understand and master the methods necessary for effectively solving this problem.
Students were divided into two groups for the study: a control group and an experimental group. In the control group, participants attempted to write code independently after watching lectures. The experimental group, in turn, worked on a task, training an AI agent based on ChatGPT to solve it. This study evaluates the effectiveness of different approaches to teaching programming and interacting with artificial intelligence.
In the study, the authors configured the AI agent to simulate student behavior. Students participating in the experiment served as tutors for the chatbot. This required them to complete the following tasks:
- assess the bot's initial "understanding" of the topic being studied;
- clearly explain the training material to it;
- answer the bot's questions;
- teach it to solve a puzzle step by step;
- check how well the AI student "learned" the material and provide feedback until it achieved the goal.
After the experiment, the students completed a series of tests aimed at assessing the knowledge and skills they had acquired. The testing allowed them to determine the level of material assimilation and identify areas requiring additional attention. Test results will help teachers adjust the educational process and adapt it to the needs of students, which will contribute to a deeper understanding of the topic and successful mastery of the educational material.
- knowledge of the solution to the "Eight Queens" puzzle and the algorithm used to solve it;
- the quality of code writing - that is, its correctness and readability;
- the ability to self-regulated learning - this is about motivation and the ability to build an educational strategy.
Reading is a key aspect of personal development and broadening horizons. It helps deepen knowledge, develop imagination, and improve critical thinking skills. The importance of reading cannot be overestimated, as it enriches our inner world and promotes personal growth. Reading books, articles, and scientific publications can significantly improve the level of education and professional competence. In addition, regular reading improves concentration and memory, which has a positive impact on everyday life. Don't miss the opportunity to enrich your knowledge and skills through reading, as it is one of the most accessible ways to improve yourself.
Self-regulated learning is a process in which students actively manage their educational experience. This approach allows students to set their own goals, plan their actions, monitor progress, and evaluate results. Self-regulated learning is important for both educational technology (EdTech) and higher education institutions because it promotes critical thinking, independence, and responsibility in students.
In the context of EdTech, self-regulated learning enables the creation of adaptive platforms that take into account the individual needs and interests of students. This improves learning effectiveness and promotes deeper understanding of the material. It is important for universities to implement self-regulated learning practices because they help prepare students for real-world work environments, where independence and self-organization are key competencies.
Therefore, the development of self-regulated learning not only improves students' academic performance but also develops skills that will be in demand in the labor market. This makes this approach relevant and necessary in modern educational systems.
What the experiment showed
The experiment organizers expected that participants in the experimental group would complete the tasks faster. The results confirmed this assumption: on average, participants in the experimental group completed the task in two attempts, while students in the control group needed almost three attempts. This indicates more effective approaches and teaching methods used in the experimental group.
Students who taught the AI bot to solve puzzles demonstrated a significantly deeper understanding of programming concepts compared to the control group. This suggests that the process of explaining algorithms to the "student" contributed to the formation of a more conscious and profound understanding of the subject. This approach to teaching not only increases the level of knowledge, but also develops critical thinking skills and an analytical approach to problem solving.
Participants in the experimental group demonstrated the ability to write more understandable and readable code. The process of explaining their code to the "student" likely contributed to a better structuring of their own solutions. Furthermore, the experiential learning approach may have inspired students to focus on the clarity and accessibility of their code, an important aspect of software development. Improving the readability of code not only makes it easier for other developers to understand, but also promotes more effective collaboration within teams. Students who interacted with the AI bot demonstrated improved self-regulation skills. Compared to the other group, they were significantly better at developing an individual learning strategy and effectively self-monitoring the learning process. This confirms that the use of AI technologies in the educational process can contribute to improved student academic performance. Research shows that when students assume the role of a teacher, it promotes their sense of expertise. This, in turn, significantly increases their self-esteem and self-confidence. Additionally, other research confirms that observing the success of their students helps students in the role of a teacher realize their potential. This experience helps them refine their skills and strengthen their confidence in their abilities.
Reading is an important aspect of personal development and knowledge acquisition. It opens new horizons, broadens horizons, and helps deepen understanding of various topics. It is important not only to read, but also to choose high-quality sources of information. This could be useful literature, articles, research studies, or blogs that provide relevant and verified data. Regular reading helps improve critical thinking skills, as well as develop imagination and creativity. Therefore, it is important to devote time to reading and actively seek out new materials that can enrich your knowledge and experience.
