Science and Falsification of Data

[Gil Hamilton]

That's what needs to change. There needs to be a societal shift where "Science" isn't viewed as a commodity you can buy to give your product/process/political agenda an air of legitimacy. There is a conflict in the scientific community about why we do this; and expanding the sum of human (ALL human) knowledge is only one viewpoint. If we assume that "Science" is a commodity, we must give up on academic integrity; after all, a suit didn't buy you and three other PhDs to listen to you whine to them about silly things like "rigor" and "honesty" from their point of view.

I suppose this is unfair, though, there are alot of companies out there that treat their scientists very well. For example, Intel and alot of other high-tech manufacturers are stand up outfits to work for. Others treat you well because they need your body. The example of that is companies like Exxon who go out of their way to hire PhDs in "Green Science" entirely so they can advertise that they do so, then make them glorified managers, which isn't to say they treat them badly.

What needs to end is the notion that scientific integrity can be bought. Unless that is done, there is no point in punishing fraud or talking about academic integrity at all. All integrity becomes is something you do if the money isn't good enough or you might get caught, and the whole system breaks down.

[Alyrium Denryle]

What needs to end is the notion that scientific integrity can be bought. Unless that is done, there is no point in punishing fraud or talking about academic integrity at all. All integrity becomes is something you do if the money isn't good enough or you might get caught, and the whole system breaks down.

The only way to do that is to black list scientists who's work can be shown to be fraudulent. The Sygentia panel for example. They should be barred forever from getting public grants or publishing in journals. The community itself needs to step up more than it does. There need to be people working for professional societies who's job it is to weed out fraud. Look for intentionally poor experimental design, patterns relating to funding source, stuff like that.

[LionElJonson]

Institutions like the CRU set up to prove a theory would also be a good place to start. Proving something true is scientifically impossible; all you can do is prove something untrue. If you try otherwise, at best you'll get the scientists failing to do what you want. At worst, they're drinking the same kool-aid you are, and actively attempt to cover up their number-fudging.

[D.Turtle]

And where exactly did the CRU falsify data? They (and all the scientists involved in the so-called "Climategate" fake-scandal) were cleared of all wrongdoing by large panels of scientists. They exist (along with many other institutions) to study reality, aka ongoing global warming.

And I'll skip over the laughable claim that it is impossible to prove something true...

[Uncluttered]

I'm actually surprised by the number of people arguing that anybody who has commit scientific fraud is completely irredeemable.

I'd be satisfied if there were some way to punish and then reform a rogue researcher.
The problem is, that the scientists who falsify data are often pressured by management.
I've been flat out asked by executives to "massage" data. They just don't get it.
I once managed to get a detailed explanation of the request in email, then I told him "no" in my reply. Then I get IT.. to invoke sarbanes-oxley and archive the email to cover my ass. The sociopath wanted to fire me and made it difficult to work. I.T. "discovered" that the company was too small to require sarbanes-oxley. Hard copies are too easily faked. I left. My lawyer couldn't find any charges to bring him up on.

If a law was passed making it illegal to falsify data, the managers/executives would have to be punished for creating conditions that lead to it.
Until then, shame is a cheap motivator.

I suspect the main cause is scarcity of funding. This is testable.

[LionElJonson]

And where exactly did the CRU falsify data? They (and all the scientists involved in the so-called "Climategate" fake-scandal) were cleared of all wrongdoing by large panels of scientists. They exist (along with many other institutions) to study reality, aka ongoing global warming.

The panels only really examined the emails; they never examined the source code that was released, and my understanding of it is that said source code includes a number of arbitrary fudge factors hardcoded into it. That is, they're evidence of actual scientific misconduct, instead of just the scientists being not spectacularly nice people (like most of the emails are). I haven't read said source code myself, so I might be wrong, but that is my understanding of the situation.

That said, things like this do not exactly inspire confidence.

And I'll skip over the laughable claim that it is impossible to prove something true...

