Motivated by the Answer

I was motivated to join this program by my love-hate relationship with instructional technology. I love the potential that ed tech offers but I have yet to find a software tool that can teach students with significant learning impairments in the way that they need to be taught (i.e., with the sensitivity to students’ cognitive and emotional needs that an experienced teacher might have). I hope to create a game-based learning project, ideally an actually usable and enjoyable game that would encourage kids to learn something (I am leaning toward math or physical science in my head). I honestly do not play a lot of games but I used a lot of games when I taught kids (mostly games that I made up or that the kids, with some guidance, came up with themselves). Games work, better than any other method of instruction in my opinion (assuming that the the game is well designed & simple enough for kids to play while still learning stuff) but I am nervous about my journey into this project because 1) I am not a gamer and 2) I do not enjoy coding (unlike solving a difficult physics problem, editing a script and seeing it run as intended just is not very cathartic, I hope to bypass any need to code if possible). However, I do love teaching and I am confident in my ability to teach and scaffold instruction. I aspire to come up with a perfect scaffolding algorithm but I do not have a ton of faith in machine learning for instructing kids. When I taught “cognitively impaired” students, the software would typically punish them for not knowing how to accomplish certain tasks, and would not allow them to move forward into content unless they “proved themselves” by accomplishing single tasks. For instance, this one math software was supposedly designed to teach kids who were behind and so it tested skills that students should have mastered while at the same time attempting to catch them up to their actual grade level (i.e., imagine a middle to hs school level but with a lot elementary questions in it at the start). It did not work out in my opinion. If you messed up subtraction, you would never make it to fractions. A few of my students struggled (and maybe will always struggle) with subtraction problem that require a lot “borrowing and carrying of the one” but these same exact students were able to solve higher level problems. Subtracting large numbers is a skill that should be mastered between 2nd to 3rd grade while being able to create & manipulate equivalent fractions is a middle school skill. I wish that a computer could figure out when its time to “let it go”. Many students get trapped in a “loop of not getting the correct answer” and the system does not know when to just let it be and move on to the next lesson. A teacher would know when they have hit an instructional wall and be able to let it go and come back to it later (i.e., another day in the week). For example, the teacher would know that a student who struggles with subtracting large numbers should not be prevented from learning other material in the content area; that the student can still learn how to cross multiply in order to create equivalent fractions or solve other math problems that do not require subtraction. The “wrong answer loop” that kids sometimes find themselves in when working with educational software can be painful and can destroy their confidence. I witnessed it with my former students. This is also where game-based learning (and by this I actually mean gamified instructional software which most of it is, rather than a pure game) fails to motivate struggling students. Repeatedly getting the wrong answer (i.e., failing to pass a level) is a good reason to not want to play (it is why I quit. Teachers, unlike algorithms, can make decisions that take into account the full reality of the instructional experience. Teachers can see where a child is struggling to the point that it is best to stop teaching and teachers can assess when a good time is to have the student try again. I am not convinced that we can train algorithms to make these sorts of decisions but maybe in the future by way of complex AI, this may happen. I wish that I could take all of my teaching experience and upload it into a super computer that would be able to teach for me and would be able to give the proper amount of guidance, motivation, and mini study-breaks, needed to encourage struggling students to keep trying and eventually learn. Yet, I doubt that a machine learning algorithm (no matter how much data is used to teach it about a student’s learning behavior) could discern the cognitive and emotional needs of students who are struggling with content. I have not figured out the steps that are needed to make my great game idea into a reality yet but I am expecting that my “great idea” will have to be broken down into an “okay idea” in order to ensure that I finish this program on time (and because I do not think robots are quite ready to fully take over all of our jobs). I do believe that instructional tech can be highly effective at “teaching content” but I think it strongly favors the self-motivated student who has not struggled too much and therefore does not need much more than a “ding ding ding you got the correct answer” to remain motivated to try. I would like for my project to focus on the other type of student but I realize this is a tall order so we shall see what happens as I progress through the ITP program.

4 thoughts on “Motivated by the Answer”

  1. Hello Shpresa! Your post is very thought-provoking and full of your passion for teaching! Yeah, it’s truly painful and frustrating that educational games can be so “stubborn” and inflexible in scaffolding the instruction. The stringent hierarchical progress might not apply to every student, especially those with special needs. I tried to figure out how we could design an educational game less strictly relying on “correct answers”, but unfortunately, I don’t have an answer. I think educational technology nowadays oftentimes lack humanistic care for students in many ways, and hopefully, this can be fixed in the future with the development of AI and machine learning, but at this moment, teachers are irreplaceable. Last, good luck with developing your project!

    1. Hi Ming, thank you for your feedback and kind words. I agree with what you said “I think educational technology nowadays oftentimes lack humanistic care for students in many ways” but it is funny to me because it seems like games seem more humanistic than educational software. In non-educational game meant for little more than entertainment, there is often rich emotional content and I think this is what education games lack that regular games often have. The way that the emotional aspect is developed and enriched makes a game more engaging than the elation of just having the correct answer. It makes me think a lot about the external vs internal motivation and I wonder if designers of educational software think about this distinction much when they design. I am sure that the typical game designer must be considering these distinctions more deeply because games are so incredibly popular, they must be triggering both external and internal motivation of players (this could possibly explain the longevity of games’ popularity anyhow). However, I do wonder about whether in the future a special algorithm could be as good as at least a caring novice teacher, I definitely think AI has potential for increasing instruction and possibly alleviating some burden from the teacher but to design the algorithm that is not ultimately focused on “correct answers” may be too much to ask of a machine. Then again it all depends on the data we feed it so perhaps in the future if we have repositories of mistakes that are not as common, an AI can help an atypical student- certainly a machine can predict student/ user behavior but being able to understand the individual human logic behind the behavior is something I do not think a computer can be trained to do (but maybe with enough relevant data in the future it is possible). Are you familiar with the IA TA Jill Watson, ? That experiment really makes it apparent that we still need real human teachers but maybe not for every instructional purpose, it is an exciting and somewhat scary thing to think about.

  2. Thanks so much for this provocative post, @Shpresa. It seems as though you are convinced— reasonably so, imo—that game-based educational technologies that rely on algorithms cannot address the needs of all learners. I would urge you to build your inquiry by thinking about approaches that could be accessible to all learners. What are the characteristics of these games? What are some models that derive from those characteristics (like how fact retrieval -> Trivial Pursuit; strategic accumulation -> Settler of Catan; facility with words -> Boggle)? By thinking about the experiences you want the player you have (you’ve already identified the experiences you want them not to have, you might be able to arrange some elements that you can shape into a form as an initial step.

    Looking forward to seeing this develop!

    1. Hi Luke, thank you for the suggestions, I will do my best to approach the 1 steps you recommend. I feel like I keep finding myself bouncing between going too simple vs getting too complex. It is hard to determine in terms of scope, what is reasonable but I do think that educational technologies that rely exclusively on “correct answer algorithms” cannot address the needs of all learners but a part of me still hopes that there must be a “golden algorithm” out there that can better accommodate real student needs. I will happily settle for a game that motivates a “C-” student to do homework consistently if I can manage to bring life to that idea.

Leave a Reply

Your email address will not be published. Required fields are marked *