Implementing Productive Strategies with AI
Learning Objectives
1. Understanding the Impact of AI Integration in Education
2. Utilizing AI Tools for Educational Enhancement
3. Learning Strategies for Generative Education
4. Understanding the concept of Productive Failure
Introduction
In my experience as an educator, I have keenly observed various challenges faced by students, notably ineffective strategies such as disorganization, procrastination, and hesitance to seek assistance. Instances of missed assignment deadlines due to a lack of proper tracking and last-minute task completion are not uncommon. As a student, I also encountered learning methodologies that proved ineffective studying until midnight or delaying assignments until the last deadlines. Recognizing the need for effective learning strategies is preferable in enhancing students' overall learning outcomes.
Moreover, there is a distinct advantage for educators who guide students in leveraging AI tools to foster productive learning strategies. This is not only cultivating academic integrity but also enhances efficiency, enabling learners to achieve more within compressed timeframes. The utilization of AI tools by educators is not limited to facilitating learning; it extends to the creation of evidence-based, productive learning environments. By incorporating AI, teachers can design curriculums that cater to individual student needs, fostering a more personalized and effective educational experience.
Drawing upon my experiences as a teacher and as a learner, I am acutely aware of the transformative impact that effective teaching strategies and AI integration can have on educational outcomes. I have had the privilege of teaching third-grade students in Afghanistan, and presently, I am engaged in the facilitation of online English classes for Afghan girls, contributing to the ongoing dialogue on effective pedagogical approaches and leveraging technology to bridge educational gaps.This chapter delves into the intersection of generative Artificial Intelligence and productive learning strategies, offering insights into how these two domains can coalesce for maximum educational benefit.
Expert Interviews
Interview with Ms. Sayma Rahman
Sayma Rahman: There are mainly AI platforms that I use for my educational purpose to help assist me in writing my assignments and all other stuffs to learn about the new information. But I think there are three AI tools that helps me the most. Number one is Grammarly as English is my second language so there are so many things that I am still learning about English grammar. When I write a paper, if I focus on grammar too much then it takes too much of my time. So, what I do is I write my paper. With my comfort. So, if I see any mistake, I just go through the Grammarly at last and it gives me the suggestions that what grammar should I put or use for the for my papers and all number second is a ChatGPT also from chat GPT I can take, I can have a concept clear concept about. How I should write that paper? What content should I write? Like basically, I always have an idea about how to write a paper or what content should I involve or put in that paper. But still if I put the things in ChatGPT then I can have more ideas like yeah, more. More knowledge like what to do or what not to put and number three is Bing there is also kind of ChatGPT but. Like, like ChatGPT can make us can make. Information on its own it cannot. Sometimes it is not valid or sometimes it is false. But in being you can have like proper. Proper data of what is happening currently or like solid data, it doesn't give you any false narratives or anything. So yeah, it is a lot to Fact Check like if my information are correct or not so Bing is also very useful.
Razia Sultani: So, thank you for sharing that, do you have any concern about privacy when you are using AI tool for your educational platform?
Sayma Rahman: Talking about privacy like yeah, privacy is very important for me. I don't know like how AI can use my information. It is just I I share my emails and e-mail ID and login to that AI platform, but yeah sometimes I get, sometimes it links to other applications which I think might be harmful for my for my personal space. I think I don't like that much invasion of my personal space. So yeah, still now I haven't faced any such issues and I really hope not to have any issues in future too.
Razia Sultani: What ways can AI be used to create a more personalized learning for students? Sayma Rahman: So, I think when a student or something it, it is saved in the history right? In the history so. As I am from educational background, I always search things about educational related topics or policies or suggestions or contents right so I think what AI can do is it can suggest me, it can suggest me like yeah this is this is a new educational. This is an education policy that has been implemented in these countries and they are benefiting from this. Would you like to go through it? So, it can be like it can suggest me some articles or some write ups or some policies related to my carrier phase so, it can be like if I am a student from a biology background then it can also give me information about new bacteria or new. New plants or anything that I can have information so I think, yeah, it can be used. Like this for more personalized learning, I guess.
Razia Sultani: Can you tell about yourself about your background and?
Melissa Warr: Sure, yeah. I originally was actually a music teacher, taught orchestra. And then I went back to school and I got a master's instructional psychology and technology with Rick West was my advisor for a master's degree who sent you to me, who you emailed before. And then I went to Arizona State University and got a PhD in learning literacy and technologies, looking a lot at how we learn through designing things and through creativity, and then the role technology plays in that and right now, I am an assistant professor at New Mexico State University. It is in Las Cruces, New Mexico, It is it is about 20 miles north of the border of the US and Mexico. And just recently with all the all the generative AI stuff and the large language model GPT stuff. Has started kind of exploring that a lot more than I than I was, I mean it is only been out a year and so a lot of my research has shifted into really considering. How that is used in education and, and really not just, I mean good you say, how it can help, but also potential bias and problems that might cause that we need to watch out for.
