One common criticism of large natural language models based on artificial intelligence, such as Chat GPT, is that they hinder students' learning abilities and make it challenging for teachers to assign research papers as a learning product. However, I believe the opposite is true. In this post (and a subsequent one), I will outline a structured research process that incorporates various applications, aiming to transform learners into researchers in their field of study. In fact, the methods I will propose will make any research work much more engaging and relevant for the current generation of high school students.
One common criticism of large natural language models based on artificial intelligence, such as Chat GPT, is that they hinder students' learning abilities and make it challenging for teachers to assign research papers as a learning product. However, I believe the opposite is true. In this post (and a subsequent one), I will outline a structured research process that incorporates various applications, aiming to transform learners into researchers in their field of study. In fact, the methods I will propose will make any research work much more engaging and relevant for the current generation of high school students.
First, to Revolutionizing Research Processes with AI Tools, let's consider the learning objectives within the inquiry-based learning paradigm. Unlike traditional teaching methods that emphasize memorization and rote learning of definitions and concepts, inquiry-based learning requires the learner to be highly active – essentially "experiencing the life of a researcher" in the classroom under the teacher's guidance.
Broadly speaking, it involves the learner's ability to:
1. Formulate an intriguing question that doesn't have a predetermined answer.
2. Delve into the subject matter from reliable sources, assessing whether they might help answer the research question and identifying gaps in existing research knowledge. This step is crucial for refining the initial research question and formulating research hypotheses.
3. Gather information from the field through interviews, online research, and observations (depending on the nature of the research question and the field), process and organize the collected information into meaningful insights, and analyze it in a continuous and in-depth iterative process, specifying sub-questions and making initial attempts to confirm or refute hypotheses.
4. Answer the research question while adhering to the standards of quality scientific research.
5. Write the research report according to the conventions used in the field, citing sources, including sections of the research report, etc.
6. Remain open to criticism of the research, methodology, and conclusions, learn from the feedback, make improvements, and plan for future research.
Cultivating the mindset of an independent learner who believes in their ability to learn and grow through critical inquiry is no small feat. Such a learner should be aware of their biases, weaknesses in the research method, and unanswered questions that their research does not address. These are some of the qualities that need to be fostered in our students to implement this learning approach in the classroom.
It's clear that this process involves a wealth of skills, a critical perspective, and the ability to apply reflection while adhering to the standards used in the researched field. All of these will be enhanced in a learning context that incorporates new tools and applications of artificial intelligence, as I will demonstrate shortly.
Let's get started!
1.Formulate an intriguing question that doesn't have a predetermined answer
An innovative approach to addressing this challenge could involve the integration of Chat GPT tools or other large language modeling tools (such as Bard, Claude, Ruby, Bing, etc.). So, how can we employ these tools to identify the research question we'll begin with?
Here's one example:
1. Collaborate with students to craft a prompt that aids in arriving at the research question.
2. Request the chatbot to assume the role of an inquisitive child who poses numerous questions.
3. Narrow the conversation topic to the area you wish to explore (e.g., global warming).
4. Prompt the chatbot to discuss the topic with us and pose questions (you can also upload current articles in PDF format or provide links to specific websites to focus the conversation; it's also beneficial to connect the chatbot to the internet for more current and accurate responses).
5. Offer feedback to the chatbot regarding the quality of the questions and refine the conversation as it progresses.
6. Consolidate the most intriguing questions and rephrase them.
7. Request the chatbot to propose additional directions for the questions.
8. Seek assistance from the chatbot in evaluating the question based on standard parameters in the content field.
9. Select the question we'll begin with and document a reflection on the process thus far.
You can propose alternative methods for arriving at interesting questions through further interactions with the chatbot or discussions with classmates. I would appreciate it if you shared your ideas in the comments section of this post.
After settling on an agreed-upon phrasing and explaining the choice to your peers or yourself, indicate how artificial intelligence tools were utilized and assess their contribution to this part of the process.
It's crucial to conceptualize with students the added value of using these tools and the newly acquired skills, alongside the potential loss of other skills in the process. Reflection is vital for this innovative learning approach.
Proceeding to the next step:
2. Gathering information from the web related to the selected research question to deepen the understanding of the researched topic, identify information gaps, refine the research question, and formulate research hypotheses.
Since large language models based on statistics learn from information freely available on the web, which may not necessarily be research-based, their use for finding information for research is not recommended due to biases and hallucinations. For more information, see this post.
At this stage, it is advisable to use artificial intelligence tools that are connected to sources agreed upon by researchers in the fieldthe following video presents the possibilities of gaining control and precision in creating paragraphs based on research-based information. using Perplexity.ai.
This step is crucial in the research process, but if your students let the machine do everything, what will they learn from the process?
At this point, it's highly beneficial to include a significant step where students deepen their knowledge in the field and organize that knowledge in their own human memory.
I propose using artificial intelligence once again (!!!) to train the students and help them establish their knowledge in the researched field. Studies show that deep learning is effective when learners have to retrieve knowledge and apply it in new situations, which is exactly what I'm suggesting now.
So, how can we deepen human understanding using AI tools?
I recommend writing a new prompt for the chat and asking it to train me, the learner, on the knowledge by constructing questions at different knowledge levels. Start with easy questions about key concepts, and then continue asking about processes, challenges, and unresolved issues. All of these should appear in the text that the machine helped the learner write in the first step. Draft a prompt asking the student to play the role of a teacher in the field of inquiry. Enter the text formulated in the previous step as the learning source for the deep understanding step (by pasting or attaching a PDF file). Ask the chat to ask you knowledge questions about the text and give you feedback if you knew the answer. Add an instruction that every time you made a mistake, the chat would reformulate an easier question, and if you answered correctly, continue to more difficult questions. Give feedback on the questions and continue until you feel that you have mastered the text. You can continue at this stage with additional texts and the like.
You can also use tools like https://nolej.io/ , where you can add a file/topic/video and ask the artificial intelligence to build you (in seconds) a course for learning the topic found in the file. The course can be designed and edited, and of course, save you the time of drafting the prompts.
In the next post in the series, I will continue to describe the next steps. In the meantime, you can register as a member of the site to receive notifications to your email and write comments at the end of the post. Click here to join - Log In.
Additionally, remember that you can also order an inspiring lecture on ChatGPT technology and innovation in education today.
Good luck to all 21st-century teachers!