|
Read the rules in the OP before posting, please.In order to ensure that this thread continues to meet TL standards and follows the proper guidelines, we will be enforcing the rules in the OP more strictly. Be sure to give them a re-read to refresh your memory! The vast majority of you are contributing in a healthy way, keep it up! NOTE: When providing a source, explain why you feel it is relevant and what purpose it adds to the discussion if it's not obvious. Also take note that unsubstantiated tweets/posts meant only to rekindle old arguments can result in a mod action. |
On December 02 2013 10:20 Roe wrote:
...What? No scientist worth anything would say correlation proves a causal relationship. And nothing meaningful ever comes from correlational studies either.
Your so called "correlational studies" are also called "experiments", and are the foundation of every scientific field. And yes we are trying to prove our models(causal relationships) by conducting experiments. We gather a lot of data, and if the correlation is really high, we say we have proven something. Many scientific fields have different thresholds for when something is considered "proven". In particle physics, for example, you would consider something proven if it passes the five sigma barrier, which would mean that the chance that all your correlation just happened randomly is roughly about 1 out of 3.5 million.
(Read more: http://blogs.scientificamerican.com/observations/2012/07/17/five-sigmawhats-that/)
|
Reagan steps into office, and in his first two years unemployment skyrockets.
Reagan starts spending shitloads of money, oil prices drop precipitously, and the US/UK/IMF force deregulations on the rest of the world in order to make it a friendlier environment for American capital, and suddenly unemployment in the United States starts going down.
I can see why you would be confused as to what causes what.
|
I'm reading through this report
http://www.demographia.com/dhi.pdf
which fits with the other stuff I've been reading: That there are a number of places where housing and development regulation or limitation is excessive, and causing high prices that make it hard for people to afford housing.
|
On December 02 2013 09:41 Nyxisto wrote:Show nested quote +On December 02 2013 08:26 IgnE wrote:On December 02 2013 08:04 Nyxisto wrote:On December 02 2013 06:33 Roe wrote:On December 02 2013 05:04 JonnyBNoHo wrote:On December 02 2013 04:21 Roe wrote:
Where was supply side economics a success? As examples in the US, railroad deregulation and alternative energy tax credits have both worked as intended. I need a causation up in dis here place, not correlation. God i fucking hate this, although correlation!= causation is technically correct, it doesn't matter, as basically every scientific field takes a positivist and experimental approach, meaning that correlation is indeed viable evidence if you want to prove something. And regarding the supply side vs demand site debate. I don't like the supply side approach either(because i think it's pretty obvious that all the trickle down stuff is not really working and inequality is super high) but you have to admit that it keeps unemployment low, and seems to produce somewhat of reliable growth. what are you talking about? supply side neoliberalism intentionally implements policies that result in higher unemployment. see reagan's US, thatcher's UK, pinochet's chile, etc. Sorry, but what are you talking about? Supply Side economy, at least by following the general accepted definition, means getting taxes down, getting rid of regulations and thus stimulating job creation, consumption and economic growth. The whole idea is based on the fact that the lack of tax income is compensated by job creation. (At least regarding "newschool" supply side economics. "Oldschoolers" just disliked the state and thought taxes were evil) Also at the end of the Reagan era, unemployment was at 5.4%, which seems pretty low. http://en.wikipedia.org/wiki/Reaganomics#Unemployment_ratesAnother example would be Germany in the last 10 years. Tax cuts and a deregulated job market lead to a TON of jobs. We actually are practically at full employment. As i said before there are hundreds of things wrong with this kind of economic policy, but job creation is probably the one thing that's working pretty well.
Keynes himself talked about economic stimulus through tax cut, and he did not considered it as "supply side economic". It's demand. The main problem is that he consider that stimulus through tax cut is less effective because of savings (Haavelmo proved that through his "theorem").
Most economists I've came by actually considered "Reaganomics" to be keynesian in disguise: the tax cut worked because there was an under employment equilibria, a situation where the economy was not using its entire production potential.
