Math comic challenge

This one is my favorite:

It is a big image

image

I thought the math thing was supposed to be for kids lol never mind also she looks like velma

Ok that made zero sense

People in places like China, that started poor and became (relatively) rich, have become taller over several generations, and have also started eating more calories per day. (In places like the US, more calories correlate with people getting wider, but in poorer countries more calories correlate with people getting taller.)

So, that's a correlation: When X gets bigger, Y gets bigger too.

So it's tempting to think that one of these caused the other. Because people get taller, they need more food to sustain themselves, so they eat more. Right?

Oh, you think it's not right? You think that because they eat more they get taller? Prove it!

Right, you can't prove it, not if all you know is that the two are correlated. To prove it, you'd have to take two groups of poor people chosen at random, give the people in one group more food, and see if they get taller. That would be an experiment, an intervention in one variable without changing anything else. (Maybe the real causative agent is some third thing: Maybe people who live in northern latitudes work harder, or something, and so they get taller from all the exercise and also eat more because they have access to more food. To avoid that possibility, you have to take two groups for whom everything else is the same, and then randomly choose half of them to get whatever intervention you think will cause the other variable to change. This is why it takes a while to prove that some medicine helps cure or prevent covid-19.)

So the first person in the cartoon is saying he used to think that if two things are correlated, one causes the other, until he took a statistics class in which the teacher said more or less what I just said, to convince the students that knowing two things are correlated doesn't tell you which causes which. (Maybe there's no causation at all involving those two variables; they just happen to both get bigger over time, as lots of variables do. Life expectancy goes up over time; population density goes up over time; literacy goes up over time; the national debt goes up over time. None of those things causes any other of them, nor is there some hidden third variable causing them.)

He used to think that correlation implies causation, until the stats teacher went through an argument like this one, and that convinced him that correlation doesn't imply causation. So the other person comments that taking the class taught him that correlation doesn't imply causality. Her observation is that taking a stat class correlates with understanding that correlation doesn't imply causality, and her theory is taking a stat class causes understanding that correlation doesn't imply causality.

She's obviously right. He took the stat class and it taught him about causality. He now understands it because he took the stat class. But, having learned some stats, he replies that it's not proven that the class taught him the idea.

They're both right. She's clearly right that it's the class that taught him the idea. But he's right that merely noticing the correlation doesn't prove the causation. What makes it funny is that her claim of causality is so obviously true, and taking the class has made him respond pedantically instead of sensibly.

Okay I didn’t understand half the things you just said

Hmm there's only so much explaining of a joke you can do before it's not funny any more, but let's try this:

Correlation means that two quantities rise and fall together. If square A has a bigger area than square B, it also has a bigger perimeter. That one's quite rigid, at least on a Euclidean plane; there aren't any exceptions. Correlations can also be looser: If family A has a bigger house than family B, they probably also have a higher income (supposing they live in the same town), but there might be a few exceptions. (There can also be negative correlations: if one quantity rises, the other falls.)

You know what causation means, even if the word is new to you; it's what people mean when they say "because." If family A has a higher income than family B, that enables them to buy and maintain a bigger house. But if family A somehow gets a bigger house without buying it (let's say their ancestors 200 years ago bought it), that doesn't give them a higher income.

When people learn about correlations, we almost always start looking for causation. If someone tells you that more poets have moustaches than engineers do (I have no idea if that's true; I just made it up) you probably respond with "why?" Or at least you think it. I can't think of any reason why having a moustache should make you a poet, so I imagine I'd guess that being a poet makes you want to grow a moustache, perhaps because you hang out with people who think moustaches are cool. That's a hypothesis about causality. It's much more satisfying as a hypothesis than "No reason, they just do."

What you learn in stats class is that correlation doesn't imply causation. They train you not to make any assumptions about something causing something else.

So the guy in the cartoon notices a correlation: He took a stats class, and he no longer thinks correlation implies causation -- if X happens and then Y happens, it doesn't mean X caused Y.

The woman in the cartoon suggests that he no longer thinks etc. because he took the stats class. She's been told about a correlation, and she thinks that behind that correlation is a causation.

He then takes what he's learned in stats class and applies it to this situation. Correlation doesn't imply causation, so just because he was taught that in stats class doesn't imply that he learned it because he was taught it. That's what he means by "well, maybe."

But what makes it a joke is that she's obviously right, not because of a statistical argument about correlation -- you can't really talk about a correlation when there's only one example to look at -- but because you learn things when you're taught them.

Does that help?

Hot off the presses:

That one's actually pretty good.

Yes but I am most likely never taking a statistics class

Has anyone done this yet?
image

behold


25 dollar handheld monster that you MUST operate with 2 hands
also does 3d graphs and has the same cpu as the sega genesis

What that video leaves out is the relatively recent discovery of how environmental history modifies our genes. I don't completely remember the mechanism, but it has something to do with the non-gene stuff in the DNA, formerly called "junk DNA" until they discovered that it isn't junk at all.

They can sell them so cheap because you have to have one of exactly that model to bring to some of the science/math SAT subject tests. I'm tempted to buy one, but I'd never actually use it, because I have a computer.

tbh best 25 bucks i had spent in 2020, thing's an absolute monster and can do multiple math equations at once when you put a buncha stuff in a list

Yeah, they worked really hard at designing it. Does the current model still provide Geometer's Sketchpad? (Or Cabri, I forget which.)

What on earth

so. much. comics.

Mine came with it, albeit i bought it for 25 bucks off of ebay instead of 300 bucks from ti howevermanyyearsago when they still officially sold it, but god, what a good buy

Did my comic get deleted? I can't find it.