How to use AI to Help, Not Hinder, Your Learning
Objectives
By the end of this chapter, you will be able to:
- explain the role of motivation in learning, and how AI can help, not hinder, motivation.
- explain how knowledge is formed, retrieved, and stored, and how AI can help, not hinder, knowledge building.
- identify use cases for various generative AI tools to help you learn.
Dr. Vasquez teaches an undergraduate class in the Gender Studies program called “Women in American Society”. She is concerned about some of her students. Several of the latest essays that were turned in were obviously generated by an AI chatbot without any original thought from the student. Some students just didn’t turn in the essay, and she is worried about whether they will be able to stay caught up in class. She asked her TA, Ahmed, to reach out informally to some of these students to find out what was going on.
Ahmed reports back that Martina admitted to using ChatGPT because she just wasn’t really interested in the topic and had other assignments she was trying to complete at the same time. Greg admitted to using Co-Pilot to write the essay and said that they are not comfortable with writing and just don’t know how to get their thoughts out in academic text. Louisa had not turned in the essay because she works full-time and had a sick child at home. She had done the reading and outlined her essay but had not had the chance to draft it yet. Dr. Vasquez wonders how she can help her students like Martina, Greg, and Louisa to use generative AI to help, not hinder their learning.
The Problem
Generative AI chatbots burst onto the scene with widespread adoption at the end of 2022, surprising many people with their ability to generate a variety of types of text that are generally indiscernible from human-written text. Some students have found that AI chatbots make it so much easier to cheat, while other students are afraid that AI will ruin our lives and want nothing to do with it. It’s unlikely that schools will be able to prevent students from using AI in their learning and effective use of AI chatbots is fast becoming a requirement in many occupations, so it is important that teachers and students alike learn how AI can help the learning process and when it should be avoided because it disrupts the learning process.
Motivation
Aspects of Learning In general, AI should not be used to do the work of the learning objective. If the goal of a learning task was to use well-written prose to describe an important event in your life, then asking ChatGPT to write this for you would not only be unethical, but it would cheat you out of the learning experience. The goal was to improve your writing ability and generative AI did that work instead. However, if the goal of the learning activity was for you to understand the interplay of factors that brought the Cold War to an end, then asking an AI chatbot to help you think through the various events and situations can actually be useful in helping you meet this goal. If you do the work of understanding and explaining the end of the Cold War with the help of generative AI, then you are doing the work of the learning objective. If writing skill is not the objective of the learning activity, then generative AI can be helpful in providing you feedback on your writing or ways of phrasing your ideas about the end of the Cold War that may make them clearer to the reader. Further resources for distinguishing appropriate use of AI for learning are presented by Ditch that Textbook and Kate Meyer.
In order to determine when it is appropriate to use generative AI for your learning, it is first important to understand how learning happens. In this section, we will explore some of the most important components of your learning journey: motivation and the formation and retention of knowledge. Each section will provide examples of how you can use generative AI to help and not hinder these aspects of learning.
What is it and how does it work?
Why do you do the things you do? Some things you do without thinking too much about them because they are habits or automated, like the procedure for riding a bike. Some things you plan ahead for and go on autopilot because they are part of your routine, like riding the subway to campus. Most of the rest of your actions are either a reaction to a trigger or are things you choose to do for a reason. This reason is your motivation. Motivation also sustains your actions toward achieving a goal. Interestingly, success at a task, like learning, increases your motivation for completing similar tasks. So, even if you don’t feel like learning about something, by forcing yourself to get started, you might grow the motivation you need to continue.
You may have heard of extrinsic motivation, where you choose to complete a task for an external reward, such as a grade or money, and intrinsic motivation, which stems from internal rewards like satisfaction and personal growth. Researchers have defined and studied motivation in several ways, but we will unpack one of these models here so you can see how AI can help or hinder your motivation for learning.