Germany is not an exemple of supply side again, Germany is one of the most hypocritical country ever. They actually invested a lot in their economy, their debt GDP ratio before the crisis was bigger than most EU countries, they had (have) a big investment program in green energy and in the reconstruction of Berlin. Not to mention the euro was greatly helping them (low inflation rate in other european countries, with high inflation in Germany, an euro underevaluated for them and overevaluated for the rest of the euro, there is a piece on Krugman blog about that http://krugman.blogs.nytimes.com/2013/11/12/germanys-lack-of-reciprocity/).
If you want to find supply side economy, find someone who make tax cut and deregulation in a period of growth, like Bush, and see the result.
|
On December 02 2013 10:20 Roe wrote:
...What? No scientist worth anything would say correlation proves a causal relationship. And nothing meaningful ever comes from correlational studies either.
All statistics are correlational in nature. Correlation does not prove causation, but correlation along with correction for confounding factors and a solid theory about how A causes B, does provide evidence for causation. Inference from correlation is pretty much the basis of all empirical evidence in the social sciences, and the only way to test a hypothesis.
When confronted with statistical evidence, no scientist worth a damn would go "correlation doesnt equal causations" becasuse that is incredibly infantile.
|
On December 02 2013 20:03 Crushinator wrote:Show nested quote +On December 02 2013 10:20 Roe wrote:
...What? No scientist worth anything would say correlation proves a causal relationship. And nothing meaningful ever comes from correlational studies either. All statistics are correlational in nature. Correlation does not prove causation, but correlation along with correction for confounding factors and a solid theory about how A causes B, does provide evidence for causation. Inference from correlation is pretty much the basis of all empirical evidence in the social sciences, and the only way to test a hypothesis. When confronted with statistical evidence, no scientist worth a damn would go "correlation doesnt equal causations" becasuse that is incredibly infantile.
To add to this point, correlation is a necessary but not sufficient condition to prove causation.
Elephant in this room: history is not a science. EDIT: scientists don't say "correlation doesn't equal causation" because most scientific experiments are tightly controlled and most scientists suck at statistics. They do argue about methodology all the time.
For example, the famous correlation that economic growth has been better under Democratic presidents than Republican presidents does not necessarily imply that the US would have been better off or would have avoided recessions if it only elected Democrats to the White House.
|
On December 02 2013 20:03 Crushinator wrote:Show nested quote +On December 02 2013 10:20 Roe wrote:
...What? No scientist worth anything would say correlation proves a causal relationship. And nothing meaningful ever comes from correlational studies either. All statistics are correlational in nature. Correlation does not prove causation, but correlation along with correction for confounding factors and a solid theory about how A causes B, does provide evidence for causation. Inference from correlation is pretty much the basis of all empirical evidence in the social sciences, and the only way to test a hypothesis. When confronted with statistical evidence, no scientist worth a damn would go "correlation doesnt equal causations" becasuse that is incredibly infantile.
Maybe because you actually never did any epistemology, but of course all scientists who actually question themself on their practice start by "correlation doesn't equal causations".
In hard science it's pretty easy because a causation is testable, falsifiable, so the question is kinda eluded as long as the experiment can be repeated and repeatedly shows the same result.
For social science, the question of correlation and causation introduced the idea of interpretation and historical context. A correlation can be considered as a causation if there is a valid interpretation of the phenomena (this is why social scientist doesn't respond "it's not causation" when they see statistical evidence, but rather question the interpretation of the statistical evidence), and all the causations are always causations in relation to the context in which they have been interpreted.
Sadly, most of the time (always actually) the context cannot be truthfully exhausted (it's too vast for the human mind) so that the "truths" in social sciences are always incomplete.
|
Norway28716 Posts
Science is not constrained to positivist boundaries. History and other social sciences are sciences and subject to scientific methods of research -but yes, the knowledge attained is "less certain" than in hard sciences. It's still absolutely science though.
|
On December 02 2013 23:45 Liquid`Drone wrote:
Science is not constrained to positivist boundaries. History and other social sciences are sciences and subject to scientific methods of research -but yes, the knowledge attained is "less certain" than in hard sciences. It's still absolutely science though.
It depends on the definition you want to use. If you want to use the romantic, broadly-defined version of anything that attempts to add to our collective knowledge or change our perspective about things, then yes, social sciences are sciences. If you want to use the anti-religious, humanist, narrowly-defined version which people like to use as the end-all be-all of what is possible to know, then no, social sciences are not science.
