Except to make the analogy accurate you have to be constantly increasing the stress on the tower (turning up gravity I guess) and at the same time your big brother walks by every once in a while and shakes the table -- the effect of other earthquakes. While you MIGHT knock the tower over prematurely, in general you'll be better off by getting as many blocks out while gravity is weak, before the whole thing gets knocked over by your brother.
No, it's not. A correlation coefficient is a number calculated by a formula to indicate the strength of a correlation, which is a relationship between two variables. Another number calculated by a formula is a p-value, which indicates the likelihood that the observed correlation is real or due to chance.
No, it is not. A correlation is a statistical measure and is therefore probabilistic. You can never point to a series of numbers and say "they are correlated." If you measure a correlation properly you get a p-value. Roughly interpreted, that's the probability that they are NOT correlated, even though they appear to be. It is never zero, although it can be made arbitrarily close by collecting more data.
If two things are actually correlated then they are causally related in some way. If they only appear to be correlated, but are not, it's called "co-incidence."
This is a very simple misunderstanding of what correlation is, and leads to a lot of silly things being said about it. Perhaps it stems from engineers studying convolution and learning that correlation is convolution except that neither function is reversed. This is true, and closely related to the correlation you're talking about when you say two things are correlated, but lacks the statistical disclaimers.
That doesn't make sense. The equations of quantum mechanics describe the world, they don't create it (otherwise the world would only be a century old). We've found equations that work, but just like any equations they only tell us what's happening, not why. Newton's or Einstein's equations or the ideal gas law are no different.
No, you don't get to peek at the cat's state. You can peek at some other property, like how fast it's transitioning between (quantum) states, but you can't see the cat's state.
To use the analogy of the thought experiment, maybe you could peek in the box and determine whether there is a cat in the box at all, or what colour it is, but if you look long enough to determine whether the cat is alive or dead then you'll destroy the superposition.
Lots of universities had research reactors (a few still do). They had no more security than some bored grad students working in the outer lab. If it was an open house even they would have been too busy to look after every wandering kid.
Ah, I thought we were talking about what SHOULD be. Since you want to just talk about law, let's do that.
Assuming you're American, the US federal FOIA applies to information and documents controlled by the US government. A (soon to be) exception (kind of) is data produced by research using federal funds. Not e-mails, notes, or anything else.
Even supposing university researchers WERE employees of government agencies (they are not), internal e-mails, notebooks, etc. fall under exception 2. Also, release of information about individuals is prohibited by the US privacy act, unless the release is to the individual the information is about.
You said it yourself, governments are subject to FOI acts. Researchers and university research labs are not governments. I guess it was your statements that were misinformed.
A Google search for "scratched iphone lens" turns up a lot of hits, including a few videos about how to polish the scratches out (not very successful) or replace the camera back. I'll be taking a look at everyone's phone next time I'm at work though.
What you're suggesting is that there is no correlation, just coincidence. Coincidence is not correlation, although it can appear so if you've done your stats poorly.
I read what you said. Adding stress is a non issue. We just don't use enough energy to pump water down to add any serious amount of stress.
I didn't reply to your second paragraph because it's just making things up. Relieving stress through small earthquakes removes energy and lowers the risk of bigger earthquakes later. If the slip in part of the fault increases stress on another part, it would have eventually anyway.
What if there are inner-space humans living deep underground and we're drowning them and the earthquakes are their cries of despair?!
If the earthquakes are indeed caused entirely by injecting water and not by any pent up geological stress then there's nothing to worry about. Unless the oil companies are using nuclear pumps (as in bombs), they won't be putting enough energy into the ground to do any serious damage.
Frakking has some potential issues, but earthquakes aren't one of them. The earthquakes caused by frakking range from irrelevant to beneficial.
There have been unusually large numbers of (largish) quakes everywhere in the last few years. The hypothesis is that the big Indonesian one shook everything up and we're still feeling the effects.
The current Canadian government yes. Previous ones were into balanced budgets and paying back the national debt (gently) before it got all trendy and, um, urgent.
Um, real austerity is cutting things so you HAVE no deficit, so you can PAY BACK your debt. A balanced budget should be seen as normal, not austere.
Your sneering at "European welfare states" is kind of ironic considering that the US (that bastion of "capitalism") is deeply in debt, with huge deficits, and doesn't seem to be too concerned about doing anything about it. Half of the major US political parties still think large national deficits are a good thing.
Why should I talk to him? He seems to be a bit of an "it's not a problem I notice therefore it doesn't exist" type.
I just looked at my iPhone 4 lens cover. It's in pretty good shape, but it does have some scratches, a nice big one right over the lower right quadrant of the field of view. I haven't noticed any serious problems in the pictures but another scratch like that could definitely cause some.