Razia Sultani: Great to hear, so the first question is, can you? Explain about the concept of generative learning and why these strategies are considered productive for learning.
Melissa Warr: Kind of like it is just, I mean it makes sense. It goes along with like maybe a constructivist perspective where we are generating information to learn. Is that kind of what you are thinking about? Like we were actually producing information and that is helping us learn.
Razia Sultani: Yeah, yeah, yeah.
Melissa Warr: It is just not a term I heard. I think, you know, there is my work in design. A lot of a lot of design is like you, you make something to learn so you make to learn and you learn through making. So, if I am going to. Try to for example, I have been I have been learning some coding lately and ChatGPT helps a lot with that and it is because I needed it for some studies I have been doing and so, I don't just sit and watch a video about how to do it right, I go to ChatGPT and I say, hey, I need to write a code that does this. How do I do that? And at first, I had to ask for a lot of details and have it well, what program do I even download to run this code? Like every single little thing, but I was doing it with a purpose. Right, I was doing it to learn to do something and then as I have gotten better at it, sometimes I am better at like the prompts. I write and I can be like OK, I need a code that takes these inputs and gives me this output. How you know what is , what are a couple ways I could do that and so I am generating those codes and I am pasting them in and I am looking at them and evaluating them and just through that process I am learning cause I, I have knowns some other programming languages, but I am programming Python which I hadn't before and I am I am kind of learning the details of what we call, the syntax, the language and stuff. But not from someone teaching me, from generating it on ChatGPT, putting it in and trying things out and then finding bugs and working back and forth. Is that kind of what you are thinking of that way?
Razia Sultani: Yeah, yeah, that is exactly what you are looking for.
Melissa Warr: And I, I like that term I gotta remember that. That is, that is a Have other people been using that term?
Razia Sultani: Yeah, I think we in this semester we have used like ChatGPT and other AI tools that can help us in our education.
Melissa Warr: Yeah, I was just wondering because, you know, they are, they are generative AI tools. I just hadn't connected that with a generative learning phrase. So that is that is just an interesting, interesting phrase and it makes sense. It is, it is a really interesting perspective actually now that I think about it because it is it is making, but it is making through. This generation versus through, I have to think about that, right? And I hope to myself.
Razia Sultani: So, and the next question is about another strategist that helping students in their educational journey like teaching, teaching others is as one of learning strategies. Involving. Explaining learned material to pairs and. In your idea, what distinguished teaching from self-explaining as a generative learning strategy?
Melissa Warr: Yeah, I mean, you know, anytime, I mean even as a teacher, like I am talking to
you here explaining something to you and it, it triggers new thoughts in my mind and it makes me break down more like what I am doing and what I am understanding so I think any, you know, that is that is why it is really good. I think for, for teaching other people is it really. First of all, you have to make sure that you know well enough to teach it. We know that, right. But in that communication process. It is changing that knowledge almost into a little bit of a different. Almost a different form that allows you to communicate with someone else so like what I just told you about that programming like I have never really thought of it that way exactly, but through explaining it to you, it kind of changed my perspective on it just a little bit, which gives me some new ideas and, and for me, learning is really about seeing things differently so when we learn something, it helps us see the world differently. It helps us understand, you know, in different is not better or worse necessarily but, but it helps us, you know, take a different perspective on something, and I think when we are teaching, we are having to think through, OK, what does the other person know? What would make sense to the other person? And then break down our own understanding and articulate or show it in a way that can support that and so I, I think that is one reason that is that helping others, you know, teaching others and really why education should be social overall because it is really should be a conversation that we are learning and discovering together versus just you know memorizing something by ourselves because we, we produce that understanding through connecting with people and connecting ideas across.
Razia Sultani: The next question is about are there any AI tools or platforms that support the summarization process effectively for students?
