Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Is watching the 1984 Ghostbusters movie killing people?

covid-datascience.com

Post reply on HN
Print view View original

231–240 of 248 posts

Re: Is watching the 1984 Ghostbusters movie killing people?

#231

Post by harry8 »

ohwellhere wrote: post #191
calibas wrote: post #160

Earlier quoted context omitted.

I don't get this attitude, if things are correlated, it should at least make a scientist wonder why. It certainly could be random chance, but correlation can also lead to establishing a causal model or discovering a third variable. If two things keep happening in conjunction, it at least merits further investigation. It seems like there's this extreme reaction against people behaving like correlation equals causation…

I share this criticism. It's almost like you have to scream at people that although correlation doesn't imply _direct_ causation, it most certainly does imply some causal chain. Ruling out chance through significance and power, what phenomenon in the world is correlated without being causally related _somehow_?

Take the worst intersection in the country this week for traffic accidents. Close the roads to the intersection at 3am and organise a contemporary expressive dance performance to rid the intersection of its evil and re-open it 30 minutes later.

Repeat every week with the worst traffic intersection for accidents.

What you find is. IT WORKS! HURRAH! These intersections are more than usually not the worst the following week! The evidence is clear. The correlation is utterly compelling. It is significant. It has power. How could it possibly be unrelated?

Now if we stop it being comically silly in our example and make it a red light & speed camera, see how the issue is much more difficult. There is a clear line of potential causation of fixing dangerous intersections. But is it really better than folk dancing for 20 minutes at 2am? [1]

[1] This example should not be interpreted as being an opposition to all red light and speed cameras.

Re: Is watching the 1984 Ghostbusters movie killing people?

#232

Post by candeira »

My explanation of the base rate fallacy is that most car crashes are caused by people who hold a valid driver's permit. The number of crashes caused by unlicensed drivers is negligible. Therefore, we should stop issuing driving licenses.

Most people I explain this to recognise the fallacy, even if they're not trained in statistics.

A huge problem with vaccine denialism is that it stems from motivated thinking. People aren't being persuaded by bullshit arguments; rather, they already agree with the conclusions, and look to these spurious correlations as emotional support.

Re: Is watching the 1984 Ghostbusters movie killing people?

#233

Post by jeremysalwen »

hedora wrote: post #169
josho wrote: post #125

Earlier quoted context omitted.

Care to explain what you mean?

Not the person you are responding to, but I challenge you to find a single COVID related article making the social media rounds that demonstrates a reasonable understanding of conditional probabilities, and presents data in a clear, level-headed, unbiased fashion. For instance, this article has a silly example of confounding variables (which ghostbusters you watched as a child is correlated to your age, and age is co…

> The first graph specifically controls for *age*, vaccination rates, and size of population.

> Age 10-59

Hmmmm....

Re: Is watching the 1984 Ghostbusters movie killing people?

#234

Post by makeitdouble »

lordnacho wrote: post #10
makeitdouble wrote: post #6

Another Spurious Correlations [0]. The message will always bear repeating. [0] https://tylervigen.com/spurious-correlations

No, most of those are "actually spurious" in the sense of the two time series having nothing to do with each other. A lot of them are pretty short series as well, so it's easy for them to be correlated. What we are looking at is Simpson's Paradox, where the true causal relationship is obscured by information that isn't obvious from the plot. Now before you correlation != causation, there is actually a causation here…

It's still interesting to note that we assume "nothing to do with each other" with some cases, "obscured" relationships on other like the OP's second point.

Where it falls relies on the viewer's knowledge of the problem space, which can also be limited enough to lure them into false causations.

My point would be that a single graph showing two trends without any further info should never be taken as more information than "there is two trends". You'll still be free to decide there is true causality based on other information you believe.

Re: Is watching the 1984 Ghostbusters movie killing people?

#235

Post by rg111 »

calibas wrote: post #160
jkhdigital wrote: post #75

Earlier quoted context omitted.

Statistics are meaningless without a rigorously examined causal model of the phenomenon under investigation. In my experience of statistics education, the art of crafting causal theories was scarcely addressed.

I don't get this attitude, if things are correlated, it should at least make a scientist wonder why. It certainly could be random chance, but correlation can also lead to establishing a causal model or discovering a third variable. If two things keep happening in conjunction, it at least merits further investigation. It seems like there's this extreme reaction against people behaving like correlation equals causation…

> if things are correlated, it should at least make a scientist wonder why

Not really.

See: https://www.tylervigen.com/spurious-correlations

Re: Is watching the 1984 Ghostbusters movie killing people?

#236

Post by calibas »

rg111 wrote: post #235
calibas wrote: post #160

Earlier quoted context omitted.

I don't get this attitude, if things are correlated, it should at least make a scientist wonder why. It certainly could be random chance, but correlation can also lead to establishing a causal model or discovering a third variable. If two things keep happening in conjunction, it at least merits further investigation. It seems like there's this extreme reaction against people behaving like correlation equals causation…

> if things are correlated, it should at least make a scientist wonder why Not really. See: https://www.tylervigen.com/spurious-correlations

You should still be curious as to why those correlations exist. It's still important even when the reason is p-hacking, since then you can identify dubious statistics.

Re: Is watching the 1984 Ghostbusters movie killing people?

#237

Post by raxxorrax »

tejohnso wrote: post #28
nanna wrote: post #12

Nice example of 'correlation does not imply causation' and if you submit this kinda thing to me in my data analysis class I'll fail you.

Makes me cringe a little bit when global warming presentations start out with a co2 vs global temperature chart. Especially if that forms the entire basis of the analytic part of the presentation. Because I know someone (most) is just thinking "correlation != causation" and dismissing the entire thing.

But in this example we know that there indeed is a correlation between greenhouse gases and temperature. Of course you cannot restrict your argument to that, but at some point you start with some axioms.

Re: Is watching the 1984 Ghostbusters movie killing people?

#238

Post by raxxorrax »

morsch wrote: post #80
everybodyknows wrote: post #71

Earlier quoted context omitted.

Source? The long temperature graphs I've seen look quite irregular, rather than cyclical.

Here's the smoking gun graph he's hosting: http://donath.org/Photos/TempChange.PNG This is what he's talking about: https://en.wikipedia.org/wiki/Milankovitch_cycles Here's a graph similar to his, hosted on German Wikipedia: https://de.wikipedia.org/wiki/Milankovi%C4%87-Zyklen#/media/... Nobody brings it up, because scientists don't believe the observed changes in temperatures are due to Milankovitch cycles, since gl…

Scaling the time axis sensibly will quickly reveal that the rising temperature and this cycle cannot explain the rise we see.

There are certainly overlapping effects, but the model would fail to predict temperatures.

Post reply on HN