O_o It's not. That's why scientific experiments are set up to disprove things, why scientists talk about "falsifiability" and not "provability", and why statistical analysis is all about deciding whether or not to reject the null hypothesis. I'm not a scientist, but even I know that much.

[adam_grif]

And I'll skip over the laughable claim that it is impossible to prove something true...

That hypothesis can never be proven, only supported / disproven was drilled into us quite hard in first and second year psychology.

[D.Turtle]

I can prove that I am taller than my little sister. I can prove that the universe is older than 6000 years. I could prove that it is warmer today in my house than yesterday. I can prove that it is generally warmer today than 30 years ago. I can prove that an increase in atmospheric CO2 leads to an increased "greenhouse" effect. I can prove that there is an increase in atmospheric CO2 because of human output.

Those are all claims that can be scientifically proven.

What can not be ultimately proven is a theory to explain those facts. These can only be proven beyond reasonable doubt - something that AGW certainly falls under.

The panels only really examined the emails; they never examined the source code that was released, and my understanding of it is that said source code includes a number of arbitrary fudge factors hardcoded into it. That is, they're evidence of actual scientific misconduct, instead of just the scientists being not spectacularly nice people (like most of the emails are). I haven't read said source code myself, so I might be wrong, but that is my understanding of the situation.

That is complete bullshit. From Realclimate:

HARRY_read_me.txt. This is a 4 year-long work log of Ian (Harry) Harris who was working to upgrade the documentation, metadata and databases associated with the legacy CRU TS 2.1 product, which is not the same as the HadCRUT data (see Mitchell and Jones, 2003 for details). The CSU TS 3.0 is available now (via ClimateExplorer for instance), and so presumably the database problems got fixed. Anyone who has ever worked on constructing a database from dozens of individual, sometimes contradictory and inconsistently formatted datasets will share his evident frustration with how tedious that can be.

...

Fudge factors (update) IDL code in the some of the attached files calculates and applies an artificial ‘fudge factor’ to the MXD proxies to artificially eliminate the ‘divergence pattern’. This was done for a set of experiments reported in this submitted 2004 draft by Osborn and colleagues but which was never published. Section 4.3 explains the rationale very clearly which was to test the sensitivity of the calibration of the MXD proxies should the divergence end up being anthropogenic. It has nothing to do with any temperature record, has not been used in any published reconstruction and is not the source of any hockey stick blade anywhere.

That said, things like this do not exactly inspire confidence.

The FOI request handling was the one and only thing in the entire "climategate" scandal that had any relevancy - and was criticized by the inquiry boards. Unfortunately for the skeptics, it did not in any way affect the science, and there are numerous independent temperature databases that show the same thing as the CRU database did. In addition the vast majority of the data used in the CRU database was publicly accessible.

[adam_grif]

I can prove that it is generally warmer today than 30 years ago. I can prove that an increase in atmospheric CO2 leads to an increased "greenhouse" effect. I can prove that there is an increase in atmospheric CO2 because of human output.

No, none of these things can be proven, just strongly supported. You can show that the difference between average temperatures today and 30 years ago are higher to a degree which is statistically significant; you can show that atmospheric CO2 coincides with increased temperatures to a degree unlikely to be due to chance, and you can show that atmospheric CO2 levels are strongly correlated with the spread of industrialization.

The things you just listed are correlational designs, not experimental ones. This is basic science and statistics. Don't lump me in with the guy trying to talk crap about climategate or whatever, I'm just saying. "Facts" are things you measure, theories explain relationships between them.

[LionElJonson]

I can prove that I am taller than my little sister.

No, you can't. There will always be error in measurements; at best, you can give an estimate of the probability you're taller than your sister, with a confidence interval.

I can prove that the universe is older than 6000 years.

No, you can't. Not to turn this into an argument about Creationism, but can you create an experiment to falsify the hypothesis that the universe was created in such a fashion to look older than 6000 years old, short of building a time machine?

I could prove that it is warmer today in my house than yesterday. I can prove that it is generally warmer today than 30 years ago. I can prove that an increase in atmospheric CO2 leads to an increased "greenhouse" effect. I can prove that there is an increase in atmospheric CO2 because of human output.

