We all know that LLM, especially ChatGPT, is infiltrating educational centers. Resistance to prevent this influence does not seem wise because many educational experts and teachers have become supporters of these technologies and have evidence to prove their claim. As a teacher with more than ten years of teaching experience, I favor empirical research evidence and prefer to research them instead of fanatically resisting and rejecting the effectiveness of these technologies, or at least trust the results of experimental research that others have done! In this short article, I have summarized some effects of LLM on education systems.
While introducing a notable research achievement, I aim to encourage readers, especially educational technology researchers, to pursue further interdisciplinary and experimental studies.
Advanced Learning with the Help of LLM
In the previous article I published in Medika Life, I explained one of the strong psychology theories: Vygotsky’s sociocultural theory. That article also discussed the connection and application of social theories in LLM.
In general, socio-cultural theories believe that human learning occurs in the highest possible state in the context of culture and social interactions. I call this learning model community-centered personal development.
The main focus is on individual needs and skills, but the social context and constructive interactions support and guide the learning process.
This principle is embedded in LLM chat technology, particularly when used correctly. Intelligent and well-informed critics know that LLM technology does not eliminate the key role of teachers and human interactions but rather the role of a teacher who knows how to use LLM to facilitate the teaching-learning process.
In summary, some key factors of the effectiveness of LLM on effective education are:
- Learning in the context of social interactions:
These two-way interactions occur between information systems, technology, learners, teachers, and all the components present in the educational context.
This flow increases the speed of learning and prevents wasting time.
These interactions are based on the knowledge of individual needs and skills. So, it is better to call them personalized interactions because they focus on individual strengths and potentials. By identifying these potentials, they provide the basis for individual progress toward achieving the high goals of educational learning.
This continuous recognition and evaluation of individual potentials cannot be realized without interactions and social context.
- Gaps and deficits in learning are filled with the help of technology:
LLM fills gaps by giving quick feedback to the learner. The key point here is that LLM and GPT technology continuously evaluate the learner and give him feedback.
The learner’s and teacher’s awareness of the learning and teaching process has double and even multifold efficiency.
A Key question: Why is the efficiency higher?!
LLM gives quick and immediate feedback on the learner’s progress. The learner learns about his progress in small steps based on his personal speed. The steering wheel is in the learner’s hands. Therefore, awareness and self-confidence progress without comparing with others.
All these steps are done in the shortest possible time.
On the other hand, the teacher is informed about the flow of learning and teaching with LLM feedback. Therefore, technology evaluates the teacher’s teaching method quickly and systematically.
As a result, the efficiency and effectiveness of the teaching and learning process, compared to the process without the use of technology, are two or more times faster and more rhythmic. While increasing the quality, it prevents the wastage of time and manpower. Therefore, according to quantitative criteria, it is fruitful and profitable.
- The third key element in increasing the effectiveness of this process is the supportive teacher:
As mentioned earlier, deep-thinking experts know that GPT technology, if used correctly, does not eliminate human interactions and gives them a new meaning. Here, the role of the teacher is not removed or even reduced, but a new update is defined for it. This new role affects both the teacher’s knowledge and teaching and learning results.
The teacher has the same role which Vygotsky’s sociocultural perspective calls scaffolding.
In addition, the teacher is a supporter, and in addition to social interaction and constructive dialogues with the learner, the teacher uses LLM technology to increase the quality and quantity of the teaching process.
In this process, an artwork and outcome called” learning” are created. Learning is the product of the effort and interaction of the learner, teacher, peers, and educational facilities. LLM controls and guides this interaction.
Please see this article: https://medika.life/llms-as-a-modern-partner-in-Vygotsky’s-zone-of-proximal-development/
So, the LLM partner is this flow and helps the outcome (learning) fix its possible defects sooner and better. It does this with feedback, evaluation, speeding up, and enriching the information.
Therefore, it prevents the waste of energy and time and increases the quality and quantity.
In the next section, a report of experimental research is presented that confirms the effectiveness and facilitation of the education flow by ChatGPT. It is mentioned in the reference of this research. Those interested can refer to this research article for more references about the studies done have access.
A Research Report
As mentioned, research resources are needed to examine the relevance and effectiveness of LLM in the educational atmosphere. Fortunately, these detailed experimental investigations have been seriously started in different places. For example, consider a research paper derived from an actual experimental study. https://www.nature.com/articles/s41539-024-00273-3.pdf
As you can see in the link, this research has investigated the effectiveness of GPT4 chat and LLM in analyzing classroom dialogues and improving the quality of teaching.
In addition, the GPT method has been compared with traditional coding methods.
The subjects included middle school students, and the elective subjects in the classes included mathematics and Chinese language. The results of the research and statistical analysis were very interesting, which are summarized as follows:
- The use of GPT for coding saves users’ time, which cannot be compared to traditional methods. This time saving positively and significantly affects the teacher’s performance in teaching and evaluation.
- In addition, due to the fact that reliable methodology methods were used to check the efficiency of the performance, a significant reduction in the performance time and validity of the methodology was confirmed.
- This research, which was based on cultural-social perspectives, confirmed the effectiveness of dialogues on high-level learning of students, such as reasoning, collaborative problem–solving, and autonomy during learning.
- As expected, it was found that using LLM educational technology effectively increases dialogue analysis and important learning outcomes. The time range and speed of dialogue analysis are significantly reduced.
- For example, the dialogue analysis time by ChatGPT was 5 minutes, and in the traditional methods used in this research, it was 41 minutes, which is a significant difference. In general, the results of the comparisons showed that GPT surpasses the traditional methods in terms of accuracy and speed. So, the differences are both quantitative and qualitative.
Conclusion and Suggestions
Integrating LLM technologies such as chat GPT with the educational environment, regardless of the purpose and topic of education and who the learners are, can increase the quality of education and help teachers and trainers. Along with this qualitative achievement, it optimizes time resources and avoids spending extra energy and time.
In this article, we discussed some unique advantages and benefits of LLM for the educational environment and experimental research that included a specific group of subjects and teaching materials.
This research, along with other academic and experimental research, can make us more optimistic about the formation of a new world of education with the help of LLM, provided that we have a future-oriented and logical perspective.
The most important point is that we are at the forefront of conducting systematic and academic research on GPT and education. In other words, nothing without sufficient research in different situations, different educational subjects, different geographical and cultural conditions, ecological and social issues, biological variables, language, and finally, ethical issues and many other components contribute to the structure of artificial intelligence scientific research.
The future of artificial intelligence and education is meaningless without considering socio-cultural variables and everything related to humans. Therefore, as mentioned before, variables should be studied in the context of interdisciplinary research with the cooperation of experts from different groups and in experimental research. Another guideline is that researchers in the cognitive age must focus on the educational goals in other systems on high cognitive functions such as problem-solving, creativity, critical thinking, and reasoning rather than superficial educational functions and goals.
As a last guideline, whether teachers, coaches, parents, or any group and society that intends to use LLM for education and to facilitate educational performance, they should first be involved and update themselves with current knowledge.
Culture and knowledge of the correct use of technology take precedence over its use. This is an obvious principle in the application of all technologies, including artificial intelligence technologies!