According to the Situated Expectancy Value Theory (SEVT) by Eccles and Wigfield (2020), our motivation is determined by contextual and situational factors, like social identity, background, and previous experience, as well as goals and academic self-concept, or what you believe about yourself in terms of learning. These all influence the key determinants of motivation: expectations for success (ES) and subjective task values (STV). Your choices and performance all return to feed your previous experiences, goals, and academic self-concept, influencing further motivation (see Figure 1). Let’s explore this in a little more detail so we can understand what these components mean for you.
Figure 1
Simplified Model of Situated Expectancy-Value Theory by Eccles and Wigfield
Note: This figure summarizes the main components of the SEVT described in Eccles & Wigfield, 2020.
Eccles and Wigfield (2020) define expectancies for success as “individuals’ beliefs about how well they will do on an upcoming task” (Eccles & Wigfield, 2020, p. 3). Your judgment about how well you will do on an assignment or assessment (like a quiz or exam) is determined by your social identity, background, previous experiences, your goals, and your academic self-concept. Whether you think you will do well on the learning task is not enough to motivate you, however. Motivation is also influenced by your unique view about how valuable it is for you to complete a given task.
Subjective Task Values (STV) are viewed differently by each individual learner. Overall value is determined by intrinsic value, attainment value, utility value, and cost. Intrinsic value is “the anticipated enjoyment one expects to gain from doing the task for purposes of making choices and as the enjoyment one gets when doing the task” (Eccles & Wigfield, 2020, p. 4). You might not think that a learning activity can be particularly enjoyable, but they can be! Utility value is “how well a particular task fits into an individual's present or future plans” (Eccles & Wigfield, 2020, p. 5), or a means to an end. Even though you may not realize it, every time you get started with homework, you are thinking to yourself, “How useful is this task? Will it help me reach any of my goals?” If the answer is that it is not useful and won’t help to get you where you want to go, you are much less likely to do the assignment. Attainment value is “the relative personal/identity-based importance attached by individuals to engage in various tasks or activities” (Eccles & Wigfield, 2020, p. 5). In other words, how much better will you feel about yourself by doing the learning activity well? When we weigh up the cost to benefit ratio of completing a learning task, we assess:
- Effort cost – the perception of how much effort it will take to complete a task and whether it is worth doing so;
- Opportunity cost- how much doing one task takes away from one’s ability or time to do other valued tasks; and
- Emotional cost -the emotional or psychological costs of pursuing the task, particularly anticipated anxiety and the emotional and social costs of failure. (Eccles & Wigfield, 2020, p. 5)
It is important to note that all of these factors can be influenced by others within your social sphere: parents, friends, classmates, etc. For example, if you hear a friend talking about how much they enjoyed a class because of how much they learned, this might increase your intrinsic value, or interest, in taking the class yourself. You may also be subject to stereotype threat where thinking about negative societal expectations of your identity (e.g., girls are not good at math, people of color are not academically successful, etc.), can lead you to believe in these stereotypes, avoiding the learning task because of the emotional cost and low attainment value (Beilock et al., 2007; Steele, 1997).
One final note about motivation: mattering matters! What you are learning has to matter to you in some way. If there is no emotional connection to the content, such as curiosity or empathy, the information will probably not be remembered. The brain is too efficient to learn something that is not meaningful (Immordino-Yang, 2015). Emotional connections to course content can provide further task value.
How can AI help?
How can you use generative AI chatbots to help with your motivation? Let’s examine some of the components of the SEVT model.
Expectation of Success: You have an exam coming up. What are your expectations for success? It would be nice to have some verification of how well you know the content so you can be more accurate in your expectation for success. Unfortunately, we often over-estimate how well we know course content because we are just familiar with the content (Brown et al., 2014; Deslauriers et al., 2019). We might be able to recognize it, but really knowing it requires you to come up with the information on your own and apply it. You can use an AI chatbot to test your knowledge of the content. Take out your study guide, then ask ChatGPT, Bing, Gemini, or Claude to prepare you for the exam. Here’s an example.