I think people here are trying to use the more narrow definition, the one in which "science" is always right.
|
On December 02 2013 23:32 coverpunch wrote:Show nested quote +On December 02 2013 20:03 Crushinator wrote:On December 02 2013 10:20 Roe wrote:
...What? No scientist worth anything would say correlation proves a causal relationship. And nothing meaningful ever comes from correlational studies either. All statistics are correlational in nature. Correlation does not prove causation, but correlation along with correction for confounding factors and a solid theory about how A causes B, does provide evidence for causation. Inference from correlation is pretty much the basis of all empirical evidence in the social sciences, and the only way to test a hypothesis. When confronted with statistical evidence, no scientist worth a damn would go "correlation doesnt equal causations" becasuse that is incredibly infantile. To add to this point, correlation is a necessary but not sufficient condition to prove causation. Elephant in this room: history is not a science. EDIT: scientists don't say "correlation doesn't equal causation" because most scientific experiments are tightly controlled and most scientists suck at statistics. They do argue about methodology all the time. For example, the famous correlation that economic growth has been better under Democratic presidents than Republican presidents does not necessarily imply that the US would have been better off or would have avoided recessions if it only elected Democrats to the White House.
To expand, the variables Democratic and Republican are ill-conceived to begin with. You would need to isolate the underlying economic policies that drove this difference in growth.It is not at all apparent that Democraticness or Republicanness have anything at all to do with economic growth. There is no theoretical basis.
I am not intimately familiar with US politics, but I know enough to understand that both groups are very diverse, and I suspect both Republican and Democratic administrations have pursued a wide range of policies over this time period. Isolating the economic policies that led to success or failure would be impossibe. And even if it were possible, we would not have enough data to exclude the possibility that the differences were due to random chance.
This is why these statistics are almost meaningless, not because there is something fundamentally wrong with correlation.
|
On December 02 2013 23:45 Liquid`Drone wrote:
Science is not constrained to positivist boundaries. History and other social sciences are sciences and subject to scientific methods of research -but yes, the knowledge attained is "less certain" than in hard sciences. It's still absolutely science though.
Saying that science here and there are different doesn't mean at all that one is science and the other is not. I come from sociology, and despite the fact that I love sociology, I would never accept anyone saying that sociology is science in the same way as physics (same for economy). This actually lead to some important fact on social science : there are no "rules" that one should follow no matter the context.
It's a matter of fact that the assumption in social science and "hard sciences" have different value. It is actually what makes social sciences more interesting to me, they can be discussed to no end because the context in which the actions are made is inexhaustible. It is not that it is less certain, it is a difference in the caracteristics of the assumptions.
|
Norway28716 Posts
There might be an Atlantic gap here, and I guess I shouldn't enter the US politics megathread imposing my Norwegian views.. Anyway though, here, you generally don't want to use a positivist scientific approach outside hard sciences, and every other field of study would be very offended by the notion that they are unscientific. It's actually cool though, because I'm having an exam on scientific method tomorrow and I'm kinda in this mindset anyway, which I guess is why I'm engaging. To me, the notion that knowledge has to be exact and permanent for it to be scientific sorta reduces science to an incredibly narrow field without like, any real world application outside math/physics? I think that when you're examining people, and the economy fits under people just as much as under numbers, then contextualizing is absolutely crucial - but I'm just a student in this field and far from attaining scholarly status - I'd be delighted to be convinced otherwise.
|
No science does, can or should function as the hardcore positivists thinks it should anyway.
Physics has not claimed getting positive result for about a century now. Maths are the only one who could argue for that I guess, but they do not use the scientific method, so it's not a science either.
The changes in the definition of science have just been an ideological attempt at deciding what kind of knowledge is acceptable. Imo, people in social science should just tell people claiming they are not a science "whatever man, I don't care", and keep on doing important stuff.