Phones are tossed in pockets and left on counters, and all sorts of other things I'd never do with my several thousand dollar SLR lenses without a lens cap on. Mine also goes kayaking, hiking in deserts and hang gliding, so it might get a bit more wear than is usual. Either way, I'd rather have a better lens cover with a minor sacrifice in flare resistance (that's how the actual professional camera reviewers rate it).
They still screwed up the outdoor one. One photo is taken at least several minutes after the other.
It looks like Apple traded a bit of flare resistance in challenging photographic conditions for other benefits - a scratch proof lens being not the least of them. Personally, I'd much rather have a sapphire lens cover for my cell phone camera than the ability to take slightly less shitty pictures in shitty conditions.
Google uses those things to get you to look at their ads, but they also use those things to collect information about you. I don't really see a problem with that either, so long as people understand the price they're paying.
What I do find strange is that rabidly privacy oriented Slashdot seems entirely blinded to what Google really is. I guess the geek's penchant for shiny overwhelms even his paranoia.
Artists don't like precise, or technical, and they depend on interpretation. Actually, I found the comments by the biologist part of the group more disturbing. Yes, he's a biologist, but for someone working with bacteria and heavy metals he really should know something about chemistry.
I'm not criticizing making art projects out of science. I've collaborated with opera singers and instrumentalists and am a photographer myself. Science is beautiful. My objection was to the OP calling this art science.
Support for OS X 10.5 didn't suddenly end when 10.6 was released either. And XP support is only going until 2014 because of overwhelming demand. Microsoft has tried to cancel it several times over the last few years.
C causes A and B is one of the three types of causal relationships. Correlation does not imply direct causation would be correct.
No, correlations are always real. Our claim of a correlation may be in error. We can only detect correlation with a finite (but arbitrarily high) probability. However, as in your example, you'll find that "false correlations" are usually statistical errors. In your example, if you properly correct for multiple comparisons you will probably not find any correlations in your random data. The correlations you predict you'll find are actually false claims due to improper statistical procedure.
The correct response to a measured correlation is to ask whether one or more of the three causal possibilities is ruled out by other data, then do an experiment to determine with of the remaining is true. Correlation is not sufficient to claim a particular causal relationship, but the throwaway phrase "correlation does not imply causation" is in fact incorrect.
Slashdot Mirror
Except to make the analogy accurate you have to be constantly increasing the stress on the tower (turning up gravity I guess) and at the same time your big brother walks by every once in a while and shakes the table -- the effect of other earthquakes. While you MIGHT knock the tower over prematurely, in general you'll be better off by getting as many blocks out while gravity is weak, before the whole thing gets knocked over by your brother.
No, it's not. A correlation coefficient is a number calculated by a formula to indicate the strength of a correlation, which is a relationship between two variables. Another number calculated by a formula is a p-value, which indicates the likelihood that the observed correlation is real or due to chance.
http://en.wikipedia.org/wiki/Correlation_and_dependence
No, it is not. A correlation is a statistical measure and is therefore probabilistic. You can never point to a series of numbers and say "they are correlated." If you measure a correlation properly you get a p-value. Roughly interpreted, that's the probability that they are NOT correlated, even though they appear to be. It is never zero, although it can be made arbitrarily close by collecting more data.
If two things are actually correlated then they are causally related in some way. If they only appear to be correlated, but are not, it's called "co-incidence."
This is a very simple misunderstanding of what correlation is, and leads to a lot of silly things being said about it. Perhaps it stems from engineers studying convolution and learning that correlation is convolution except that neither function is reversed. This is true, and closely related to the correlation you're talking about when you say two things are correlated, but lacks the statistical disclaimers.
That doesn't make sense. The equations of quantum mechanics describe the world, they don't create it (otherwise the world would only be a century old). We've found equations that work, but just like any equations they only tell us what's happening, not why. Newton's or Einstein's equations or the ideal gas law are no different.
No, you don't get to peek at the cat's state. You can peek at some other property, like how fast it's transitioning between (quantum) states, but you can't see the cat's state.
To use the analogy of the thought experiment, maybe you could peek in the box and determine whether there is a cat in the box at all, or what colour it is, but if you look long enough to determine whether the cat is alive or dead then you'll destroy the superposition.
Lots of universities had research reactors (a few still do). They had no more security than some bored grad students working in the outer lab. If it was an open house even they would have been too busy to look after every wandering kid.
Ah, I thought we were talking about what SHOULD be. Since you want to just talk about law, let's do that.
Assuming you're American, the US federal FOIA applies to information and documents controlled by the US government. A (soon to be) exception (kind of) is data produced by research using federal funds. Not e-mails, notes, or anything else.
Even supposing university researchers WERE employees of government agencies (they are not), internal e-mails, notebooks, etc. fall under exception 2. Also, release of information about individuals is prohibited by the US privacy act, unless the release is to the individual the information is about.
You said it yourself, governments are subject to FOI acts. Researchers and university research labs are not governments. I guess it was your statements that were misinformed.