AI Chatbot Example: Study for an Exam
Intrinsic Value: You’re in college on a football scholarship and economics is the last thing you want to learn. If you can’t find a reason for wanting to learn the course content, then you are going to struggle to stay motivated in the class. Maintaining your GPA to keep your scholarship might not be enough. Ask your favorite chatbot to give you a reason to learn economics. Adjusting the prompt and settings can really help you get a useful response. Check out this example:
AI Chatbot Example: Increase Interest in Course Content
Cost: Suppose you think of yourself as “not a math person” and are now thinking about the assignment for your stats class. Could you just plug the questions into Wolfram Alpha to get the answers and submit them? Sure. Would this help you learn the content? Absolutely not. But you’re weighing the costs of completing this assignment: the effort is probably more than it’s worth, it’s going to take way too much time that you could be spending on other assignments, and the emotional cost will be high, too—you hate math and know that the whole experience will be frustrating and you’re not sure you will be able to understand the assignment in the first place. Can generative AI help? Check out this example of using a chatbot as a companion tutor to help you through a stats assignment:
AI Chatbot Example: Personalized Stats Tutor
How will AI hinder?
Anything that interferes with the motivation cycle described above can potentially hinder your motivation for learning. You may also be surprised to know that if something is too easy, in other words has no cost associated with it, you will likely have no interest in the task either. If you let generative AI do an assignment for you rather than with you, you risk lowering your motivation for continuing to learn. As mentioned above, feeling successful with learning will increase your motivation to continue learning, but if you don’t give yourself an opportunity to be successful, you will miss out on this spark.
Memory and Knowledge
What is it and how does it work?
The goal of learning is to distill new knowledge and skills from our experiences to such a level that we can then apply them in other ways. In order to do this, we need to form the memory, remember and use knowledge or skills at the right time, and be able to store these memories for long-term use. This section of the chapter breaks down each of these three steps with examples that you may see in your own learning.
Forming Memories. Forming memories happens when we have experiences that we process in some way. If you want to remember something, it will be important to first pay attention to it, then be sure to process it deeply so that the memory is formed. This is called encoding, and it actually changes your brain. Each time a memory is formed, new connections are made between brain cells. In fact, all of your memories are really just electrical signals moving throughout your brain with unique collections of brain cells activated.
Processing can happen in several ways with the most effective being thinking through how the new idea fits into what you already know. After all, “memory is the residue of thought” (Willingham, 2009, p. 54). If you are presented with new information that you cannot relate to in any way, you will likely not remember it. In this case, you can try using a metaphor that reminds you of something you do know well so you can use that structure to help build memories of the new concept.
When I mentioned the electrical signals moving throughout your brain, did you picture it? Try picturing it now. Imagine the idea of a new connection being formed between two brain cells to represent learning this concept. You can use the metaphor of an anchor and a boat connected by a chain. The anchor at the bottom of the ocean represents the knowledge you already have, like knowing that the brain is made up of clusters of little cells, while the boat is the new information about forming connections between cells to form memories. Now, imagine a bolt of electricity moving from the anchor to the boat and back again. That is your new memory of how learning happens.
If you do have some background information about what is being taught in class, then use that to help you understand the new information. Think about how it connects or contrasts to what you've learned previously. If the learning experience is connected to a strong emotion or sensory stimulus, then that helps to more deeply encode the memory. For example, if you are outraged by the injustice of some statistical information on a marginalized group, you will probably remember it.
Memory Retrieval. Once you have formed the memory, you have to be able to recall it when needed. Especially when it is time to take the test. Remember how I suggested that you try to think about new information in terms of what you already know? These can be cues to help trigger your memory. If you think about a memorable experience that you can relate in some way to the new information you are learning, then thinking about that experience can help you to recall the new information when you need it. Perhaps you are learning about the immune system in your biology class, and you are finding it difficult to remember how antibiotics interact with the cells. You remember how tense you felt watching the season finale of the Last Kingdom. Attacks were coming at Uthred’s castle from all sides. How did he defend himself? Use this metaphor to help you to understand and remember how pathogens invade the body, the components of the immune system, and their functions. Then, when it’s time to remember the role of antibiotics in the next class discussion, you can remember the Danish army coming to the rescue in the battle against the Scots.