PS : yes the knowledge physics get is more certain than what history gets, but I'm sometimes tempted to ask : so what ?
|
On December 03 2013 00:27 Liquid`Drone wrote:
There might be an Atlantic gap here, and I guess I shouldn't enter the US politics megathread imposing my Norwegian views.. Anyway though, here, you generally don't want to use a positivist scientific approach outside hard sciences, and every other field of study would be very offended by the notion that they are unscientific. It's actually cool though, because I'm having an exam on scientific method tomorrow and I'm kinda in this mindset anyway, which I guess is why I'm engaging. To me, the notion that knowledge has to be exact and permanent for it to be scientific sorta reduces science to an incredibly narrow field without like, any real world application outside math/physics? I think that when you're examining people, and the economy fits under people just as much as under numbers, then contextualizing is absolutely crucial - but I'm just a student in this field and far from attaining scholarly status - I'd be delighted to be convinced otherwise.
You're absolutely right in your attitude, both that a positivist scientific approach is not appropriate for non-positivist sciences and that being non-positivist doesn't make a science any less legitimate or important.
But the piece you're missing is politics. Pretending to carry the positivist scientific banner makes it much easier to sound very convincing when you insist one party is better than the other. And if you're on that side, it's very comforting to know you're on the side that is "supported" by science, and equally comforting to know your opponents are dumb enough to try to oppose science.
This arguably happens among sciences too. Economics is so desperate to be seen more as a science than a social science or god forbid a branch of philosophy that economists will curl into pretzels to make stuff fit into complicated-looking equations that nobody can understand. Statisticians are so desperate to be seen as more than trash collectors doing the work nobody else wants to do that they call themselves "bioinformaticians" and "data scientists".
|
On December 02 2013 13:09 IgnE wrote:+ Show Spoiler +Reagan steps into office, and in his first two years unemployment skyrockets. + Show Spoiler +Reagan starts spending shitloads of money, oil prices drop precipitously, and the US/UK/IMF force deregulations on the rest of the world in order to make it a friendlier environment for American capital, and suddenly unemployment in the United States starts going down. I can see why you would be confused as to what causes what.
Unemployment went up early in his first term because the Fed raised interest rates up to 20% to fight inflation. That caused a rough recession and sparked the S&L financial crisis. After, stagflation had ended.
Reagan did spend a lot, though some of that was due to high interest rates on the debt.
You'll have to explain how you think making the rest of the world friendlier for US capital lowered the unemployment rate in the US. Are you making a free trade argument?
|
On December 02 2013 20:03 Crushinator wrote:Show nested quote +On December 02 2013 10:20 Roe wrote:
...What? No scientist worth anything would say correlation proves a causal relationship. And nothing meaningful ever comes from correlational studies either. All statistics are correlational in nature. Correlation does not prove causation, but correlation along with correction for confounding factors and a solid theory about how A causes B, does provide evidence for causation. Inference from correlation is pretty much the basis of all empirical evidence in the social sciences, and the only way to test a hypothesis. When confronted with statistical evidence, no scientist worth a damn would go "correlation doesnt equal causations" becasuse that is incredibly infantile.
No, but they wouldn't be publishing conclusive studies of the nature we were talking about either. When looking at correlations/statistics, a scientist would simply shrug because it obviously wasn't an experiment and isn't proving anything meaningful.
|
Norway28716 Posts
On December 03 2013 00:41 coverpunch wrote:Show nested quote +On December 03 2013 00:27 Liquid`Drone wrote:
There might be an Atlantic gap here, and I guess I shouldn't enter the US politics megathread imposing my Norwegian views.. Anyway though, here, you generally don't want to use a positivist scientific approach outside hard sciences, and every other field of study would be very offended by the notion that they are unscientific. It's actually cool though, because I'm having an exam on scientific method tomorrow and I'm kinda in this mindset anyway, which I guess is why I'm engaging. To me, the notion that knowledge has to be exact and permanent for it to be scientific sorta reduces science to an incredibly narrow field without like, any real world application outside math/physics? I think that when you're examining people, and the economy fits under people just as much as under numbers, then contextualizing is absolutely crucial - but I'm just a student in this field and far from attaining scholarly status - I'd be delighted to be convinced otherwise. You're absolutely right in your attitude, both that a positivist scientific approach is not appropriate for non-positivist sciences and that being non-positivist doesn't make a science any less legitimate or important. But the piece you're missing is politics. Pretending to carry the positivist scientific banner makes it much easier to sound very convincing when you insist one party is better than the other. And if you're on that side, it's very comforting to know you're on the side that is "supported" by science, and equally comforting to know your opponents are dumb enough to try to oppose science. This arguably happens among sciences too. Economics is so desperate to be seen more as a science than a social science or god forbid a branch of philosophy that economists will curl into pretzels to make stuff fit into complicated-looking equations that nobody can understand. Statisticians are so desperate to be seen as more than trash collectors doing the work nobody else wants to do that they call themselves "bioinformaticians" and "data scientists".