Melissa Warr: Yeah, I mean I think any of these AI tools, some are better than others. I have mostly worked with ChatGPT of course, Open AI with their model and then their four model which is the paid version. I lately I have been working with Gemini a little bit, which is Google's AI and with Anthropic, which is Claude. I don't know if you've heard of that and then I have done a little bit with perplexity and then I am trying to get into some of the like the Metas AI, which is Lambda, I think. I just haven't you have to get permission anyway I am in the process of looking at that and I think as far as the thing with base models of these large language models is that. They are just based on correlations of words, right so, it and you, I think you probably already know this so, it is it is, you know they train it to just look at a lot of data and guess what the next word would most likely be. So, if you are asking for information about something, if it is information that it saw a lot and its data like, like. I don't know the, the length of a river, right? It might have seen that river length is with the that a whole lot of times in its data. So, then it knows the length but if it is not a river that it saw very much because it is not very popular, it is just gonna guess, right? And it doesn't know when it isguessing. It doesn't know if what the correlation it is pulling up is true or not. And so, as these models develop. I think what is gonna get better at them to some extent is understanding when it is actually looking something Well, and it doesn't. The newer ones like Chat CPT it will actually go look up some things right instead of giving their best guess and so, it understands that when I am asking for a fact like a river length that it probably should go look that up and not just guess right. and so, I think when we come to which models are better. It is it is kind of going to be that which do they know when they are making stuff up and when they are not. Do they know when they need to refer to a document? So, and I am not sure if you mean by this by summarizing, but you know I have a friend that her students, you know take an article and they ask for a summary of it first and then they read it right and they are they are they are multilingual and so it is a little bit harder to read the English sometimes because that is not their native language, but that helps them with that language barrier but if you do that, some models will still make stuff up, right? Even, even though I am giving it APDF and saying summarizes PDF, if the model is not advanced enough, it will still try to fill in holes with stuff. And so, I think that is the main thing to watch out for when you are doing this type of thing and choosing models is, is is knowing well, and you just have to check it too. I mean, it comes down to right now at least you have to check it you can't trust it right off the bat but whenever you are doing that type of summarizing activity, just really making sure that you are not relying only on the summary, but you are losing you are using a tool that is going to directly reference that text and that is going to know that it shouldn't be adding in things that are not in that text you asked it to look at, right. Is this answering your question at all? Tell me what you mean by is this a summarizing? You are thinking of where you get something and it writes a summary for you. Is that what you are thinking of?
Razia Sultani: Yeah, same like this before a strong school, before this course that we studied. We were studying strong school in this semester and before that I was not very familiar with AI tools, but during this strong school course we have studied many AI tools like ChatGPT and we are for example, we have learned that we can give an article to ChatGPT and it can summarize and provide us the key points and import points.
Melissa Warr: Yeah, so I mean, ultimately ChatGPT forms a foundation there are also things like read PDF or that are built on ChatGPT, right? So, they use the same model, they just condition it in different ways and I think I, I know. Like, it is so funny. It was just like it was just a few months ago before ChatGPT 4 was letting you upload PDFs as easily like I would use read PDF to summarize and get information from PDFs, but really it was using the same language it was just applying it slightly different so it added parameters so it knew what to do and I think when we are looking at these other little platforms and tools people are developing most of them are built right now on Chat GPT, although there might be some there is some built on cloud, but ultimately they are just kind of adding a framework over that model to kind of condition what it can do and so yeah, it is it is interesting to see what is happening with that. That is probably off topic, but I am just thinking about models and what that means.
Razia Sultani: Interesting and fantastic tools that are now helping the students so the next question I am, can I ask the next question?
Melissa Warr: Yeah, sure,
Razia Sultani: Can you discuss how AI can enhance the mind mapping strategies, especially in creation of mind maps, concept maps or graphic organizer with AI?
Melissa Warr: That is an interesting question it is not a strategy that I have used yet, I have talked with my students about this. My biggest concern with AI for a mind map. I mean, you know, right now there might be tools that do the visual the most right now like if you ask it, it will give you like an outline of like what it should look like or whatever, right? But someday it is going to actually show, show you the graphic. For me, the purpose of building mind maps is to think about what you know and to make the links right so I actually worry, I am not sure and I could be wrong and this could be a use that ends up being really effective and I am sure in some cases it is I think in my classes, you know, if I ask my students to do mind maps, I want them to be making those connections themselves, right? So, I want them to be thinking through what that is in their mind, what, what pieces connect, and how the structure of this knowledge they are developing is so it actually is not a use that I would necessarily ask them to use AI for because I Most of the time, I don't think that. I think it would more get in the way than help. I am sure there is exceptions to that. So, I am just thinking. I am kind of thinking out loud here. It might be that they have a new topic area they need to look at and so getting a mind map might give them subtopics to start looking into, and to help them structure what they are understanding and how it fits together so that might be an appropriate use that way. I just overall I would prefer that they made their own connections but I think it might be a good way to give an overview of something that you know you do not have deep knowledge of yet and you don't have time to like research everything on it. You know, I think what I would prefer to do if I were to do this in my classes is for them to ask ChatGPT will you give me a list of topics related to this idea? And get the list and then have them map out the list with the relationships among them. I think that is probably what I would prefer, even asking for topics and information about the topics and then have them map the connections. It is probably more what I would do, but I haven't thought a lot about that. I have heard of some people doing it and I have talked to my students a little bit about it, but yeah, sorry, I am not, you know, it is funny, everyone's doing different things with these tools so, internationally it is different too, so I would be curious to hear what you've been doing with mind mapping.
