Everytime I hear people trot out the "correlation is not causation" thread, I feel like thwacking them over the head with a statistics book.
Here's the correct formulation (paraphrased from textbooks): Correlation MIGHT be causation if other factors can be analyzed and accounted for. As a corollary: If you can use different regions and times, but have only the one variable (leaded gasoline) changed, you can make a strong argument for causation.
From the MJ article:
Meanwhile, Nevin had kept busy as well, and in 2007 he published a new paper looking at crime trends around the world (PDF). This way, he could make sure the close match he'd found between the lead curve and the crime curve wasn't just a coincidence. Sure, maybe the real culprit in the United States was something else happening at the exact same time, but what are the odds of that same something happening at several different times in several different countries?
The MJ article goes through the history of this research and multiple researchers using different datasets and techniques reached the same conclusion. This may go beyond "might" by correlation.
There's a much simpler, basic reason: most people believe the $1000+ hearing aids are better. They aren't really, but that is the value seniors and other place on them, and no manufacturer wants to leave money on the table. Even seniors without insurance would want to pay that price because they believe quality has to cost that much. If they bought a $300 model, seniors would be lobbying their kids for the $2000 model.
So much of health care spending is 'irrational', and as long as our comfort and longevity are on the table, we'll spend what ever it takes even if it is out of our own pockets.
Wow. You manage to bring in one thing to explain this thing and get it spectacularly wrong. As someone else pointed out, the Columbus Day weekend is the traditional ramp down time for refineries in the U.S. as they rejigger their formulation for fall (You didn't know refineries changed formulas for the season?). Also, several major supply routes got messed up:
"Among the recent disruptions, an Aug. 6 fire at a Chevron Corp. refinery in Richmond left one of the region's largest refineries producing at a reduced capacity, and a Chevron pipeline that moves crude to northern California also was shut down. There also was a power failure that affected an Exxon Mobil Corp. refinery in Torrance, but the refinery has resumed normal operations."
As for Krugman and this being all the fault of QEx: there's a reason gas is not part of the core measure of inflation. Last I checked, we aren't in an inflation cycle yet. Gas is a volatile price (no-pun intended) that jumps way up and down responding to things like, you know, refineries having fires and pipeline shut downs. It's left out of most inflation conversations among economists.
Anyway, thanks for playing! Here's a home version of the game "The Eeeevil Fed Is Coming For Your Savings!!"
Basically, there was some growing problem (no pun intended) with tree-ring based temperature studies, and this study helps figure out how to take into account those problems, and a probably cause.
What I *am* surprised about is that they're admitting to it this quickly. I expected it to be a decade or two before TEPCO or the government would admit that anything but the earthquake/tsunami were to blame. And that they're even blaming their own culture of discipline... wow. That's some harsh self-criticism.
Exactly. Japanese Parliamentary reports are usually cover-ups or whitewashes of political and industry screw ups. This is probably a first in Japanese post-war history!
I tried to read it, but couldn't get past the first 100 pages. It was trying soooooo hard to be cool & edgy it turned me off. It was a like an Ritalin-addicts pastiche of William Gibson.
I think everything he wrote from from Cryptonomicon on are his best works. The Baroque Cycle was fantastic! It's kind of sad that it's Snow Crash that gets all the attention.
He scares the left because he's basically about leaving the states to their own resources, and most states (especially the Red States), don't generate enough GDP to do anything on their own. Also, state politics are notoriously corrupt and prone to special interest groups (see California). The U.S. Federal government is the easiest institution to create a social safety net, control big corporations, etc. The things Ron Paul wants to let go of.
I can understand your point of view, but it doesn't change the fact the Japanese government had begun the process of feeling for surrender. What ground soldiers did is separate from what the politicians wanted to do. Did LBJ want American soldiers massacring civillians at Mai Lai? No, but insane shit happens in war. The soldiers that tortured and murdered your grandfather were war criminals, and one can only hope that those Japanese soldiers either committed sepiku or were brutally beaten & killed by American marines.
I find Stanislaw a little hard to read at times (i.e., boring), but after reading a couple of his books from the library and ready to give up on him, I finally read Solaris. And wow, Solaris is different from the movies. It's not really about the planet, or what happens to George Clooney on the planet. It's a question about "Can science ever really know the unknowable?"
The book is more like a future history of research into the planet Solaris and the failure of humanity to understand the how or why of the planet. Humanity meets an intelligence (Solaris), and neither side can understand or communicate with each other. Very haunting for me.