Again, these are all measurements. There will always be error with measurements; the best you can give to any of these things is an estimate of the p-value of the analysis of the measurements, and then make a decision on whether or not to reject the null hypothesis.

Those are all claims that can be scientifically proven.

There is no such thing as "scientifically proven". At best, they can be shown to be "very likely".

What can not be ultimately proven is a theory to explain those facts. These can only be proven beyond reasonable doubt - something that AGW certainly falls under.

There are alternate theories (for instance, increases in solar radiation) which have not been adequately disproven to my knowledge, and AGW does not, to my knowledge, explain such things as the Medieval Warm Period, which is warmer than it is today. I'm no climate scientist; I'm just attesting to the evidence I personally have read. Additionally, any science that excessively politicized is suspect; odds are just as good that AGW is modern Lysenkoism as that its oil-industry funded detractors are.

The panels only really examined the emails; they never examined the source code that was released, and my understanding of it is that said source code includes a number of arbitrary fudge factors hardcoded into it. That is, they're evidence of actual scientific misconduct, instead of just the scientists being not spectacularly nice people (like most of the emails are). I haven't read said source code myself, so I might be wrong, but that is my understanding of the situation.

That is complete bullshit. From Realclimate:

HARRY_read_me.txt. This is a 4 year-long work log of Ian (Harry) Harris who was working to upgrade the documentation, metadata and databases associated with the legacy CRU TS 2.1 product, which is not the same as the HadCRUT data (see Mitchell and Jones, 2003 for details). The CSU TS 3.0 is available now (via ClimateExplorer for instance), and so presumably the database problems got fixed. Anyone who has ever worked on constructing a database from dozens of individual, sometimes contradictory and inconsistently formatted datasets will share his evident frustration with how tedious that can be.

...

Fudge factors (update) IDL code in the some of the attached files calculates and applies an artificial ‘fudge factor’ to the MXD proxies to artificially eliminate the ‘divergence pattern’. This was done for a set of experiments reported in this submitted 2004 draft by Osborn and colleagues but which was never published. Section 4.3 explains the rationale very clearly which was to test the sensitivity of the calibration of the MXD proxies should the divergence end up being anthropogenic. It has nothing to do with any temperature record, has not been used in any published reconstruction and is not the source of any hockey stick blade anywhere.

I'm not entirely sure what that means, but it sounds a lot like "It wasn't giving us the results we wanted, so we fudged the numbers until it did," and it could very well be cover-your-ass-ese for exactly that. That's not good science.

That said, things like this do not exactly inspire confidence.

The FOI request handling was the one and only thing in the entire "climategate" scandal that had any relevancy - and was criticized by the inquiry boards. Unfortunately for the skeptics, it did not in any way affect the science, and there are numerous independent temperature databases that show the same thing as the CRU database did. In addition the vast majority of the data used in the CRU database was publicly accessible.

And an adversarial mindset in a heavily politicized field isn't going to skew the results? It's well-known that when politics get involved, the human mind stops thinking rationally; it's a hang-over from when political struggles wound up with people dead or exiled from the tribe. The "us-vs-them" mindset is a symptom of it, right along with ignoring evidence that doesn't fit in with your political views, and ignoring evidence that contradicts your theory is a cardinal sin in science.

[D.Turtle]

[edit]This is in response to adam_grif.

Wrong on all counts.

You can show that the difference between average temperatures today and 30 years ago are higher to a degree which is statistically significant;

If I have all temperature measurements from 30 years ago and compare them to temperature measurements from today, I can see a big difference. There fore it is generally warmer today than 30 years ago. And if you still don't understand: If I take a measurement of the temperature in my room from today, and compare it with one from tomorrow, then I can definitely say whether it was warmer today or tomorrow. Same thing with global temperatures - except that there are thousands of measurements that can be compared.

you can show that atmospheric CO2 coincides with increased temperatures to a degree unlikely to be due to chance,

Look at my claim again:

I can prove that an increase in atmospheric CO2 leads to an increased "greenhouse" effect.

