Rationality for Kids?
post by goose000 · 2020-09-16T22:58:21.300Z · LW · GW · 6 commentsThis is a question post.
Contents
Answers 18 AllAmericanBreakfast 4 ChristianKl 3 G Gordon Worley III 2 James_Miller 1 goose000 1 Doc Paradox 1 kotrfa 1 Xkcd 1 kerspoon None 6 comments
UPDATE 11NOV:
I came up with a game to use as an icebreaker. And I'd love ideas for future variations. It's a combination of Credence Calibration, 20 Questions, and Taboo. The children are trying to determine which of three possible states exist on the card which I have face down (for my first iteration, the possibilities will be "Cat", "Rat", and "Dog"). Every kid gets 30 poker chips to allocate to each of the three possibilities. Kids will then take turns asking a yes or no question, but before each Q, I roll a six sided die. If it comes up six, all chips placed on a wrong answer are turned in, otherwise, they ask their question, I answer with something on a scale of "Never" to "Always", and they are permitted to reallocate their chips. But there is a catch: they are not permitted to use certain words (i.e. cat, dog, rat, meow, bark, pet, etc.) in their questions.
The point is to find tests which can serve as evidence between the possibilities and recognize how confidence should change according to evidence.
Would be interested in other possible states for future iterations
END UPDATE
So I really appreciate the lessons I've learned from "Rationality", but I wish I had learned them earlier in life. We are now homeschooling my kids, and I want to volunteer to teach my kids plus others who are interested lessons about thinking rationally.
Does anyone have recommendations on how to put together a curriculum which gets at the core ideas of rationality, but is oriented towards young kids? Some criteria:
Children will likely range from 7-11, meaning they should be simple concepts and require very little prior knowledge and only the simplest math.
Lessons should be interactive.
Lessons should include TRUE experiments (not just doing fun stuff with chemicals).
Lessons should be fun and appealing enough that parents will want to sign their kids up.
Any other suggestions on the course (wording that will be appealing without sounding too "nerdy" or alarming to the conservative types who usually homeschool) are welcome.
UPDATE: the Inflection Point Curriculum appears to be the middle school version of what I am looking to do: https://drive.google.com/file/d/1tcUJXRlZXeKjAWeU9Y37FcPKv3lj6PsX/view?usp=sharing
I currently envision the course as a combination of game type exercises like Credence Calibration, Zendo, and Meta-Forms, and experiments like adjusting the air composition of a room and investigating bernoulli effects using things like paper and shower curtains. Other ideas: investigating citrus batteries, water absorption by celery, and the light spectrum of various sources as split by a prism.
Answers
I’ve been teaching this age range for the last ten years. Knowing that the dynamic and priorities of each family are different, I hesitate to over-advise.
But I have a few areas to focus on:
- Closely observing the natural world, asking leading questions to get them to articulate details of what they see.
- Learning a bit about key scientific principles, and using them to explain the world they’re familiar with.
- Use paradox to awaken the puzzle-solving itch.
Ask them what would happen if a store started charging $100 for their favorite candy bar. Or what if the principle of their school had to buy all the kids’ school supplies and decide who gets what. Explain evolution and the basic needs of a plant, then dig up a garden weed and use those ideas to explain its structure. Ask them “When you put ice cubes in a glass of water, why does the ice always melt rather than the water turning all to ice?” Or tell them that “you said abracadabra when you got a cut on your finger and it got better a few days later, so saying abracadabra makes cuts get better,” then ask them how they know you’re wrong.
Together, this gives them practice in all the basic methods of science. Keep it light, keep it fun. When their puzzle-solver’s habit is well established, they‘ll be more likely to feel a need for the clarity that more formal methods can offer.
↑ comment by goose000 · 2020-09-17T18:30:33.142Z · LW(p) · GW(p)
I like these ideas, and you're right that these KISS type questions are good at getting at the heart of mechanisms and generalizing outside of context.