Razia Sultani: In this semester we used like, I am not familiar with many AI tools, I have used ChatGPT for mind maps and for just giving some ideas from AI that how we can create a good mind map and connect between them yeah, we have used.
Melissa Warr: Yeah, I do. I am a big fan of using ChatGPT for idea production so like, I am always trying to write better titles for articles that are more interesting, right? but, I think there is there is a scholar, Arizona State University. His name's Ron Baghetto, and he is made some little applications. He studies creativity and he's made some little applications about how these tools can spark creative thinking, which seems counterintuitive to a lot of people, right? Like this is the opposite of creativity, right? But I think it helps, it can help you think in different ways and from different angles and so having it help you brainstorm stuff as a thought partner, right. Not it just telling you things, but like we did, we had this theoretical framework we were trying to some colleagues and I were trying to refine and we were a little bit stuck with it, and so we actually had a conversation say, hey, you know, we are struggling with this framework. We don't know what to do what are some different possibilities we could do? And it gave us some possibilities and one of them, we found it specifically interesting so then we said, oh, that sounds really interesting, like what if maybe we what else could we do with that part? but it took our collaboration and ultimately the approach we took, it was not anything it directly told us, but what the conversation we had helped us think of it kind of like, you know, when you talk with someone you get new ideas sometimes that is that, is kind of what it was like so.
Razia Sultani: So, the last question is, what is your opinion about the future role of AI in education?
Melissa Warr: Such a big question so what I really like AI for? Umm, I really like it for practicing, conversations and role- playing so, and this is just one application, right? But well, two things. I like it for learning stuff like I am learning coding, right? It is providing a support. I couldn't do what I am doing without it right now because I would, I could, but it would take me years because I'd have to learn all the stuff, foundational knowledge, and then it would take me forever to figure everything out. So, I like it when it provides a shortcut to something that enables you to do more right it is providing me with some code that helps me do my job better and so, it is providing that support, what I really like in education as well. Turn off my notifications, is this role-playing so for example, if I was going in for a job interview and I and I was going to be interviewed for a job as a communications manager at a edtech company, right? I could tell ChatGPT I want to practice some interview skills. You are going to be the interviewer at this such and such company. I am going to practice responding to your interview questions and then you can actually do a practice interview and ask for feedback afterwards. So, so things like that we are doing it right now with some with future teachers, so people were training to be teachers. We are asking ChatGPT to be like a third-grade student who doesn't understand fractions and then our future teacher is trying to help that third grader that is ChatGPT understand fractions, and so it comes up with, you know, it says, well, I thought 2/3 + 2/3 was just 46, Why would it be any different? Or you know, it comes up with stuff and then they have to explain and then after they do that, they can actually say, hey, give me feedback, what did I do? Well, what could I do differently? Do you have any other angles you could take that could help the student? And so, as I look at the uses of this, I, I am more of a fan of some of these interactive type of things in education then then some of the more direct the more direct uses that people are talking about, like having it somewhere not summarized. I think summaries are really good because I think it If it is used as a bridge to something, right, It is not used as the end all. It is used as a bridge for learning that is what I like, I do worry a little bit about they talk a lot about personalized learning and how it can act as a tutor with the student. I have been doing some studies that show how biased it is and I am worried with that application how it can perpetuate some inequities in society. So, I don't need to get into that, but I am a big fan of using it where the user is in control so, me as a user I am choosing what I want to do with it. I am very nervous about a student sitting on it and it telling the student what to do, right? So, any of these uses, like summarizing stuff where you are taking the power and it is working for you, I am a fan of those IA lot of the talk though is about OK, now it can be a personal tutor for every student right and that relationship is usually the tutors in charge and that worries me because you know, as kids learn and the way that we talk to them, their identities develop so if the if the AI tutor is talking to them in a certain way, it is going to impact what they think of themselves, and the way it is talking is pulling it on the current discourse in society right now, so it is going to perpetuate and continue some, some problems we already have so I think I think there is definitely going to be some of that that personalization type aspect, I am not a fan. I am a fan of producing things I am also a fan of getting feedback right so sharing your writing and saying, you know, how could I improve this? What could I do to do better? Where you are asking it, it is doing something for you, right? You are not doing something for it. I think that is the important distinguishing factor in my mind right now is what I think is appropriate. And I am hoping people are more going towards that. I am also hoping we kind of get past our fear of plagiarism that and cheating. Everyone's so afraid of cheating that their students are cheating, right. I am hoping it is going to push people to think more about what is a meaningful assignment because if ChatGPT can do the assignment for them, then why are we even having them do the assignment and that is you know there is some basic writing you need to learn and stuff so that is that is an extreme position but I am hopeful that it will, it will kind of help people think more about what it means to learn and. That is not necessarily like we can have really good ideas even if we don't have all the writing skills and it can help us develop those ideas and fill in some of the holes, we might need from time to time. But I am probably way off you’re your question, but those are kind of my thoughts about where it's maybe going, where I think it should go or shouldn't go, I think. Yeah.