Note that I did not claim that an increase in atmospheric CO2 leads to an increase in global temperatures - only to an increased greenhouse effect. That CO2 was a greenhouse gas wasproven over 150 years ago. It is a basic fact that atmospheric CO2 absorbs (and re-radiates) heat. The matter gets more complicated when you compare it with temperatures, as there are numerous effects besides the CO2 greenhouse effect. But thats why I only claimed the increased greenhouse effect as a fact.

and you can show that atmospheric CO2 levels are strongly correlated with the spread of industrialization.

You can do better than that. From Realclimate:

CO2 produced from burning fossil fuels or burning forests has quite a different isotopic composition from CO2 in the atmosphere. This is because plants have a preference for the lighter isotopes (12C vs. 13C); thus they have lower 13C/12C ratios. Since fossil fuels are ultimately derived from ancient plants, plants and fossil fuels all have roughly the same 13C/12C ratio – about 2% lower than that of the atmosphere. As CO2 from these materials is released into, and mixes with, the atmosphere, the average 13C/12C ratio of the atmosphere decreases.

The things you just listed are correlational designs, not experimental ones. This is basic science and statistics. Don't lump me in with the guy trying to talk crap about climategate or whatever, I'm just saying. "Facts" are things you measure, theories explain relationships between them.

And all the claims I posted were facts, that can be - and are - measured.

[D.Turtle]

I can prove that I am taller than my little sister.

No, you can't. There will always be error in measurements; at best, you can give an estimate of the probability you're taller than your sister, with a confidence interval.

LMAO. You have such a laughable interpretation of truth and proving facts, that is completely useless for anything concerning the real world.

There are alternate theories (for instance, increases in solar radiation) which have not been adequately disproven to my knowledge, and AGW does not, to my knowledge, explain such things as the Medieval Warm Period, which is warmer than it is today. I'm no climate scientist; I'm just attesting to the evidence I personally have read. Additionally, any science that excessively politicized is suspect; odds are just as good that AGW is modern Lysenkoism as that its oil-industry funded detractors are.

All already addressed by scientists. Its not their fault that idiotic idiots don't want to believe them. As for conspiracy theories: Oh please... Hundreds of labs worldwide all conspiring together, no one betraying the conspiracy, etc. There are numerous independent scientists and researches who have independently researched the matter and have almost all (except for a few discredited scientists reached the same conclusion.

I'm not entirely sure what that means, but it sounds a lot like "It wasn't giving us the results we wanted, so we fudged the numbers until it did," and it could very well be cover-your-ass-ese for exactly that. That's not good science.

Since you apparently missed the relevant part, I'll quote it again:

IDL code in the some of the attached files calculates and applies an artificial ‘fudge factor’ to the MXD proxies to artificially eliminate the ‘divergence pattern’. This was done for a set of experiments reported in this submitted 2004 draft by Osborn and colleagues but which was never published. Section 4.3 explains the rationale very clearly which was to test the sensitivity of the calibration of the MXD proxies should the divergence end up being anthropogenic. It has nothing to do with any temperature record, has not been used in any published reconstruction and is not the source of any hockey stick blade anywhere.

And an adversarial mindset in a heavily politicized field isn't going to skew the results? It's well-known that when politics get involved, the human mind stops thinking rationally; it's a hang-over from when political struggles wound up with people dead or exiled from the tribe. The "us-vs-them" mindset is a symptom of it, right along with ignoring evidence that doesn't fit in with your political views, and ignoring evidence that contradicts your theory is a cardinal sin in science.

Well, then lucky us, global warming is not dependent on any one scientist, lab, detection method, or whatever. It is robust.

[LionElJonson]

I didn't particularly want to get into a climate debate; I was simply picking an easy target due to the way they let politics influence their research. Frankly, until it ceases to be a politicized subject, any climate research data is suspect, regardless of what the results gathered are. Additionally, because it is politicized, people will dig in and fortify themselves, rather than engaging in proper scientific discussion. I'm not arguing for any sort of grand conspiracy; just a lot of scientists looking out for themselves and/or drinking their own kool aid.