I'll mention now though, that I've been rightly advised to not disregard the flashy stuff kids like to see, because it is effective at getting them excited about science. Do you have any specific recommendations on how to take some of the classic "experiments for kids!" stuff you can find with a google search and add in a dose of "construct a falsifiable model and attempt to falsify it"? Some way I can keep the flash, but still teach them to the importance of models which allow them to make bold predictions?
Replies from: AllAmericanBreakfast↑ comment by DirectedEvolution (AllAmericanBreakfast) · 2020-09-17T20:05:48.210Z · LW(p) · GW(p)
This hasn't been my experience with kids, honestly. Nor is it my interpretation of the education literature that there's overwhelming evidence that making science education flashy is an optimal strategy. It's hard to do good science education while keeping the flash, and flash isn't the most durable emotion.
Instead, I find that kids like novelty and play, and they also like to feel capable and appreciated for their knowledge. Things don't have to be flashy to be novel and playful to children. Close observation of the world can reveal novelty to them even in things that are familiar, like the patterns on a garden spider's back, or seeing a rainbow in water sprayed from a hose. Making a routine practice of pointing out these phenomena, asking them about it, and making exploring the world in this manner a part of your relationship is the way I would approach things.
I question whether making young kids invent falsifiable models and do controlled experiments is really the best way to kick off their science education. Science education, even in college, is far more about having them read about other people's discoveries and observing the world closely than it is about lab work. Undergraduates rarely if ever invent their own experiments or models.
Nothing wrong with having your kids do an experiment here and there if it's fun. But if it were my own children, I'd have them peer through telescopes, look at bugs with a magnifying glass, follow an ant to see if they can find its nest, build a Halloween costume that incorporates some home-made electronics, learn to program a computer game, help you cook a recipe that requires them to double all the proportions of the ingredients, and other things like that. I earnestly believe that the desire to analyze the world follows from a habit of observing it closely.
Replies from: goose000↑ comment by goose000 · 2020-09-18T14:48:44.728Z · LW(p) · GW(p)
Yes, I agree that doing good science is hard with flash, I've just had everyone telling me that that's what hooks them. Good to know that's not really true.
I'm thinking along the lines heavily leading to/giving the model, not necessarily having them come up with it themselves and then testing it. But part of the reason I'm asking here is to see if anyone has ideas regarding models which are discoverable by kids this age so that they can get there by more of their own processes.
Replies from: AllAmericanBreakfast↑ comment by DirectedEvolution (AllAmericanBreakfast) · 2020-09-18T15:13:46.931Z · LW(p) · GW(p)
That’s fair! I think that’s a good idea to explore and I think it’s great to try things out. If you try something and the kids don’t take to it, no harm done :)
One thing you could try is some probability. There’s a classic intro stats demo where you have a class come up with fake sequences of 20 coin flips in a row, and generate some real sequences of 20 coin flips as well, all while the teacher is out of the room. Then the teacher comes in and guesses which are real and which are fake.
They can do that because people tend to generate fake sequences with too few stretches of repeated heads and tails.
Kids can flip a coin, they’d have fun trying to trick you, and when you guessed right, it might seem like a magic trick. You can also teach them a few things about probability and dice rolls and help them see how it applies to board games.
Replies from: goose000↑ comment by goose000 · 2020-09-19T19:37:50.945Z · LW(p) · GW(p)
I like the coin flip idea. I have done something along these lines as a single session with homeschool kids where I gave them two decks of cards and had them stack the deck while I was out. When I came back I used an Excel VBA program I had made to continually reassess the maximum likelihood for the red/black proportion and updated it as I drew cards. Didn't go quite as well as I had hoped, mostly because I didn't emphasize that in order to get quick results they needed to really stack the deck, and they had made it 24 red, 28 black, or something similar.
