Just saw this mentioned on social media, and it gave me enough of a giggle that I thought it'd be worth sharing here.
https://www.tylervigen.com/spurious-correlations
It's a website that has a bunch of graphs with spurious correlations (essentially, two variables that look related but aren't). Here are a few of my favorites so far:
Hilarious...I don't need that rabbit hole....
In reply to Beer Baron 🍺 :
That one's the classic, isn't it?
Only on GRM...
Recon1342 said:Only on GRM...
I think that's how I already know about this in the first place
I was expecting sandwiches, due to recent trends.
Sadly
In reply to RevRico :
Well now I'm not sure if the relationship between sandwiches and myself is spurious...
But you also get data of events that actually *are* related, but not causally.
NickD said:
What I got from that one is that the cheese marketing board is probably very happy with a 10+% increase in per capita cheese consumption between 2000 and 2007.
Yep, I teach an eight day crash course on engineering statistics and I pull up this site about every other month when I get to the "seven ways to do bad science" part of my lectures.
Correlation does not prove causality...it can be used to predict but not necessarily control.
Keith Tanner said:NickD said:
What I got from that one is that the cheese marketing board is probably very happy with a 10+% increase in per capita cheese consumption between 2000 and 2007.
I didn't even notice that. Damn, the early 2000s were really a strong era for cheese!
You know, the more I eat Chipotle, the more I wonder about going into outer space. I see a correalation.
Correlation doesn't imply causation, but it does waggle its eyebrows suggestively and gesture furtively while mouthing 'look over there'.
In reply to wearymicrobe :
See also...
RX Reven' said:In reply to wearymicrobe :
See also...
I spend a good chunk of time working with a predictive AI group. trust me I KNOW>
In reply to wearymicrobe :
Whenever I demonstrate the Box-Cox transformation to my students I hear George Carlin's old skit about the seven words you can't say on TV.
Had a stats prof in college who rolled out that there was a direct correlation between the birth and suicide rates in New York City.
Of course it wasn't a causal relationship... maybe
prodarwin said:
Definitely can't get killed by geese on other planets. Just saying.
You'll need to log in to post.