A word of caution about analogies and metaphors is needed. They work so well to get you started, but it's important to look for where the metaphor does not match up anymore with the content you are trying to learn. Otherwise, you could be learning the wrong information, which inevitably means unlearning and relearning the correct information. Use metaphors and analogies to get you started, then be sure to recognize when it does not fit.
Every time you recall a memory, those connections between the cells get stronger, making it easier to remember in the future. Think of it like first tying a thread from the anchor to the boat, then each time the bolt of electricity runs between them, signifying the recall of that memory, the thread becomes thicker until it becomes a rope, and eventually a chain. If you can recall content without any cues or hints, then that makes the memory even stronger. This is difficult to do, but struggling during encoding and retrieval is actually good for you as it makes for stronger memory traces.
Memory Storage. It’s the beginning of the semester and you have just learned a new concept in your introduction to economics class: price and quantity controls. You think you have a good grasp on it now, but what happens in 6 weeks when it’s time for the midterm exam? Will you remember it well enough to answer questions correctly? Will you remember it well enough to be able to discuss the concept in an essay? Getting the information in your brain and out again is important for learning, but for the information to really be useful, you have to be able to store it for the long term.
One of the best ways to store information in your mind is to access it at regular intervals. If you have to struggle to remember the information, even better! Desirable difficulties like this can improve learning. Think of it like going to the gym: if the workout is effortless, you're not building muscle or stamina. While you are asleep at night, your brain begins to sift through everything it has experienced that day and in recent days, looking for connections to what you already know and trying to decide if it is worth keeping or getting rid of. There is only so much room in your skull for all of those new connections to form, so part of the maintenance of your brain each night while you sleep is to trim the connections that are not needed and clean out any other brain refuse. If you spend six weeks without thinking about price and quantity controls, your brain will clean house and that memory will likely be lost or much more difficult to retrieve. Research suggests that you should revisit this information at least 10-20% of the amount of time before you will be tested (Carpenter et al., 2012). If the midterm is in 6 weeks, then you should add a note to your calendar somewhere between six and twelve days from now to study that concept again, preferably by quizzing yourself or attempting to summarize it without notes. This will provide you with a noticeable boost in remembering the information.
If you really want to improve storage, you can use the results of an interesting experiment conducted over 100 years ago (Murre & Dros, 2015). Herman Ebbinghaus taught himself lists of nonsense words, testing himself on his ability to recall them after increasing periods of time. Unsurprisingly, the longer the gap before he tried to recall the list, the less he could remember. You can see the results in Figure 2.
Figure 2
Ebbinghaus Forgetting Curve
Fortunately, Ebbinghaus tried recalling some of the lists multiple times and found that the curve became less steep with additional efforts to recall the words until it almost disappeared. This proved that actively recalling information strengthens the memory, especially when it is spaced to occur just before the information would otherwise be forgotten. A lot of research has been conducted on spaced studying since then, and some scientists have even developed a formula to predict an optimal study schedule to retain new knowledge (Pashler et al., 2009). You can see a sample study schedule in Figure 3.
Figure 3
Forgetting Curve Modified by Spaced Recall
A word about the word “review”. This does not mean to just look at the same material again. You have probably been taught to highlight as you are reading and to re-read your notes to study. You might be surprised to find that these techniques are generally ineffective (Dunlosky et al., 2013; Leonard et al., 2021). Instead of “review”, consider “retrieve”. The more time you spend effortfully trying to remember or retrieve knowledge, the better you will remember it in the long run.