Simply put, politics are poison to good science, and I think this is something we can all agree on.

[D.Turtle]

Politics makes science a lot harder, true. However, it is not always possible to simply wait and do nothing until it is no longer a matter of politics. What you have to do is look at such an area of science a lot closer and more skeptically than a non-controversial subject.

Having said that, in natural sciences there simply comes a point when evidence is overwhelming. It is a lot more difficult in social sciences, as there no simple experiments, etc. are possible. So economic theories or political theories would be such an area where it is quite difficult if not impossible to hold a debate on a pure scientific basis.

So, yes politics is poison to good science, but it does not make good science impossible.

[Zixinus]

Frankly, until it ceases to be a politicized subject, any climate research data is suspect, regardless of what the results gathered are.

So you just slap on your knee and say that no one scientific institution can make a valid conclusion on a politically charged topic, regardless of how well done the research is?

You are aware that one of the points of the scientific process, and academia in particular, is to weed out personal bias?

I do agree that politics can influence scientific research, but more due to forceful involvement of the politicians (or their lackeys) than the political (or sometimes religious or philosophical) inclination of the individual scientists. That is (one of) the points in maintaining academia and peer review.
That is why you are a bit accused of talking about a conspiracy: scientists' work are checked by other scientists and generally by everyone in academia. Any scientists in the same field who looks over their work and finds serious faults (based on scientific analysis, rather than political) can and will call them out on it. For, say, a paper to get trough with bad science in it, it has to get trough the peer review board and has to go unnoticed by every other scientist. The only way I can imagine this happening is if the peer review board is bribed and the paper goes into a too obscure topic.

I mean, do you believe that scientists know how suns are born? You probably do, because the data behind the theories are logical and supported by observation, as well as the theory of another field (nuclear physics). Yet, would you believe it when I point out that these theories are not experimentally verified, because no one artificially collected one solar mass worth of material into one bundle to see what happened?

[adam_grif]

If I have all temperature measurements from 30 years ago and compare them to temperature measurements from today, I can see a big difference. There fore it is generally warmer today than 30 years ago.

If only it were that simple; Research Methods 2 would have been a joke of a subject. We don't measure a temperature today and compare it to a temperature last year and say "This year is hotter than last year." We collect datasets, and then we have to determine the probability that the differences between temperature today and last year were due to probabilistic variation.

There are a few ways this can be done. The nice simple way is using something like Student's t-test. Even then, we don't "prove" that temperatures are trending upwards, we demonstrate that the differences are statistically significant, and reject the Null Hypothesis that the temperatures are "the same". This lends support to the alternative hypothesis that it is getting hotter (doesn't prove it).

This may seem like a bunch of useless busywork and distinctions, but this is how statistics is done. Before you bring up some tangent about how "I CAN DETERMINE THE AVERAGE TEMPERATURE THEN AND NOW AND COMPARE AND THEY ARE HIGHER NOW SO ITS HOTTER", let me preemptively address that. Yes, you can show that the numbers are higher, but that's not what we're discussing. The hypothesis is that the higher temperatures are not random. Since there is always a possibility that it is, nomatter how supportive your dataset, we still only express it in probability terms (p < .001, for instance).

Also, I'm not necessarily supporting the use of t-tests and p-values as the be all end all, and I love CI's as much as the next guy. Before someone jumps down my throat about that.

Note that I did not claim that an increase in atmospheric CO2 leads to an increase in global temperatures - only to an increased greenhouse effect. That CO2 was a greenhouse gas was proven over 150 years ago. It is a basic fact that atmospheric CO2 absorbs (and re-radiates) heat. The matter gets more complicated when you compare it with temperatures, as there are numerous effects besides the CO2 greenhouse effect. But thats why I only claimed the increased greenhouse effect as a fact.