Academic motivation, academic performance, and self-esteem are closely interconnected and play an important role in the educational process. Academic motivation determines a student's desire to learn and their willingness to overcome difficulties. High motivation promotes deeper assimilation of the material and improves academic performance.
Academic performance, in turn, influences self-esteem. When a student achieves good results, their self-confidence increases. This creates a virtuous cycle: high academic performance increases motivation to study, which, in turn, leads to even greater achievements.
Negative self-esteem can reduce academic motivation and, as a result, academic performance. Students who lack confidence in their abilities may avoid challenging tasks, leading to poor performance. Therefore, developing self-esteem is an important aspect of increasing academic motivation and, consequently, improving academic performance. Understanding the relationship between these factors allows us to develop effective teaching and support methods for students, thereby promoting their successful development and high achievement. Students in the experimental group made a similar number of coding errors as students in the control group, and in some cases even more, although this difference was not statistically significant. The researchers suggest that students interacting with the AI bot may have created code with errors already during the training phase. Successful problem solving was possible because the neural network generated completely correct code. However, this has its drawbacks: by delegating the task to a neural network, students are deprived of the opportunity to develop their own code-finding skills, which is critical for their learning and professional development.
The ChatGPT-based student bot demonstrated abnormally rapid learning, raising questions about its ability to maintain a consistent learning experience. Interactions with it ranged from a complete beginner to an advanced learner, making it difficult to assess its true knowledge. Furthermore, the bot showed no progress during the learning process, calling into question its effectiveness as a teaching tool.
Due to the unique characteristics of this neural network, this AI student is unable to fully rely on prior knowledge or demonstrate partial understanding of the material. Interaction with it differs from interaction with a human student, who gradually and consistently achieves understanding. This aspect limits the potential of AI in teaching, making its approach to information acquisition less flexible and adaptive than human learning.
The idea of integrating AI learners into the educational process has many advantages and can significantly improve the student experience. However, to achieve the desired results, the AI bots need to be further developed. The authors of the experiment noted the importance of allowing for the neural network to make errors. This will allow students not only to identify shortcomings but also to develop skills for correcting them, a key aspect of learning. Thus, the correct implementation of AI in the educational process can improve learning effectiveness and prepare students for real-world challenges.
The authors noted that for the learning process to be effective, it is necessary to develop tasks in which students explain to neural networks the terms and rules important for solving the assigned tasks. This approach promotes a deeper understanding of the material by the students themselves.
How else can such an AI bot be improved?
British online learning expert Philippa Hardman discussed the results of a recent study in her blog. She offered educational developers several recommendations on the skills an ideal AI learner should possess to effectively support students in the learning process.
Philippa believes that the ideal bot should have "controlled imperfection." She emphasizes the importance of neural networks periodically giving incorrect answers, which will allow the "teacher" to correct errors. Furthermore, the bot should be able not only to make mistakes in answers, but also to partially "learn" information, which implies the possibility of errors in its judgments, which it voices to its "teacher." This approach improves interaction between the bot and the user and enhances the quality of learning.
It's important for the AI bot to ensure continuity in the learning process. This means the bot should retain information about previous conversations and leverage its learning experience to enhance interaction. It should demonstrate progress over time, avoiding sudden transitions from novice to expert. This approach will create a more holistic and consistent learning process, which will promote better absorption of the material and increase user motivation.
The bot should exhibit natural behavior, which includes displaying difficulties in understanding the material. It should request explanations in a more accessible form, simulate confusion in appropriate situations, and ask increasingly complex questions. This will create a more interactive and engaging user experience.
The difficulty level of interaction with the AI bot must be adjustable depending on the student's level of knowledge. If the student has not yet thoroughly mastered the material, the AI learner should not ask overly complex questions or make errors that the student cannot detect. The neural network should be configured to ensure consistent skill development and deepen understanding of the topic through interaction with the AI learner. This will create an adaptive educational environment that promotes effective learning and knowledge growth.
Filippa believes that the "explain to someone else" method is one of the most effective pedagogical approaches, and artificial intelligence can play a key role in its scalability. To achieve this goal, developers need to focus on creating more realistic behavior for AI bots. Researchers should study how different AI behavior styles affect the educational process, and developers of educational solutions should actively experiment in this area and share their achievements and experiences. This will optimize educational technologies and make learning more effective.
