|
On September 23 2014 12:15 Nyxisto wrote:not automatically, but the way the people use the "fallacy" on the internet is idiotic, too. Obviously two things correlating strongly may indicate that there is a causal relation between them, that's why we make experiments.
its not that obvious, there could be a third hidden factor that correlates with the two observed factor but the two observed factors are not in fact impacting each other, you just dont see that third factor and are tricked into it.
|
On September 23 2014 12:32 Sub40APM wrote:Show nested quote +On September 23 2014 12:15 Nyxisto wrote:On September 23 2014 05:06 SoylentGamer wrote:
does correlation = causation? not automatically, but the way the people use the "fallacy" on the internet is idiotic, too. Obviously two things correlating strongly may indicate that there is a causal relation between them, that's why we make experiments. its not that obvious, there could be a third hidden factor that correlates with the two observed factor but the two observed factors are not in fact impacting each other, you just dont see that third factor and are tricked into it.
Yes, but the correlation=! causation stuff that's repeated so often implies that there seemingly is no connection between the two, which is wrong. If person x punches person y in the face ten thousand times, and it hurts ten thousand times, no body would argue that this correlation doesn't imply anything. You can't prove causation, you simply repeat an experiment so often until you are confident to say: "okay now the chance that the correlation happened by accident is really, really small".
|
On September 23 2014 12:39 Nyxisto wrote:Show nested quote +On September 23 2014 12:32 Sub40APM wrote:On September 23 2014 12:15 Nyxisto wrote:On September 23 2014 05:06 SoylentGamer wrote:
does correlation = causation? not automatically, but the way the people use the "fallacy" on the internet is idiotic, too. Obviously two things correlating strongly may indicate that there is a causal relation between them, that's why we make experiments. its not that obvious, there could be a third hidden factor that correlates with the two observed factor but the two observed factors are not in fact impacting each other, you just dont see that third factor and are tricked into it. Yes, but the correlation=! causation stuff that's repeated so often implies that there seemingly is no connection between the two, which is wrong. If person x punches person y in the face ten thousand times, and it hurts ten thousand times, no body would argue that this correlation doesn't imply anything. You can't prove causation, you simply repeat an experiment so often until you are confident to say: "okay now the chance that the correlation happened by accident is really, really small".
I get that the correlation/causation thing is overdone in internet discussion, often applied incorrectly. See also, the misuse of "red herring," "Occam's razor" or even generally "logical fallacy."
But the punching thing is not the issue. The issue is cases like how violent crime is related to ice cream sales. Even if the relation were very strong, it would not be because ice cream causes crime, or visa versa. Both are related to heat, which is an independent factor that happens to cause both of the above. There is no causal relationship between the effects, and to say there is a causation is incorrect.
For a popular real example of this, look at pretty much any study reported by a media outlet on things that "cause" IQ increases, better job performance, etc. Of course there are causes to these things. But saying that eating fast food makes you dumber is simply failing to control for the variables of wealth/class/region.
|
And that's not even talking about spurious correlations...
When you start searching for patterns in large enough amounts of data you will find them. Just look at bible numerology for an (in)famous example. However, it takes quite a bit more go explain the data than a simple statistical test. All correlation in a data set can do is indicate that there is something worth investigating; not that you've actually found anything.
|
Is it common to commute long distances for people in the US? Here the hardcore commuters I know do maximum 60 km to go / 60 to come back. Do people in the US commute like 200-300 km? I heard tales about that 
|
On September 23 2014 19:19 SoSexy wrote:Is it common to commute long distances for people in the US? Here the hardcore commuters I know do maximum 60 km to go / 60 to come back. Do people in the US commute like 200-300 km? I heard tales about that
I'm not sure I have heard of anyone driving 200-300 km to commute (120-180 Miles by us measurements) but its somewhat common for people to travel that far or further on Monday and stay for the work week, then fly / drive home on Friday. This is usually only the case in places where its cheaper to do that then live where you work (New York City, Los Angeles) or if the job pays for the travel, also its possible if you're a very high profile (You make enough that the travel is negligible compared to what you are payed) but do not want to live inside a big city. For example, lets say wealthy people who work in New York City sometimes own houses outside of New York state, such as Rhode Island or Connecticut where their family resides, and they practice this.
Take all this with a grain of salt, everyone's case is different and I dont actually live in an area where this takes place.
|
On September 23 2014 19:46 .kaz wrote:Show nested quote +On September 23 2014 19:19 SoSexy wrote:Is it common to commute long distances for people in the US? Here the hardcore commuters I know do maximum 60 km to go / 60 to come back. Do people in the US commute like 200-300 km? I heard tales about that I'm not sure I have heard of anyone driving 200-300 km to commute (120-180 Miles by us measurements) but its somewhat common for people to travel that far or further on Monday and stay for the work week, then fly / drive home on Friday. This is usually only the case in places where its cheaper to do that then live where you work (New York City, Los Angeles) or if the job pays for the travel, also its possible if you're a very high profile (You make enough that the travel is negligible compared to what you are payed) but do not want to live inside a big city. For example, lets say wealthy people who work in New York City sometimes own houses outside of New York state, such as Rhode Island or Connecticut where their family resides, and they practice this. Take all this with a grain of salt, everyone's case is different and I dont actually live in an area where this takes place.
This was very informative, thanks 
|
Is it true that people in korea almost never see the blue sky because of the pollution? Or is it just a myth?
|
Zurich15328 Posts
In Korea in general? Not even a myth.
In Seoul specifically? Just a myth. It's overcast a lot though.
|
I see the blue, blue sky here in Seoul more often than overcast skies. However between ~May - late August it's pretty hazy from a combination of the humidity and pollution.
|
On September 23 2014 19:19 SoSexy wrote:Is it common to commute long distances for people in the US? Here the hardcore commuters I know do maximum 60 km to go / 60 to come back. Do people in the US commute like 200-300 km? I heard tales about that
Really depends where you are. I lived for a while in the NYC area. The commutes there can be very long in time, though they are generally shorter in distance. Longest I knew of was 60 mi (~100km), most are half that in distance or less. Traffic however can push even a 30 mi (50km) commute into being several hours long on a bad day. I've also lived in the rocky mountain west. There are absolutely 100km+ commutes out there along the valley floors, but the gnarliest ones are the (usually shorter) commutes over treacherous passes. A commute from Victor Idaho to Teton village is like 50km, and is a half hour on a good day, but will often be double that, triple that, or even literally impossible depending on snow conditions.
|
Does Facebook let you post that you're in a relationship with a sibling or cousin?
|
Does anyone know where to find that "drink and a pizza" music video they kept playing during Dreamhack this weekend?
|
Is Pat Condell a comedian?
|
Why shouldn't we nuke Mt. Everest?
|
|
|
Nor does it have any racial tensions to milk for conflict
|
If you where to guess how much it would cost to ship the stuff a normal 1 bedroom apartment would have(1couch, queen size mattress and boxspring, 1 dresser and just "stuff") from one end of canada to the other how much do you think it would be?
|
|
|
I don't know if that's an accurate estimate, but I personally was expecting it to be quite a bit more than that.
|
|
|
|
|
|
EPT