A similar process happens in your brain when you are learning a skill, whether it be mental or physical. Building the memory requires connecting the skill to one you already have, and instead of just trying to remember the skill, you actually have to use the skill. In other words, practice, practice, practice. The plus side is that these procedural memories stay in your brain for most of your life, and often just need a little brushing up when they haven’t been used for a while.
How can AI help?
Now that you know how memories are formed, let’s look at how generative AI can help in this process. You can use AI tools to assist with forming the memories, retrieving the memories, and improving the storage of memories. Here are some examples of each:
Forming Memories: You are taking Introduction to Biology and are learning the parts of the cell. All of the parts look like aliens, they have long difficult-to-pronounce names, and you’re not quite sure why you even need to learn this! Remember what we said about improving encoding? You can use generative AI to request metaphors to help you understand this foreign content, elicit emotional reasons for learning the content, and give you ideas for learning the parts of the cell through multiple sensory experiences. Here is an example:
AI Chatbot Example: Forming Memories
Retrieving Memories: You have an essay to write for a class, but you are very anxious because you “are not a good writer”. You feel completely comfortable talking about the topic, but just freeze up when it is time to put thoughts on paper. While grammar and mechanics are included on the grading rubric for the essay, the purpose of the paper is not to demonstrate your writing ability, but to demonstrate your ability to analyze the topic. Here is an example of a conversation with Claude that shows how this tool can be used to help you organize your thoughts and generate content to help you write the essay that says what is really in your mind.
AI Chatbot Example: Writing an Essay
Improving Memory Storage: The memory that is stored is only as good as the one that was initially formed and then strengthened with recall. Instead of just reading a chapter multiple times or skimming it again before the exam, make sure you understand what you are reading first, and then set up a schedule of spaced retrieval to help you retain the information over time. Here is an example of a conversation with Claude that shows how you can improve the knowledge formation and retention so that the information stays in your memory until the exam.
AI Chatbot Example: Remember Chapter Reading
How will AI hinder?
Anything that interferes with or disrupts the memory formation, retrieval, and storage system described above will lead to forgetting or more difficulty in recalling and using information you have learned. It is so easy to cut corners with AI, which can save you time and effort that you can devote to other endeavors, but you don’t want to cheat yourself out of the learning experience. If you are learning a skill, like editing a paper in your English class, don’t let AI do the work for you. Do the editing yourself first to practice the skill, then you can ask ChatGPT to edit the document to check your work.
Suppose your professor assigns some dense research articles for you to read for homework. It would be so much easier to just upload the paper to Claude and ask for a summary, as described above. However, you will be missing out on a lot of the thinking that needs to happen for the memories to be formed. A better idea is to ask Claude for the summary and some things to think about as you read the article, and then read the article yourself. You’ll actually find that it is easier to understand once you have read the simpler summary.
Solving the Problem
Martina used ChatGPT to write the essay on the importance of women in government because she just had no interest in the topic and had many other assignments due that week, too. Martina could have had a conversation with ChatGPT, Claude, Gemini, Co-Pilot, or any of the other generative AI chatbots to help her see why writing the essay was relevant to what matters in her own life. According to the Situated Expectancy Value theory of motivation, Martina did not value the task because she was not intrinsically interested, did not think it was useful, and thought the effort and opportunity cost was too high because it would keep her from completing the other assignments for her other classes. A conversation with an AI chatbot could have helped to increase her interest, find value in the assignment, and help make the process easier by helping her to organize her thoughts and provide feedback on drafts of her essay.
Greg used Co-Pilot to write the essay for a different reason: he had estimated low attainment value and high emotional cost for completing the essay assignment because of his perceived writing ability. He had thoughts about women in American government, but he just didn’t know how to organize them in a way that expressed what he was trying to say. Instead of asking Co-Pilot to write the essay for him, he could have used the tool to help him write the essay as described in the example above.