I think you misunderstand the term audiophile here. Originally, it did mean someone like you who liked good sound. Now, it means a pretentious half-twit who measures the quality of his system by the price he paid than the performance it delivers.
There is studio equipment, which is awesome and any true audiophile would have bought, and there is Best Buy/Amazon.com/etc. "audiophile" equipment which is actually sub-par compared to the studio equipment but costs 2 to 5x more than than the studio equipment.
This looks like a case where a company successfully innovated in the marketplace, and then took out patents to help secure their position.
The position adopted by the author of TFA was not subtle and, in my view, does not help the discussion.
Yes, that's the impression I got too. Although Honeywell's behavior suggests they want Nest out of the market entirely. For starters, they didn't serve papers to Nest -- Nest found out via the press release from Honeywell. Also, Honeywell wants Nest to simply cease & desist. I wish these kinds of lawsuits never happen -- a legitimate claim being used to obliterate a true innovator. If I was Honeywell, I would have used Coase's theorem to negotiate with Nest, pointing out that Honeywell was in the right, but they should work together to find a mutually beneficial outcome.
Gee wouldn't it be better for the developers to notify the manager IMMEDIATELY that the users were available for testing so they're going to put it off? You need to have a daily meeting to communicate?
Yup. Our company overloads our managers to the point they get 200+ e-mails a day. My company is stupid, and with any luck, I will be getting out soon. *knock on wood* Lots of companies are like that. Thus the popularity of the stand-up meeting: all signal, no noise. E-mail: 99% noise.
And that's the problem of Agile in a nutshell - you open up for not meeting schedules because you have to be agile, allowing changes. So you either deliver an unfinished product or go over time. Seriously so. The correlation between not meeting deadlines with a finished product and Agile is so high it converges towards 1. Really.
You can't have something for nothing, and in Agile you pay for your gains by not being able to deliver a full product on time. If upper management aren't informed of this, they need to be.
Interesting. Because when we were in modified waterfall, we had to allow changes or the clients would not be happy. Change happens in custom Enterprise software development. If you are working in a "product" shop, bless you. You are probably much happier than me. I have to deal with users who have conflicting ideas of how they do their own jobs. In fact, they have imaginary ideas about the business requirements which we find out half way through is wrong. Change is our environment. Management wants us to be able to react to changing user requirements quickly, document & cost them, etc. With the modified waterfall, there was a huge bureacracy to do that, and people would argue if it was a "clarification", "discovered requirement", "scope increase", etc.
Which again points to an important point about any process you follow: it must match your reality. If you are working in products, you have the luxury to aim for a relatively well defined set of requirements for a given release. You could probably do well from modified waterfall or some sort of iterative waterfall. For us, management's #1 priority is to make our clients happy and give them custom developed software that meets their needs, even if the clients don't even understand their needs. For us, agile is a better fit. For you, something else.
What do we do? We talk to each other. Emails, IMs, the occasional skype call, and very rarely a face-to-face so we can whiteboard. It's all ad hoc, entirely as needed, and only the people who need to be present are invited.
I've been pushing my department of developers from using the corporate approved modified waterfall method, which has lead to massive budget overruns since corporate pushed it down on us, to an Agile-like process (You do know you are allowed to modify Agile to your environment?). The last two projects we did that way came in under budget and only slipped the original schedule due to client-introduced requirement changes.
One of the things we did was the 2-minute stand-up meeting (there were only 3 people for that project). Kept it focused to: What I did yesterday, what I will do today and what problems I'm having. When we had the weekly meetings, we usually found out about roadblocks a week after they happened. Now, the project manager found out things within 24 hours and could fix them quickly. "Yeah, the users aren't available for testing so they're going to put it off--" "I'll talk to [their manager] and get it fixed."
So when I read all these skeptics and haters, I'm shocked. For those of us who used stand-up meetings, they are so much better than the old sit-down 1-hour meetings. Then I dug into the criticism and I think I know what's the problem: cargo cult management. That's where clueless managers follow the form without understanding the motivation or why it works. So the punishments, like singing and running a lap, makes my skin crawl. The agile manifesto explicitly says People over Process. Just a simple "You're late!" is sufficient. I've read other Agile horror stories on Slashdot over the last two years, and it seems like those shops followed the motions without understanding the why. One guy complained that he saw a bug but wasn't allowed to fix it because of some bullcrap like "You don't have the token to work on that". WTF? That's not Agile! If something's broken, anyone is allowed to jump in and fix it. But that shop seemed more interested in following the liturgy than actually being Agile. Remember the first rule of Agile:
Individuals and interactions over processes and tools
The Agile horror stories I'm hearing are teams choosing processes and tools over people.