I would ask you to stop abusing the words "proof" and "fact".

I reiterate; nothing in science is ever proven. "Facts" in science are direct measurements, not their associated explanations. It ain't a "fact" that more CO2 = higher temperatures, nor is it a "fact" that it absorbs and re-radiates heat. That it does that is an explanation for facts garnered from experimentation ("the more CO2 we put in the higher the temperatures got", or whatever it was). Facts are things that "did" happen, not things that "will" happen. Predictions are never made with 100% certainty, and there's always the off chance that next time you drop a ball it'll float up.

Jesus, I feel dirty even saying things like this, because I acknowledge that GW happens, recognize that competing explanations are hokum and support governmental action taken to minimize it. But you're not using the nomenclature correctly, and must be corrected on minor points!

[Simon_Jester]

To be honest, the idea that there is no such thing as "proof" in science is a fetish created by the Popperians; it's a philosophical statement, not a practical one. Real scientists routinely regard the really nailed down ironclad principles as "proven." Maxwell's laws are "proven" for any value of "proof" that human beings can talk about without being hypocrites.

Note that a real scientist will not depend on the notion that "anything which we have proven will never ever need to be examined." Things can be true without being sacred truths, and arguments "proven" true by ample arguments have turned out to be wrong when new facts came to light. But we shouldn't be shy of using the word "proven" to describe scientific theories that are infinitely more certain than would be required in, say, criminal court to "prove" a verdict.

[adam_grif]

To be honest, the idea that there is no such thing as "proof" in science is a fetish created by the Popperians; it's a philosophical statement, not a practical one. Real scientists routinely regard the really nailed down ironclad principles as "proven." Maxwell's laws are "proven" for any value of "proof" that human beings can talk about without being hypocrites.

Note that a real scientist will not depend on the notion that "anything which we have proven will never ever need to be examined." Things can be true without being sacred truths, and arguments "proven" true by ample arguments have turned out to be wrong when new facts came to light. But we shouldn't be shy of using the word "proven" to describe scientific theories that are infinitely more certain than would be required in, say, criminal court to "prove" a verdict.

I'm well aware of the necessity for practicality and that "p < .001" is just as good as really proving something functionally. But like it or not, it's true, and there's no use denying it. I never feel comfortable lying about things to the public just so we don't confuse them (like when people say "you never prove anything in science" as an argument against evolution or something).

[Simon_Jester]

Thing is, in the language real human beings normally use, "proven" doesn't mean "this proposition is certain with probability 1.000..." out to fifty or a hundred zeroes. We simply don't use that kind of confidence interval. By human standards of proof, evolution or Maxwell's laws are proven. Global warming is somewhere between "damn near proven" and "proven."

And I think it's misleading the public to say otherwise. When they hear authoritative sources say "nothing is ever certain in science" when the issue is as certain as anything they depend on in daily life... that misleads them much more than the same scientist saying "yes, I am certain, we are certain, and the great bulk of the scientific community is certain enough to stake their careers on this" or some such.

[adam_grif]

Well then, we come back to the necessity of adequate education in the sciences.

[D.Turtle]

snip temperature stuff

If I have a temperature reading from 13:00 today showing it is 28°C in my room. And I show you one from 13:00 yesterday that showed the same thermometer with 35°C in my room, would you be willing to admit that it is a fact that it was warmer in my room yesterday at 13:00 than it is today?

I would ask you to stop abusing the words "proof" and "fact".

I reiterate; nothing in science is ever proven. "Facts" in science are direct measurements, not their associated explanations. It ain't a "fact" that more CO2 = higher temperatures,

I NEVER EVER EVER FUCKING CLAIMED THAT. So stop repeating that false claim. I claimed that it is a fact that CO2 is a greenhouse gas, and that human output has increased the amount of CO2 in the atmosphere.

nor is it a "fact" that it absorbs and re-radiates heat.

It is a simple characteristic of CO2 that is blocks heat radiation, shown through numerous experiments. It is just as sure as the fact that water freezes at about 0°C, boils at about 100°C (and I'll add at sea level pressure, just for you), steel melts at higher temperatures than water, etc.