Louisa did not turn in the essay at all because of challenges with balancing work, family, and school, a common barrier for nontraditional learners in higher education. She had done the readings and outlined her essay, but she just didn’t have the time to actually write the essay. A generative AI chatbot could have been her companion during this process to speed up the process in her limited available time. Louisa could have used the speech to text feature on her ChatGPT app to dictate her thoughts, provided the outline to the chatbot, then asked the tool to help her edit the dictated draft. This would give Louisa a draft to read over and make adjustments to in order to ensure that her thoughts on the topic were well represented.
Dr. Vasquez is right to be concerned about how her students are and are not using generative AI in their learning. Without guidance, many students are unsure about when it is okay to use these chatbots, while others may use them to save time because they think they won’t get caught. Understanding the learning process can be helpful in identifying appropriate uses of these tools when completing learning tasks, whether graded or not. An AI policy statement like the one created by Dr. Lorien Lake-Correl and Dr. Torry Trust that outlines specific use cases for generative AI based on the learning process would be helpful for her students.
Discussion Questions
- Use the Situated Expectancy Value Theory to describe your motivation for an assignment you recently completed or have coming up. In the areas that detract from your motivation, how could you use a generative AI tool to help you increase your motivation?
- Consider the memory formation, retrieval, and storage process described above. Use this process to describe something that you learned last year. How could a generative AI tool be used to improve this process?
- Even if you are using generative AI in a way that helps and does not hinder your learning, there may still be other ethical issues to consider. What might some of these be?
References
Beilock, S. L., Rydell, R. J., & McConnell, A. R. (2007). Stereotype threat and working memory: Mechanisms, alleviation, and spillover. Journal of Experimental Psychology: General, 136(2), 256–276. https://doi.org/10.1037/0096-3445.136.2.256
Brown, P. C., Roediger, H. L., & McDaniel, M. A. (2014). Make it stick: The science of successful learning. Belknap Press: An Imprint of Harvard University Press.
Carpenter, S. K., Cepeda, N. J., Rohrer, D., Kang, S. H. K., & Pashler, H. (2012). Using spacing to enhance diverse forms of learning: Review of recent research and implications for instruction. Educational Psychology Review, 24(3), 369–378. https://doi.org/10.1007/s10648-012-9205-z
Deslauriers, L., McCarty, L. S., Miller, K., & Kestin, G. (2019). Measuring actual learning versus feeling of learning in response to being actively engaged in the classroom. PNAS, 116(39). https://www.pnas.org/doi/abs/10.1073/pnas.1821936116
Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students’ learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4–58. https://doi.org/10.1177/1529100612453266
Eccles, J. S., & Wigfield, A. (2020). From expectancy-value theory to situated expectancy-value theory: A developmental, social cognitive, and sociocultural perspective on motivation. Contemporary Educational Psychology, 61, 101859. https://doi.org/10.1016/j.cedpsych.2020.101859
Immordino-Yang, M. H. (2015). Emotions, learning, and the brain: Exploring the educational implications of affective neuroscience. W. W. Norton & Company.
Leonard, S., Stroud, M. J., & Shaw, R. J. (2021). Highlighting and taking notes are equally ineffective when Reading paper or eText. Education and Information Technologies, 26(4), 3811–3823. https://doi.org/10.1007/s10639-021-10448-9
Murre, J. M. J., & Dros, J. (2015). Replication and analysis of Ebbinghaus’ forgetting curve. PLOS ONE, 10(7), e0120644. https://doi.org/10.1371/journal.pone.0120644
Pashler, H., Cepeda, N., Lindsey, R. V., Vul, E., & Mozer, M. C. (2009). Predicting the optimal spacing of study: A multiscale context model of memory. Advances in Neural Information Processing Systems, 22. https://proceedings.neurips.cc/paper_files/paper/2009/hash/6bc24fc1ab650b25b4114e93a98f1eba-Abstract.html
Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and performance. American Psychologist, 52(6), 613–629. https://doi.org/10.1037/0003-066X.52.6.613
Willingham, D. T. (2009). Why don’t students like school? A cognitive scientist answers questions about how the mind works and what it means for the classroom (1st ed). Jossey-Bass.