Personally, I think that online learning is great for delivering *content*, that is, information about things. Do you want to learn how to use trigonometric formulas, or how to build websites, or ancient history, or a zillion other things? Then screen-based learning can deliver that information very efficiently, and when it's done well it's better than anything that a live human can achieve.
The problem is that basic education in mathematics (say, K-8) is not really about mastering a large body of knowledge. Instead, it's about developing basic reasoning skills and fundamental pattern recognition, and this is hard to foster with any computer-based interaction. (It's hard to foster with any pedagogic method at all!) I think this is the underlying issue: that computer-based learning can be very powerful for certain things, but those things aren't the first-order concerns when teaching basic math to a general student population.
I don't know, I think that assumes a level of motivation that is not there for every student. Attention is hard to sustain when getting students started with a new topic. I think humans can play a role there that is very different from a computer. I have also found that tech folks often defer to fancy videos or animations for beginners when they would be better off with some simpler examples and non-examples, stuff much more straightforward than typical tech-based delivery.
I don't totally agree...K-8 mastery depends upon answering lots of high-quality questions. Think of the times table, fractions, basic algebra -- you certainly want students to understand concepts, but you also want them to develop fluency. Answering lots of targeted questions really helps students build fluency. I'm not discounting the importance of reasoning/pattern recognition, but both rely upon a steady stream of good questions.
Dylan, this is an interesting issue and one I've been dealing with for many years. I do think it is possible to deliver instruction online that is effective in improving outcomes. However, as you've pointed out, the design of the instructional process is the key. Rather than explain my approach, I'll let my program deliver the message here: https://mathinsight.com/number_bonds_tm. When you log in, you will see a menu with some activities highlighted in yellow. These are active for you to access. (The others will require a paid subscription to be offered in the near future.) If you'd like to take a look, I would suggest that you complete an entire activity so you can see the teacher report at the end, along with a "reward" animation. This is typically implemented in Google Classroom. The teacher controls which activities are assigned to each (or all) students. The report shown at the end of the activity is automatically uploaded to the teacher's data file. That allows the teacher to review student performance and make assignments based on performance, which is important for keeping the teacher in the process. Any feedback (pro or con) is welcome!
Screens definitely have disadvantages. Students bring all their habits and expectations from previous screen experiences to new ones, and many of these are bad in a classroom: passive entertainment, isolation, an avoidance of difficult thinking.
It's also worth noting that the grocery truck analogy can be turned around in your favor: Suppose you were shopping at a farmer's market with your nutritionist -- would you buy different food than if you were ordering UberEats while watching Monday Night Football with your friends?
But the computer behind the screen can also do things that teachers can't. The most important is question selection. It can track every question a student has ever answered. It can attach psychometric properties to these questions based on all the student data. And it can use this information to eliminate questions that are too easy (boring) or too hard (frustrating) while finding questions that will fill gaps and strengthen weaknesses. A student who struggles with the times table may only be missing ~10 facts, but if these aren't targeted efficiently, the student may take a long time to master them. This delay will in turn frustrate the attempts to acquire many other skills. On the other hand, if you do figure out these questions, you'll increase engagement, without any gamification whatsoever. But teachers are managing so many students -- it's impossible to know what these high priority questions are for every student at every moment.
I do think you could modify your app so that students could only beat it by mastering the material you prioritize (I've been able to do it with mine). And it's not totally true that kids don't learn anything when playing video games -- in some games, they can only advance by learning about the game's world and manipulating it accordingly. And I think you can make math versions of this.
But that doesn't mean you have to passively hand your teaching over to the screens. You could have them work in groups on games. You could have them work individually but circle the class constantly to monitor/help. You could even have zero screens -- have the computer generate student-specific worksheets, then upload the data to the computer so that it can grade them and generate the next worksheets.
I agree with everything here. One of my constraints when I made my app was that I didn't have the skills to build a Google login integration with persistent memory for each user. (Maybe I could vibe code it with LLMs today? Not sure how hard or expensive it is.) Lots more would be possible with that. There are definitely advantages like you describe, and I appreciate everyone who is building tools that cater to those advantages and let the paper and pencil work come in when it is superior. The idea that obviously everything is better on a computer is just obviously wrong with the tools we have right now.
Love your content! When are you getting an offer from Alpha school? 😅
Those folks don't like me very much...