A Google search for "scratched iphone lens" turns up a lot of hits, including a few videos about how to polish the scratches out (not very successful) or replace the camera back. I'll be taking a look at everyone's phone next time I'm at work though.
What you're suggesting is that there is no correlation, just coincidence. Coincidence is not correlation, although it can appear so if you've done your stats poorly.
They need to stop depending on short term visions, and expectations of them.
Although it sounds like eliminating 90% of their printer models would be an excellent place to start, and it shouldn't take four years to do that.
Yes. And they still haven't learned that it just pisses customers off.
Not in a way that counts, anyway.
I read what you said. Adding stress is a non issue. We just don't use enough energy to pump water down to add any serious amount of stress.
I didn't reply to your second paragraph because it's just making things up. Relieving stress through small earthquakes removes energy and lowers the risk of bigger earthquakes later. If the slip in part of the fault increases stress on another part, it would have eventually anyway.
What if there are inner-space humans living deep underground and we're drowning them and the earthquakes are their cries of despair?!
If the earthquakes are indeed caused entirely by injecting water and not by any pent up geological stress then there's nothing to worry about. Unless the oil companies are using nuclear pumps (as in bombs), they won't be putting enough energy into the ground to do any serious damage.
Frakking has some potential issues, but earthquakes aren't one of them. The earthquakes caused by frakking range from irrelevant to beneficial.
There have been unusually large numbers of (largish) quakes everywhere in the last few years. The hypothesis is that the big Indonesian one shook everything up and we're still feeling the effects.
If it moves when you lubricate it, then there was stress on it.
You think maybe earthquakes cause frakking? Or perhaps an oil company exec's decisions cause both frakking AND earthquakes?
The current Canadian government yes. Previous ones were into balanced budgets and paying back the national debt (gently) before it got all trendy and, um, urgent.
Um, real austerity is cutting things so you HAVE no deficit, so you can PAY BACK your debt. A balanced budget should be seen as normal, not austere.
Your sneering at "European welfare states" is kind of ironic considering that the US (that bastion of "capitalism") is deeply in debt, with huge deficits, and doesn't seem to be too concerned about doing anything about it. Half of the major US political parties still think large national deficits are a good thing.
Why should I talk to him? He seems to be a bit of an "it's not a problem I notice therefore it doesn't exist" type.
I just looked at my iPhone 4 lens cover. It's in pretty good shape, but it does have some scratches, a nice big one right over the lower right quadrant of the field of view. I haven't noticed any serious problems in the pictures but another scratch like that could definitely cause some.
Phones are tossed in pockets and left on counters, and all sorts of other things I'd never do with my several thousand dollar SLR lenses without a lens cap on. Mine also goes kayaking, hiking in deserts and hang gliding, so it might get a bit more wear than is usual. Either way, I'd rather have a better lens cover with a minor sacrifice in flare resistance (that's how the actual professional camera reviewers rate it).
They still screwed up the outdoor one. One photo is taken at least several minutes after the other.
It looks like Apple traded a bit of flare resistance in challenging photographic conditions for other benefits - a scratch proof lens being not the least of them. Personally, I'd much rather have a sapphire lens cover for my cell phone camera than the ability to take slightly less shitty pictures in shitty conditions.
Google uses those things to get you to look at their ads, but they also use those things to collect information about you. I don't really see a problem with that either, so long as people understand the price they're paying.
What I do find strange is that rabidly privacy oriented Slashdot seems entirely blinded to what Google really is. I guess the geek's penchant for shiny overwhelms even his paranoia.
Artists don't like precise, or technical, and they depend on interpretation. Actually, I found the comments by the biologist part of the group more disturbing. Yes, he's a biologist, but for someone working with bacteria and heavy metals he really should know something about chemistry.
I'm not criticizing making art projects out of science. I've collaborated with opera singers and instrumentalists and am a photographer myself. Science is beautiful. My objection was to the OP calling this art science.
Support for OS X 10.5 didn't suddenly end when 10.6 was released either. And XP support is only going until 2014 because of overwhelming demand. Microsoft has tried to cancel it several times over the last few years.
C causes A and B is one of the three types of causal relationships. Correlation does not imply direct causation would be correct.
No, correlations are always real. Our claim of a correlation may be in error. We can only detect correlation with a finite (but arbitrarily high) probability. However, as in your example, you'll find that "false correlations" are usually statistical errors. In your example, if you properly correct for multiple comparisons you will probably not find any correlations in your random data. The correlations you predict you'll find are actually false claims due to improper statistical procedure.
The correct response to a measured correlation is to ask whether one or more of the three causal possibilities is ruled out by other data, then do an experiment to determine with of the remaining is true. Correlation is not sufficient to claim a particular causal relationship, but the throwaway phrase "correlation does not imply causation" is in fact incorrect.