Razia Sultani: Right, that's the point so, it's all. Thank you for your time and if you have anything to share or anything you want to ask form me.
Melissa Warr: Yeah, yeah, I mean, I yeah, I mean I am interested so this is a class we are using ChatGPT in the OR using generative AI in a lot.
Razia Sultani: Yeah, and this class was like if you were learning and how we can use the AI tools that can help us in during this educational journey. We have used, we have used and learned many AI tools like. ChatGPT, Bing Chat. Badges, badges, badges, and many items. We have learned how to make videos, pictory videos and also, we have learned about how to summarize videos. I didn't know about that that we can just use AI tools to summarize a video and we have learned in this course.
Melissa Warr: Yeah, that is another. That's another good one what tool are you using for summarizing videos?
Razia Sultani: It’s named. It's just the name is summarizing video, video summary tool.
Melissa Warr: Yeah. OK. I am going to look at that, that is that is a good that is a good plan. You know the other use I like it for is for academics there is have you heard of Elicit or. What's it called? Research rabbit? Or where they help you find connections of papers? And it provides those brief summaries, right, of papers that help you find, get an overview that you can then dig into better. It kind of streamlines that process a little bit, although I do worry a little bit that we might be missing some things if we only do that but it helps me a lot when I have to do something fast.
Razia Sultani: Yeah, it really helps.
Yeah, definitely so. Great well, I am curious, let me know if you can share what you write about this, I would be interested in seeing.
Razia Sultani: Sure, I will do that
The Art and Science of Learning
Learning is a complex process characterized by transformation of knowledge, the synthesis of understanding, combining of information, and establishment of connections to real-world applications. Learning is not only obtaining and memorizing information rather, it encompasses the systematic organization and construction of meaningful insights, facilitating the practical application of acquired knowledge in real-life scenarios.
In addition, significance comprehension of various learning techniques is imperative. Leveraging diverse educational resources such as textbooks, videos, audios, and visual aids plays a pivotal role in enriching the learning experience. Furthermore, active participation in group discussions within the classroom setting and the exchange of data and information with peers emerge as effective learning strategies. These collaborative approaches not only enhance the learning process but also contribute to the enjoyment of the educational journey.
Educators are pivotal in recognizing optimal learning strategies, understanding students' capacities, and determining effective teaching approaches to transfer knowledge.” As such, AI-enabled personalized education promises to yield many of the benefits of one-on-one instruction at a per-student cost similar to large university lecture classes.” (Maghsudi et al., 2021) Today, AI tools serve as alternatives, possessing the capability to assess individuals' knowledge, offer personalized guidance, break down extensive information into smaller parts, and simplify complex data efficiently at the same time decreasing financial cost for learning.
For limitations of the human brain, it is crucial to explore ways of presenting information to adjust with our cognitive processes. Productive learning involves incorporating current methods, integrating diverse strategies to form new approaches, employing proven methods backed by experiences and research, and adapting strategies that resonate with students which is possible with the help AI tools
Effective information delivery to students includes breaking down lessons into manageable segments, simplifying complex topics, organizing information logically, and involving students in active practice sessions. Teachers and AI platforms support students in comprehending and retaining knowledge more efficiently, facilitated by technology and AI tools.
Learning Strategies that Foster Generative Learning
"Summarizing, Mapping, Drawing, Imagining, Self-testing, Self-explaining, Teaching, and Enacting" (Fiorella & Mayer, 2015) represent eight impactful learning strategies that foster generative education. Each of these strategies provides students with a unique avenue for engaging in generative learning approaches.
Thus Generative Activity Principle video was created by Aaron Chia Yuan Hung, for the EdMedia conference, delves into the effectiveness of generative activities in promoting understanding during and after multimedia lessons. These activities, outlined in the research by Fiorella and Mayer (2015), include summarizing, mapping, drawing, imagining, self-testing, self-explaining, teaching, and enacting. Aaron Chia Yuan Hung also mentioned that these generative activities have positive effects on learning outcomes when implemented correctly. Summarizing, mapping, and drawing involve creating representations, with college students benefiting from summarizing, while younger students may need guidance. (Fiorella & Mayer, 2015) Mapping benefits from interactive graphic organizers, especially for younger learners. (Ponce & Mayer, 2014) Drawing and imagining, creating mental images, are well-supported by students, particularly with specific prompts.