Personally, I think that online learning is great for delivering *content*, that is, information about things. Do you want to learn how to use trigonometric formulas, or how to build websites, or ancient history, or a zillion other things? Then screen-based learning can deliver that information very efficiently, and when it's done well it's better than anything that a live human can achieve.
The problem is that basic education in mathematics (say, K-8) is not really about mastering a large body of knowledge. Instead, it's about developing basic reasoning skills and fundamental pattern recognition, and this is hard to foster with any computer-based interaction. (It's hard to foster with any pedagogic method at all!) I think this is the underlying issue: that computer-based learning can be very powerful for certain things, but those things aren't the first-order concerns when teaching basic math to a general student population.
I don't know, I think that assumes a level of motivation that is not there for every student. Attention is hard to sustain when getting students started with a new topic. I think humans can play a role there that is very different from a computer. I have also found that tech folks often defer to fancy videos or animations for beginners when they would be better off with some simpler examples and non-examples, stuff much more straightforward than typical tech-based delivery.
I don't totally agree...K-8 mastery depends upon answering lots of high-quality questions. Think of the times table, fractions, basic algebra -- you certainly want students to understand concepts, but you also want them to develop fluency. Answering lots of targeted questions really helps students build fluency. I'm not discounting the importance of reasoning/pattern recognition, but both rely upon a steady stream of good questions.
Dylan, this is an interesting issue and one I've been dealing with for many years. I do think it is possible to deliver instruction online that is effective in improving outcomes. However, as you've pointed out, the design of the instructional process is the key. Rather than explain my approach, I'll let my program deliver the message here: https://mathinsight.com/number_bonds_tm. When you log in, you will see a menu with some activities highlighted in yellow. These are active for you to access. (The others will require a paid subscription to be offered in the near future.) If you'd like to take a look, I would suggest that you complete an entire activity so you can see the teacher report at the end, along with a "reward" animation. This is typically implemented in Google Classroom. The teacher controls which activities are assigned to each (or all) students. The report shown at the end of the activity is automatically uploaded to the teacher's data file. That allows the teacher to review student performance and make assignments based on performance, which is important for keeping the teacher in the process. Any feedback (pro or con) is welcome!
Screens definitely have disadvantages. Students bring all their habits and expectations from previous screen experiences to new ones, and many of these are bad in a classroom: passive entertainment, isolation, an avoidance of difficult thinking.
It's also worth noting that the grocery truck analogy can be turned around in your favor: Suppose you were shopping at a farmer's market with your nutritionist -- would you buy different food than if you were ordering UberEats while watching Monday Night Football with your friends?
But the computer behind the screen can also do things that teachers can't. The most important is question selection. It can track every question a student has ever answered. It can attach psychometric properties to these questions based on all the student data. And it can use this information to eliminate questions that are too easy (boring) or too hard (frustrating) while finding questions that will fill gaps and strengthen weaknesses. A student who struggles with the times table may only be missing ~10 facts, but if these aren't targeted efficiently, the student may take a long time to master them. This delay will in turn frustrate the attempts to acquire many other skills. On the other hand, if you do figure out these questions, you'll increase engagement, without any gamification whatsoever. But teachers are managing so many students -- it's impossible to know what these high priority questions are for every student at every moment.
I do think you could modify your app so that students could only beat it by mastering the material you prioritize (I've been able to do it with mine). And it's not totally true that kids don't learn anything when playing video games -- in some games, they can only advance by learning about the game's world and manipulating it accordingly. And I think you can make math versions of this.
But that doesn't mean you have to passively hand your teaching over to the screens. You could have them work in groups on games. You could have them work individually but circle the class constantly to monitor/help. You could even have zero screens -- have the computer generate student-specific worksheets, then upload the data to the computer so that it can grade them and generate the next worksheets.
I agree with everything here. One of my constraints when I made my app was that I didn't have the skills to build a Google login integration with persistent memory for each user. (Maybe I could vibe code it with LLMs today? Not sure how hard or expensive it is.) Lots more would be possible with that. There are definitely advantages like you describe, and I appreciate everyone who is building tools that cater to those advantages and let the paper and pencil work come in when it is superior. The idea that obviously everything is better on a computer is just obviously wrong with the tools we have right now.
You can definitely build the Google login with the help of an LLM. Happy to help if you get stuck.