Ah! That makes perfect sense. Essentially, quantitative studies are more politically convincing because they can't immediately be dismissed due to sample size or research bias, and the flaws that originate from quantitative studies (on people) can't necessarily be claimed to originate from the political position of the researcher.
I just kinda immediately had a gut reaction to the "history isn't a science" snippet - but I guess if you were mostly commentating on how history is politically perceived and essentially interpreted differently by different political factions and thus less politically convincing of a tool than "statistics" (even when there is great consensus among actual historians) then I cannot but agree with you.
|
On December 03 2013 01:26 Roe wrote:Show nested quote +On December 02 2013 20:03 Crushinator wrote:On December 02 2013 10:20 Roe wrote:
...What? No scientist worth anything would say correlation proves a causal relationship. And nothing meaningful ever comes from correlational studies either. All statistics are correlational in nature. Correlation does not prove causation, but correlation along with correction for confounding factors and a solid theory about how A causes B, does provide evidence for causation. Inference from correlation is pretty much the basis of all empirical evidence in the social sciences, and the only way to test a hypothesis. When confronted with statistical evidence, no scientist worth a damn would go "correlation doesnt equal causations" becasuse that is incredibly infantile. No, but they wouldn't be publishing conclusive studies of the nature we were talking about either. When looking at correlations/statistics, a scientist would simply shrug because it obviously wasn't an experiment and isn't proving anything meaningful.
This is absolute nonsense. Pick up a book on ethnography, or any other methodology that relies on careful observation of the environment (using either qualitative or quantitative methods) rather than controlled experiments.
There is only a small subset of humanities, law, and economics that can realistically conduct controlled experiments. And even other areas like biology rely largely on observation in a natural environment, rather than some controlled setting. Not to mention cosmology, where "experiments" are fundamentally impossible.
The scientific method isn't founded on experiments, it is founded on falsifiable hypotheses. And even that is only part of the equation. I suggest everybody here starts reading up on the philosophy of science before spouting pseudoscientific drivel here.
|
On December 03 2013 01:39 Acrofales wrote:
The scientific method isn't founded on experiments, it is founded on falsifiable hypotheses. And even that is only part of the equation. I suggest everybody here starts reading up on the philosophy of science before spouting pseudoscientific drivel here.
Sorry but this is just outright delusional. The "Method of Falsification" a la Popper you are proposing here is not the standard scientific approach. In medicine you are not trying to prove that a drug isn't working, you are trying to prove that it is. At CERN physicists are not trying to prove that the Higg's Boson doesn't exist, they're collecting enough data so they can be pretty sure it does. Experiments and empirical verification are the core idea of every "hard science". Please show me a scientific field where that is not the case.(You made your point pretty clear, so please prove me wrong and don't expect me to provide any evidence here for my statements  )
Germany is not an example of supply side again, Germany is one of the most hypocritical country ever. They actually invested a lot in their economy, their debt GDP ratio before the crisis was bigger than most EU countries, they had (have) a big investment program in green energy and in the reconstruction of Berlin. Not to mention the euro was greatly helping them (low inflation rate in other european countries, with high inflation in Germany, an euro underevaluated for them and overevaluated for the rest of the euro, there is a piece on Krugman blog about that http://krugman.blogs.nytimes.com/2013/11/12/germanys-lack-of-reciprocity/). If you want to find supply side economy, find someone who make tax cut and deregulation in a period of growth, like Bush, and see the result.