Slashdot Mirror
Technically, that would be "meta". ;-)
Everytime I hear people trot out the "correlation is not causation" thread, I feel like thwacking them over the head with a statistics book.
Here's the correct formulation (paraphrased from textbooks): Correlation MIGHT be causation if other factors can be analyzed and accounted for. As a corollary: If you can use different regions and times, but have only the one variable (leaded gasoline) changed, you can make a strong argument for causation.
From the MJ article:
The MJ article goes through the history of this research and multiple researchers using different datasets and techniques reached the same conclusion. This may go beyond "might" by correlation.
I've seen far too many developers get excited by Node.js. Boggles my mind, personally, but the idea of a Javascript back-end is coming.
Do you know modern tradesmen? You might be surprised what kind of hobbies they like.
You can if you started posting cool YouTube videos. Hint, hint! :-)
Mike Rowe would like to talk to you.
See this documentary for more details. :-P
Like pet veterinary care.
There's a much simpler, basic reason: most people believe the $1000+ hearing aids are better. They aren't really, but that is the value seniors and other place on them, and no manufacturer wants to leave money on the table. Even seniors without insurance would want to pay that price because they believe quality has to cost that much. If they bought a $300 model, seniors would be lobbying their kids for the $2000 model.
So much of health care spending is 'irrational', and as long as our comfort and longevity are on the table, we'll spend what ever it takes even if it is out of our own pockets.
Wow. You manage to bring in one thing to explain this thing and get it spectacularly wrong. As someone else pointed out, the Columbus Day weekend is the traditional ramp down time for refineries in the U.S. as they rejigger their formulation for fall (You didn't know refineries changed formulas for the season?). Also, several major supply routes got messed up:
From California gasoline prices soar amid refinery and pipeline shutdowns By The Associated Press:
As for Krugman and this being all the fault of QEx: there's a reason gas is not part of the core measure of inflation. Last I checked, we aren't in an inflation cycle yet. Gas is a volatile price (no-pun intended) that jumps way up and down responding to things like, you know, refineries having fires and pipeline shut downs. It's left out of most inflation conversations among economists.
Anyway, thanks for playing! Here's a home version of the game "The Eeeevil Fed Is Coming For Your Savings!!"
That is correct, AC. "Excellence" indeed. For most Japanese, they associate the word with the ninja from legend.
Read RealClimate's coverage.
Basically, there was some growing problem (no pun intended) with tree-ring based temperature studies, and this study helps figure out how to take into account those problems, and a probably cause.
What I *am* surprised about is that they're admitting to it this quickly. I expected it to be a decade or two before TEPCO or the government would admit that anything but the earthquake/tsunami were to blame. And that they're even blaming their own culture of discipline... wow. That's some harsh self-criticism.
Exactly. Japanese Parliamentary reports are usually cover-ups or whitewashes of political and industry screw ups. This is probably a first in Japanese post-war history!
I tried to read it, but couldn't get past the first 100 pages. It was trying soooooo hard to be cool & edgy it turned me off. It was a like an Ritalin-addicts pastiche of William Gibson.
I think everything he wrote from from Cryptonomicon on are his best works. The Baroque Cycle was fantastic! It's kind of sad that it's Snow Crash that gets all the attention.
Yeah! Just leave that to the birds! :-)
He scares the left because he's basically about leaving the states to their own resources, and most states (especially the Red States), don't generate enough GDP to do anything on their own. Also, state politics are notoriously corrupt and prone to special interest groups (see California). The U.S. Federal government is the easiest institution to create a social safety net, control big corporations, etc. The things Ron Paul wants to let go of.
Yes, this Daniel Pearl.
I can understand your point of view, but it doesn't change the fact the Japanese government had begun the process of feeling for surrender. What ground soldiers did is separate from what the politicians wanted to do. Did LBJ want American soldiers massacring civillians at Mai Lai? No, but insane shit happens in war. The soldiers that tortured and murdered your grandfather were war criminals, and one can only hope that those Japanese soldiers either committed sepiku or were brutally beaten & killed by American marines.
I find Stanislaw a little hard to read at times (i.e., boring), but after reading a couple of his books from the library and ready to give up on him, I finally read Solaris. And wow, Solaris is different from the movies. It's not really about the planet, or what happens to George Clooney on the planet. It's a question about "Can science ever really know the unknowable?" The book is more like a future history of research into the planet Solaris and the failure of humanity to understand the how or why of the planet. Humanity meets an intelligence (Solaris), and neither side can understand or communicate with each other. Very haunting for me.