Anyway, yes, I was thinking exploring probability might have some more possibilities along these lines, so I will think about that a little more. We did optical illusions today: persistence of vision, pattern juxtaposition, etc. Then we talked about how they fool system 1 thought, but you can use system 2 techniques to defeat them, did things like measuring the apparently converging lines, slowed down the thaumatrope, etc.
Random thoughts:
Focusing, Internal Double Cruz and Belief reporting are powerful tools for aligning system I and system II. They aren't very complex and have relatively few moving parts. It might be possible that you can teach those to children but I could also imagine that it's hard to teach them. Dealing with groundlevel thinking is intimite and it needs buy-in from the child.
Zendo seems to be a game that can be fun for children and it teaches valuable lessons about building scientific hypothesis and testing them.
Credence calibration doesn't seem very complex but might be too much for a 7/8 year old. I'm unsure at what age the necessary understading of numbers between 1 and 100 exists.
I could imagine a setting where you talk with children about what they are curious about and what might be tested by experiment. Then the children gather data collaboratively and you do the math for them in Numpy.
If you have a humidity, temperatur and a CO2 sensor you can for example have experiments about how much opening the window affect the air in the room. Does open the window fully for 5 minutes do more then have it half open for 10 minutes?
Opening the windows to get better air is a task that's relevant to daily life and it's not a question where you find the answer in normal textbooks.
↑ comment by Sandi · 2020-09-17T11:31:30.362Z · LW(p) · GW(p)
Here's an Android game that works like Zendo but has colorful caterpillars, might be great for kids: https://play.google.com/store/apps/details?id=org.gromozeka1980.caterpillar_logic
Replies from: ChristianKl, tomcatfish↑ comment by ChristianKl · 2020-09-17T11:43:12.698Z · LW(p) · GW(p)
When teaching it to a group of children I would likely not do it via an app but with physical items and drawings. It makes sense to switch the kind of items with which you play it between sessions to increase the generalizability of the learning.
↑ comment by Alex Vermillion (tomcatfish) · 2020-11-11T21:26:37.944Z · LW(p) · GW(p)
I'm 2-3 times the age of the students involved and I love that game enough to recognize it from the name in the link. I highly recommend using this or even a version with beads to teach kids.
Explanation on the beads comment:
I imagine a game where you allow kids to put beads on a string and each strand is finished with either a gold or silver bead depending on whether it passes or fails. Tie them off and let them arrange them however they need to see the patterns.
I would play this bead game.
Further exploration: Try the bead game with Lego, which seems obvious in retrospect but might have flaws I don't see.
↑ comment by goose000 · 2020-09-17T14:04:13.211Z · LW(p) · GW(p)
Well, my first thought is that I need to spend some actual time on this site (I had to look up most everything you mentioned); Most of my education has simply come from Yudkowsky's book/compilation.
Zendo definitely looks promising, and should definitely be an element of the course as well as something I play with my kids. As I envision the course, however, it would be an element such as a warm up or cash out, not the core curriculum.
My thoughts on Credence Calibration are similar to my thoughts on Zendo with the following modifications: each kid would be given ten poker chips, we would play the 2 statement variant (at least initially), scoring would be simplified to the liar keeping all poker chips bet on his lie, winner would be the one with the most poker chips at the end.
Focusing and Internal Double Crux seem like they would be pretty hard to teach to elementary age children. Focusing mostly because it seems like it would require one-on-one instruction, at least initially.
Unfortunately, I do not have much instrumentation. I could buy inexpensive things, so a thermometer and a humidity sensor would be doable, but it seems like a worthwhile a CO2 sensor might be a little more (based on my brief look on Amazon). I really do like the idea of the experiment though.
Some experimentation ideas I have received: investigate air pressure changes created by shower water spray as measured by a shower curtain (also blowing over pieces of paper, blow dryer, etc.) and electricity produced by a lemon battery, water drawn through celery including dyed water, spectrum differences in light sources as shown by a prism.
Replies from: ChristianKl↑ comment by ChristianKl · 2020-09-17T21:10:02.550Z · LW(p) · GW(p)
As I envision the course, however, it would be an element such as a warm up or cash out, not the core curriculum.