Self-testing and self-explaining, involving testing and explaining understanding, are beneficial. Teaching, where students teach someone else, is effective, showing benefits for older students equal to or greater than regular instruction. Enacting, acting out lessons, is most beneficial for academically stronger students.
Summarization stands out as a potent learning strategy, facilitating enhanced comprehension of text, videos, or lectures (Fiorella & Mayer, 2015). Learners can actively engage in this strategy by taking notes, jotting down keywords, and organizing crucial information from a lesson. This not only aids in easy recall but also fosters connections with existing knowledge and personal experiences. For example, when a student articulates the essential points in their own words during a class, it contributes to a more profound understanding of the topic.
Learning by Mapping
Mind mapping emerges as a valuable instrument for brainstorming, organizing, and visually presenting information (Fiorella & Mayer, 2015). This technique aids in data collection, categorization, and provides a comprehensive overview of planned written content. It serves as a strategic approach for constructing a broad outline of a project, essay, or chapter book. To better explain it, students have the opportunity to use a graphic organizer, mapping out details for an essay, including the introduction, body paragraphs, and conclusion. This visual aid gives readers a clear understanding of the essay's organization.
Mainly there are three types of mind mapping:
· Concept Mind Mapping: Writing the key points in picture boxes
· Knowledge Mapping: putting the information and data to diagrams and maps
· Graphic Organizer Mapping: Putting the important information in charts or table
AI tools play a crucial role in aiding keyword identification and mind mapping. Notably, tools like ChatGPT and Bing Chat are valuable resources for the creation of mind maps. They contribute by offering insights and suggestions for essay and chapter book writing, as well as providing constructive feedback on the structure of mind maps to facilitate improvement. These tools serve as effective aids in enhancing the overall process of ideation and organization, making them valuable assets in academic and creative pursuits.
Learning by Self-Testing:
Self-testing involves actively participating in practice tests and quizzes without external assistance (Fiorella & Mayer, 2015). This process serves multiple purposes, including refreshing memory, fostering critical thinking, and assessing one's knowledge and comprehension of the learned material. It is a method to establish meaningful connections and the application of acquired knowledge in real-life scenarios. AI tools provide students with the opportunity for self-assessment and preparation for authentic examinations. Numerous online applications offer a platform for students to engage in practice sessions, allowing them to adapt themselves to real exam scenarios while measuring their skills and understanding. For example, the Duolingo application, which aids language learners by offering practice tests and exercises to enhance language comprehension. Notably, Duolingo's preparation of complimentary practice tests enables students to assess their proficiency, delivering an estimated score that facilitates a comprehensive understanding of their capabilities, motivating them to further refine their skills.
Learning by Drawing
Drawing is a powerful learning strategy, especially for science learners. Students can draw based on the text or be instructed to draw according to the content. The effectiveness of learning by drawing is evident when students receive proper support during the drawing process (Fiorella & Mayer, 2015). The efficiency of learning through drawing is enhanced when students possess drawing skills. Additionally, when instructors guide students in drawing, it becomes an even more effective way of learning. Numerous online AI tools are accessible to assist and provide guidance in the art of drawing. These applications, including AutoDraw, Artbreeder, Deep Dream Generator, Doodle Addicts, Morphi, and SketchAR, tailor to students' needs by aiding them in translating their ideas into visual representations. By following student instructions, these tools empower learners to effortlessly illustrate concepts studied in school or university.
Learning by imagining
Imagining strategy fosters learning by encouraging learners to create mental images related to the text in their minds (Fiorella & Mayer, 2015). This approach involves making a visual representation of the text, helping learners form a clearer understanding of the content. To effectively learn by imagining, learners need instructions on how to create mental images based on the text and connect them to the content. By utilizing AI tools, students can craft personalized stories, boost their creativity, engage in critical thinking, and compose essays and poems, thereby enhancing their learning experience. For instance, ChatGPT facilitates individual learning journeys by generating personal stories based on student instructions, adding an enjoyable dimension to the educational process
Learning by Self-Explaining
Chi and colleagues (1994) conducted a study to investigate whether prompting students to self-explain while reading a scientific text led to deeper learning (Fiorella & Mayer, 2015). The findings revealed that self-explaining resulted in improved student scores in exams. Students can engage in self-explanation by discussing the learned material with friends, writing a concise summary in their own words, asking questions, and connecting the information to their existing knowledge. From my perspective, self-explaining strategies prove beneficial during group discussions in class, providing each student with the opportunity to express their ideas and understanding of a particular topic. This approach actively engages students with the subject matter, making it more interesting and informative. AI tools can engage in conversations with students much like a human peer. They have the ability to ask questions, provide feedback, organize ideas, and express creative thoughts. These tools are accessible anytime, anywhere, allowing students to seek assistance whenever required. This advancement in technology contributes to a more personalized and enjoyable education experience. For example, the ability of students to translate words or paragraphs using AI tools like Google Translate or ChatGPT which is able to recognize, and analyze information for learners.