Sorry but the part about German investments is just outright wrong. Our rates of investment were constantly 4-5% below OECD average for the last decade. And the Euro helping us or being in a recession hasn't really anything to do with supply side economics. If you lower taxes and deregulate your markets during a recession it's still supply-side policy and doesn't magically become the opposite. (Or please show me a definition of said policy where it says:"It only is supply side economics if you do it while your economy is booming and your currency sucks". Although i slowly get the impression that your personal definition seems to be "It can't be supply side economics if it works")
|
On December 03 2013 02:14 Nyxisto wrote:Show nested quote +On December 03 2013 01:39 Acrofales wrote:
The scientific method isn't founded on experiments, it is founded on falsifiable hypotheses. And even that is only part of the equation. I suggest everybody here starts reading up on the philosophy of science before spouting pseudoscientific drivel here.
Sorry but this is just outright delusional. The "Method of Falsification" a la Popper you are proposing here is not the standard scientific approach. In medicine you are not trying to prove that a drug isn't working, you are trying to prove that it is. At CERN physicists are not trying to prove that the Higg's Boson doesn't exist, they're collecting enough data so they can be pretty sure it does. Show nested quote +Germany is not an example of supply side again, Germany is one of the most hypocritical country ever. They actually invested a lot in their economy, their debt GDP ratio before the crisis was bigger than most EU countries, they had (have) a big investment program in green energy and in the reconstruction of Berlin. Not to mention the euro was greatly helping them (low inflation rate in other european countries, with high inflation in Germany, an euro underevaluated for them and overevaluated for the rest of the euro, there is a piece on Krugman blog about that http://krugman.blogs.nytimes.com/2013/11/12/germanys-lack-of-reciprocity/). If you want to find supply side economy, find someone who make tax cut and deregulation in a period of growth, like Bush, and see the result. Sorry but the part about German investments is just outright wrong. Our rates of investment were constantly 4-5% below OECD average for the last decade. And the Euro helping us or being in a recession hasn't really anything to do with supply side economics. If you lower taxes and deregulate your markets during a recession it's still supply-side policy and doesn't magically become the opposite.
I'm not advocating Popper's falsification. But a hypothesis can only be scientific if it IS falsifiable. Otherwise we are talking about pseudoscience. For instance, Intelligent Design proposes an alternative hypothesis to Evolution through Natural Selection. However, ID is unfalsifiable: anything you throw at it can be explained away with "but God designed it that way".
I'm not saying that Popper's falsification is how scientists work, or even how scientists work... I am saying that any hypothesis that falls within the realm of science has to be falsifiable.
The hypothesis known as the Standard Model is falsifiable. Of course scientists were trying really really hard to find the Higgs boson, because the Standard Model says it should be there, and NOT finding it would mean there is something completely wrong about the SM (other than that it is incomplete).
You don't need an experiment. You need DATA. You need observations. In this case the easiest way of getting that data was creating a controlled environment where you can smash lots and lots of protons into each other and record in great detail everything that happens in this process. By analyzing these observations they found a new particle which fits nicely with all the properties that the theoretical Higgs boson needs to have. This REMOVES one of the ways in which the Standard Model could be falsified... and makes it increasingly likely that it is indeed an excellent description of how our world works (albeit incomplete, due to the lack of dark matter, the failure to explain gravity, and a number of other shortcomings).
In social sciences, biology, cosmology and many other areas, the easiest way of obtaining data is NOT a controlled experiment, but observing the environment and seeing how things happen. Stating that scientists should ignore data because it doesn't come from an "experiment" is complete hogwash.
PS. I am not arguing one way or the other about the actual data that was presented in this thread, or any hypotheses people here are trying to defend with it. I am just answering to the idea that it should all be dismissed offhandedly for not being gathered in an experiment, which is drivel.
EDIT: just saw your edit. Define hard science please. Imho cosmology and biology count as hard science... please don't conflate experiments and empirical verification to the same thing.
|
|
|
|
|
|
EPT![[image loading]](http://upload.wikimedia.org/wikipedia/commons/a/a6/US_Seasonal_Unemployment.svg)
![[image loading]](http://upload.wikimedia.org/wikipedia/commons/7/78/US_Debt_Trend.svg)