Yes, however it's worth noting that you can play Zendo in different ways. The time I played it was with arrangement of Lego stones.
You could also play it with other domains like words or sentences. I'm uncertain about how much time a child can effectively learn something from Zendo.
At the same time it won't fill the full curriculum.
Focusing mostly because it seems like it would require one-on-one instruction, at least initially.
You didn't really speak about how many children you actually want to teach at one time.
When it comes to teaching rationality to adults there's the CFAR handbook. If you are not aware of it, it might be worth looking through it and thinking about what you can teach children.
Replies from: goose000I can't find it, but I vaguely recall Julia Galef writing something about how her parents raised her and her brother such that they fit naturally with the Rationalist community, even though it didn't exist at the time of their upbringing.
Consider the board game Metaforms. It requires you to solve logical puzzles based on colors, shapes, and position.
https://www.amazon.com/FoxMind-Meta-Forms-Puzzle-Solving-Brain-Builder/dp/B0015MC2TO
Quick update: I came up with a game to use as an icebreaker. And I'd love ideas for future variations. It's a combination of Credence Calibration, 20 Questions, and Taboo. The children are trying to determine which of three possible states exist on the card which I have face down (for my first iteration, the possibilities will be "Cat", "Rat", and "Dog"). Every kid gets 30 poker chips to allocate to each of the three possibilities. Kids will then take turns asking a yes or no question, but before each Q, I roll a six sided die. If it comes up six, all chips placed on a wrong answer are turned in, otherwise, they ask their question, I answer with something on a scale of "Never" to "Always", and they are permitted to reallocate their chips. But there is a catch: they are not permitted to use certain words (i.e. cat, dog, rat, meow, bark, pet, etc.) in their questions.
The point is to find tests which can serve as evidence between the possibilities and recognize how confidence should change according to evidence.
Would be interested in other possible states for future iterations.
Hi everyone. I wanted to join LW years ago to do this work with kids. Please check out The Paradox Lab www.paradoxlab.org to see our mission and projects. Right now we are only doing online discussion classes, mostly using philosophical inquiry to develop critical thinking skills. We are a nonprofit and all the classes are pay-what-you-want. Please share. And please join our team!
I am so glad this question is here, as it's very relevant to my post a few weeks back about Effective Children Education [LW · GW].
By the way, I recommend following Duncan Sabien (referenced in the post below) on Facebook, he has good posts about children edu, e.g. his speech for sixth-graders (referenced by someone else here - but she picked the good parts).
As mentioned below, Julia Galef also sometimes mentions something related, but I haven't found much
FWIW I wanted to plug an older thread on this topic: https://www.lesswrong.com/posts/wSEGsjDPtSWkAtXce/teaching-rationality-to-kids [LW · GW]
I'm really interested in this too. I have a 1 year old and work in improving engineering education.
https://en.wikipedia.org/wiki/Philosophy_for_Children might be worth checking out.
6 comments
Comments sorted by top scores.
comment by mingyuan · 2020-09-17T06:17:02.587Z · LW(p) · GW(p)
More of a meta-suggestion but you should contact Duncan Sabien, who formerly worked for CFAR and has always been really enthusiastic about teaching middle schoolers.
Replies from: ben-schwyn, goose000↑ comment by Ben Schwyn (ben-schwyn) · 2020-09-17T08:21:18.711Z · LW(p) · GW(p)
He used to have a nice website that had lessons or philosophy for his middle school class (Thingmakery?) but I can't find it with google.
Replies from: Davis_Kingsley↑ comment by Davis_Kingsley · 2020-09-17T13:48:57.251Z · LW(p) · GW(p)
It's still around: https://thirdfoundation.github.io/#/main
I haven't read everything on the site and do not necessarily endorse it all, but the Inflection Point curriculum document might be especially relevant.
Replies from: goose000