Learning by Teaching
In 2005, fifth-grade students engaged with a computer-based program called "Betty’s Brain" designed by Fiorella and Mayer (Fiorella & Mayer, 2015) The program involved drawing maps of rivers while providing feedback based on students' information. The results showed that students who worked with Betty’s Brain demonstrated a better understanding of maps compared to those who did not. Additionally, activities such as preparing for teaching, explaining to others, and communicating with students during teaching were identified as effective learning strategies (Fiorella & Mayer, 2015). Teaching and interacting with students offer a more in-depth understanding of the topic. From my perspective, teaching significantly impacts our understanding of a topic, as it involves thorough preparation, explanation to students, and engaging in questions and answers, fostering a deeper exploration of the subject matter.
Learning by Enacting
Effective communication involves the utilization of body language and gestures as valuable tools for transferring information and knowledge to others (Fiorella & Mayer, 2015). While body language is considered an essential skill, its effectiveness can be further enhanced when children receive guidance from instructors. The enacting learning strategy, commonly employed in subjects like math and science, proves particularly useful in providing clear explanations of complex mathematical concepts to students. Additionally, incorporating physical objects, toys, and hand gestures emerges as an effective approach to engaging listeners and vividly portraying the scenes of a story. Several online applications support students in preparing presentations and providing feedback on their presentation skills. AI platforms contribute to creating reliable learning environments or experiences like real classroom activities. For instance, tools like Virtual Speech offer opportunities for practicing public speaking and interviews, serving as valuable resources to enhance communication skills. These AI tools play a crucial role in assisting users in developing their abilities to communicate effectively.
Historically, failure has been perceived as an undesirable outcome, similar to a formidable fight that individuals sought to avoid. However, a paradigm shift in instructional design has given rise to the concept of productive failure which has proven to be a highly effective method of teaching. According “The PF learning design is more effective in developing conceptual understanding and transfer than a direct instruction design” (Kapur, 2015) In the traditional teaching approach, students receive instructions before taking problems that is a common method of delivering knowledge. However, the productive failure approach diverges by presenting problems to students before initial guidance. This encourages students to engage in critical thinking, fostering the development of innovative solutions. Notably, students, in the absence of clear instructions, integrating various solution steps, leading to a more comprehensive understanding of the subject matter. Subsequently, when instruction is provided after students have faced with problem-solving and understanding a new concept, the learning experience is notably improved compared to the direct instruction approach. For example, students engaged in "comparing the variability in rainfall in two different months of a year" (Kapur, 2015). Whether worked individually or collaboratively, students undergoing both direct instruction and productive failure expressing diverse problem-solving approaches. However, the subsequent analysis of errors and misconceptions showed that the students who were guided with productive failure method has more effective outcomes compared to direct instruction. The shift towards productive failure signifies a departure from conventional teaching norms. It underscores the value of allowing students to deal with problems independently before receiving a clear instruction, fostering a depth of understanding and innovative thinking that surpasses traditional methods.
Expert Corner / Recommended Resource
The Keynote lecture at the 2023 EdMedia conference :Productive Failure and Learning Technologies” was given by Manu Kapur, a professor at ETH Zurich. Prof. Kapur discusses the concept of productive failure and its application in educational interventions. Productive failure involves promoting learning through failure and problem-solving activation, followed by instruction.
The speaker provided an example for productive failure to show the design principles applied in a math lesson on standard deviation for 14-15 year-olds. Instead of directly instructing the concept, students were given the task of designing as many ways as possible to determine the most consistent soccer player based on goal scores across several games. The students generated various solutions such as using averages, visualizations, and comparing year-to-year deviations. All these are aspects of the concept of standard deviation, and activated the learners" critical thinking and prior knowledge. The concept of standard deviation was then explained by the teacher in the direct instruction section of the learning unit.