That it does that is an explanation for facts garnered from experimentation ("the more CO2 we put in the higher the temperatures got", or whatever it was).

No, it was a characteristic of CO2 measured in a lab. Like the strengths of various materials can be tested in a lab. Its not a vague "we think it works like this", but the result of direct measurements.

Facts are things that "did" happen, not things that "will" happen. Predictions are never made with 100% certainty, and there's always the off chance that next time you drop a ball it'll float up.

Such a definition of fact or true is completely useless in the real world. Anybody calculating in the possibility that gravity might fail can be safely ignored.

Jesus, I feel dirty even saying things like this, because I acknowledge that GW happens, recognize that competing explanations are hokum and support governmental action taken to minimize it. But you're not using the nomenclature correctly, and must be corrected on minor points!

No, what you are doing is undermining/questioning the basis upon which AGW builds - a basis that lies in very simple measurements and characteristics of materials - and not in any vague theories or whatever. What you are doing is giving idiots who don't understand AGW ammunition for questioning it. Just as applying your definitions would give flat earhters, creationists, or whatever ammunition.

[Alyrium Denryle]

No, you can't. There will always be error in measurements; at best, you can give an estimate of the probability you're taller than your sister, with a confidence interval.

No moron. You can prove something that simple. Why? Because both your sample and population size is 1. Your errors in measurement would need to be larger than the difference between the two siblings. If this is not the case, you can prove the hypothesis true.

There are alternate theories (for instance, increases in solar radiation) which have not been adequately disproven to my knowledge

Actually, they have been. Increases in solar radiation have known effects that have been taken into account. If one excludes anthropogenic effects, the models currently used cannot post-dict (re-create) known past changes in temperature.

I'm not entirely sure what that means, but it sounds a lot like "It wasn't giving us the results we wanted, so we fudged the numbers until it did," and it could very well be cover-your-ass-ese for exactly that. That's not good science.

No. There was a discrepancy between a set of proxy data (temps calculated from tree rings) and actual measurements of temperature. The size of this divergence was known, so a fudge factor was put in. Additionally, this was never published. I play with data to see what it does all the damn time, it helps me understand the modeling that I am doing. That does not mean I would publish it outside of a methods paper.

And an adversarial mindset in a heavily politicized field isn't going to skew the results?

No. Real scientists are adversarial anyway. It is rare thing to get the scientific community to agree on something like AGW so quickly. Why? Because it is in our interest collectively to prove a new hotshot theory wrong. That is how you get famous, that is how you get research funding.

Real scientists do not use FOI requests to get access to someone's raw data. In fact any scientist no matter how unappreciated their field of study would be pissed at that. Why? Because it is a recipe to get scooped, plagiarized, or otherwise robbed of your publishing rights.

The "us-vs-them" mindset is a symptom of it, right along with ignoring evidence that doesn't fit in with your political views, and ignoring evidence that contradicts your theory is a cardinal sin in science.

And where exactly is this evidence?

If I have all temperature measurements from 30 years ago and compare them to temperature measurements from today, I can see a big difference. There fore it is generally warmer today than 30 years ago. And if you still don't understand: If I take a measurement of the temperature in my room from today, and compare it with one from tomorrow, then I can definitely say whether it was warmer today or tomorrow. Same thing with global temperatures - except that there are thousands of measurements that can be compared.

Strictly speaking, you cant. When you are doing climate science or any statistics you are working with samples. If I take a sample of temperatures from thirty years ago and compare them to equivalent modern temps (and bear in mind, you are taking a sample, not the entire population of temperatures from all locations at all points in time), I am creating a probability distribution. I can prove something is so likely to be different, that it is stupid to disagree... but I have not absolutely proven that the temperatures are different. It is sort of like claiming "If I lean against this wall, I will pass through it". There is a non-zero chance that you will in fact pass through the wall thanks to quantum mechanics. However, the chance is so small you can laugh at it.