Seconding "End of Eternity". In fact, I was scanning the replies looking for just this reply. :-)
I think you misunderstand the term audiophile here. Originally, it did mean someone like you who liked good sound. Now, it means a pretentious half-twit who measures the quality of his system by the price he paid than the performance it delivers.
There is studio equipment, which is awesome and any true audiophile would have bought, and there is Best Buy/Amazon.com/etc. "audiophile" equipment which is actually sub-par compared to the studio equipment but costs 2 to 5x more than than the studio equipment.
This looks like a case where a company successfully innovated in the marketplace, and then took out patents to help secure their position.
The position adopted by the author of TFA was not subtle and, in my view, does not help the discussion.
Yes, that's the impression I got too. Although Honeywell's behavior suggests they want Nest out of the market entirely. For starters, they didn't serve papers to Nest -- Nest found out via the press release from Honeywell. Also, Honeywell wants Nest to simply cease & desist. I wish these kinds of lawsuits never happen -- a legitimate claim being used to obliterate a true innovator. If I was Honeywell, I would have used Coase's theorem to negotiate with Nest, pointing out that Honeywell was in the right, but they should work together to find a mutually beneficial outcome.
Gee wouldn't it be better for the developers to notify the manager IMMEDIATELY that the users were available for testing so they're going to put it off? You need to have a daily meeting to communicate?
Yup. Our company overloads our managers to the point they get 200+ e-mails a day. My company is stupid, and with any luck, I will be getting out soon. *knock on wood* Lots of companies are like that. Thus the popularity of the stand-up meeting: all signal, no noise. E-mail: 99% noise.
And that's the problem of Agile in a nutshell - you open up for not meeting schedules because you have to be agile, allowing changes. So you either deliver an unfinished product or go over time. Seriously so. The correlation between not meeting deadlines with a finished product and Agile is so high it converges towards 1. Really.
You can't have something for nothing, and in Agile you pay for your gains by not being able to deliver a full product on time. If upper management aren't informed of this, they need to be.
Interesting. Because when we were in modified waterfall, we had to allow changes or the clients would not be happy. Change happens in custom Enterprise software development. If you are working in a "product" shop, bless you. You are probably much happier than me. I have to deal with users who have conflicting ideas of how they do their own jobs. In fact, they have imaginary ideas about the business requirements which we find out half way through is wrong. Change is our environment. Management wants us to be able to react to changing user requirements quickly, document & cost them, etc. With the modified waterfall, there was a huge bureacracy to do that, and people would argue if it was a "clarification", "discovered requirement", "scope increase", etc.
Which again points to an important point about any process you follow: it must match your reality. If you are working in products, you have the luxury to aim for a relatively well defined set of requirements for a given release. You could probably do well from modified waterfall or some sort of iterative waterfall. For us, management's #1 priority is to make our clients happy and give them custom developed software that meets their needs, even if the clients don't even understand their needs. For us, agile is a better fit. For you, something else.
Individuals and interactions over processes and tools
You're doing Agile right.
I've been pushing my department of developers from using the corporate approved modified waterfall method, which has lead to massive budget overruns since corporate pushed it down on us, to an Agile-like process (You do know you are allowed to modify Agile to your environment?). The last two projects we did that way came in under budget and only slipped the original schedule due to client-introduced requirement changes.
One of the things we did was the 2-minute stand-up meeting (there were only 3 people for that project). Kept it focused to: What I did yesterday, what I will do today and what problems I'm having. When we had the weekly meetings, we usually found out about roadblocks a week after they happened. Now, the project manager found out things within 24 hours and could fix them quickly. "Yeah, the users aren't available for testing so they're going to put it off--" "I'll talk to [their manager] and get it fixed."
So when I read all these skeptics and haters, I'm shocked. For those of us who used stand-up meetings, they are so much better than the old sit-down 1-hour meetings. Then I dug into the criticism and I think I know what's the problem: cargo cult management. That's where clueless managers follow the form without understanding the motivation or why it works. So the punishments, like singing and running a lap, makes my skin crawl. The agile manifesto explicitly says People over Process. Just a simple "You're late!" is sufficient. I've read other Agile horror stories on Slashdot over the last two years, and it seems like those shops followed the motions without understanding the why. One guy complained that he saw a bug but wasn't allowed to fix it because of some bullcrap like "You don't have the token to work on that". WTF? That's not Agile! If something's broken, anyone is allowed to jump in and fix it. But that shop seemed more interested in following the liturgy than actually being Agile. Remember the first rule of Agile:
The Agile horror stories I'm hearing are teams choosing processes and tools over people.