Conclusion
AI tools like ChatGPT and Bing Chat have the capability to efficiently summarize extensive texts, providing students an overview and a general understanding. However, complete reliance on AI for summarization may adversely affect students' summarization skills. Student-generated summaries, forms with a personal touch and in-depth understanding while utilizing AI tools for summarization, students may skip reading the entire text, as AI tools highlight important points and provide a broad overview. This approach may not be conducive to long-term student development. balancing between AI assistance and personal summarization efforts is essential to ensure optimal learning outcomes. Moreover, numerous AI tools are available to assist students in creating mind maps, but an overreliance on AI may impact students' creativity. Creating mind maps manually allows students to personalize their maps and clearly illustrate connections among different concepts. Additionally, AI tools that help students transform text into drawings are beneficial, especially for those not skilled in drawing. However, regular drawing activities enhance memory and promote a deeper understanding of concepts. Furthermore, AI tools can generate imaginative stories by following instructions. However, if students solely depend on AI for story creation, it may hinder them from expressing their personal thoughts and ideas. Seeking assistance from AI to foster their own creativity can contribute to a more productive learning experience. Many AI platforms offer AI-generated tests, enabling students to understand their strengths and weaknesses. However, relying solely on AI tools may have a long-term impact on our knowledge. Therefore, maintaining a balance between self-generated and AI-assisted tools will lead to more effective learning outcomes. AI tools, such as ChatGPT, possess the ability to explain complex concepts by breaking them down into simpler parts and providing examples to convey their message effectively. However, relying solely on AI-generated explanations may result in a lack of depth in understanding for students. Manual reading can provide a more comprehensive understanding of the concept. balancing between utilizing AI-assisted tools and traditional learning methods is important to achieve optimal learning outcomes. Lastly, AI tools can provide realistic teaching experiences and offer various platforms for engaging learning experiences. However, AI tools are structured in a way that they may make mistakes, emphasizing the importance for students to use a combination of human instruction and AI for a better understanding of the educational concepts Balancing the use of AI-assisted tools with traditional learning methods allows students to gain a more comprehensive understanding of complex concepts.
Glossary
1.Artificial Intelligence: A branch of computer programs that focuses on creating machines capable of performing tasks that would typically require human intelligence.
2. Productive Learning Strategies: Approaches and techniques employed to enhance the learning experience and improve educational outcome, Such as summarization, mapping, drawing, self-testing, self-explaining, teaching, and enacting
3. Generative Artificial Intelligence: AI systems capable of generating new content, ideas, or solutions for instance creating personalized learning materials based on personal needs and abilities or offer guidance.
4. Personalized Education: Tailoring educational experiences to individual students' needs, abilities, and preferences.
5. Mind Mapping: A visual technique used for organizing and representing information, often in the form of diagrams or charts
6. Self-Testing: A learning strategy involving active participation in practice tests and quizzes without external assistance.
7. Artificial Intelligence Tools: applications or systems that utilize AI algorithms to perform specific tasks. In education, AI tools can assist with tasks such as summarization, concept explanation, and personalized learning experiences.
8. Summarization: The process of condensing information to its essential elements. In context of education, summarization is a learning strategy that involves extracting key concepts from texts, videos, or lectures to enhance comprehension.
9. Generative Activity Principle: A concept emphasizing the effectiveness of generative activities, such as summarizing, mapping, drawing, imagining, self-testing, self-explaining, teaching, and enacting, in promoting understanding during and after learning sessions.
10. Productive Failure: A teaching approach that presents problems to students before providing initial guidance, encouraging critical thinking and the development of innovative solutions.
11.ChatGPT: A conversational AI model developed by OpenAI. It can be utilized in various applications, including summarization, concept explanation, and generating personalized content, and offer individualized guidance.
12. Bing Chat: An AI tool that facilitates interactive conversations and can be employed for tasks such as creating mind maps and providing guidance on various academic subjects
13. Human Instruction: Guidance and teaching provided by human educators.
14. Learning Outcomes: The measurable results and achievements resulting from the learning process.
15. EdMedia Conference: An event where educators, researchers, and professionals gather to discuss and share insights on educational media, technology, and innovative approaches to teaching and learning.
16. Online Learning Platforms: Virtual spaces where students can access educational resources, interact with AI tools, and engage in various learning activities
Reference:
1. Fiorella, L., & Mayer, R. E. (Eds.). (2015). Learning by Imagining. In Learning as a Generative Activity: Eight Learning Strategies that Promote Understanding (pp. 79–96). Cambridge University Press. https://doi.org/10.1017/CBO9781107707085.007
2. Edmedia.(2023) Productive Failure and Learning Technologies by Manu Kapur [Video]. https://www.youtube.com/watch?v=oMzgzu0Il8o
3. Aaron Hung. (2023) Generative Activity Principle [Video]. https://www.youtube.com/watch?v=vF4Fqj0NwAk
4. Strong Schools.(2023) AI for Language Learning [Video]. https://www.youtube.com/watch?v=r0T9S_wTTOs
5. Maghsudi, S., Lan, A.S., Xu, J., & van der Schaar, M. (2021). Personalized Education in the Artificial Intelligence Era: What to Expect Next. IEEE Signal Processing Magazine, 38, 37-50.
6. Kapur, M. (2015). Learning from productive failure. Learning: Research and Practice, 1(1), 51-65. http://dx.doi.org/10.1080/23735082.2015.1002195
This content is provided to you freely by EdTech Books.
Access it online or download it at https://edtechbooks.org/ai_in_education/implementing_productive_strategies_with_ai.