I didn't particularly want to get into a climate debate; I was simply picking an easy target due to the way they let politics influence their research. Frankly, until it ceases to be a politicized subject, any climate research data is suspect, regardless of what the results gathered are.

How wonderful for us that science is a self-correcting enterprise.

Additionally, because it is politicized, people will dig in and fortify themselves, rather than engaging in proper scientific discussion.

They do that regardless of whether it is politicized or not. You should see some of the nasty fights in evolutionary biology, behavioral ecology, and biogeography...

To be honest, the idea that there is no such thing as "proof" in science is a fetish created by the Popperians; it's a philosophical statement, not a practical one. Real scientists routinely regard the really nailed down ironclad principles as "proven." Maxwell's laws are "proven" for any value of "proof" that human beings can talk about without being hypocrites.

Yes. On the other hand if you want to be really really accurate, you need to acknowledge the abstract possibility that you will quantum-tunnel through a brick wall.

If I have a temperature reading from 13:00 today showing it is 28°C in my room. And I show you one from 13:00 yesterday that showed the same thermometer with 35°C in my room, would you be willing to admit that it is a fact that it was warmer in my room yesterday at 13:00 than it is today?

If you measured them in exactly the same place, same time, with exactly the same (down to picolumen) amount of light, and humidity, pressure etc, and you have a sufficiently accurate thermometer, then I would be willing to say that at that time in that particular spot, the room was warmer.

Otherwise, I would assign a high probability that the room is warmer one day than the next.

[D.Turtle]

Strictly speaking, you cant. When you are doing climate science or any statistics you are working with samples. If I take a sample of temperatures from thirty years ago and compare them to equivalent modern temps (and bear in mind, you are taking a sample, not the entire population of temperatures from all locations at all points in time), I am creating a probability distribution. I can prove something is so likely to be different, that it is stupid to disagree... but I have not absolutely proven that the temperatures are different. It is sort of like claiming "If I lean against this wall, I will pass through it". There is a non-zero chance that you will in fact pass through the wall thanks to quantum mechanics. However, the chance is so small you can laugh at it.

At this time there are so many different sources, using different methods, that all reach the same conclusion, that a statement like "it is generally warmer today than 30 years ago" is pretty much in the same region as your falling through the wall example. Now, I would be willing to admit that there is some uncertainty in the temperature record of the last few decades, except it is so negligibly tiny in comparison to the temperature difference as to be safely ignorable.

If you measured them in exactly the same place, same time, with exactly the same (down to picolumen) amount of light, and humidity, pressure etc, and you have a sufficiently accurate thermometer, then I would be willing to say that at that time in that particular spot, the room was warmer.

Otherwise, I would assign a high probability that the room is warmer one day than the next.

If I have 10 or so different measurements, all using different methodology, and they all show the same thing, then that probability is so close to 1 as to make no difference. If it were a matter that only interests scientists, then it might be valid to point out that uncertainty, but when you start addressing the public you do yourself no favor when pointing out the one in a billion (or less) chance that you might be wrong.

[Alyrium Denryle]

At this time there are so many different sources, using different methods, that all reach the same conclusion, that a statement like "it is generally warmer today than 30 years ago" is pretty much in the same region as your falling through the wall example. Now, I would be willing to admit that there is some uncertainty in the temperature record of the last few decades, except it is so negligibly tiny in comparison to the temperature difference as to be safely ignorable.

While I agree, I cannot have you making incorrect statements about science or statistics.

If I have 10 or so different measurements, all using different methodology, and they all show the same thing, then that probability is so close to 1 as to make no difference. If it were a matter that only interests scientists, then it might be valid to point out that uncertainty, but when you start addressing the public you do yourself no favor when pointing out the one in a billion (or less) chance that you might be wrong.

That distinction is what separates us from propagandists.

As for your temperature, what your different methodologies and measurements give you is just increased statistical power--a reduced probability of false negative.

[Zed]

I have the impression that people are pulling statistical data on probabilities out of their asses.