...Why scientific programming does not compute

Ver Greeneyes
Ver Greeneyes
Joined: 26 Mar 09
Posts: 140
Credit: 9,562,235
RAC: 0

Every scientific measurement

Every scientific measurement you make is embedded in a theoretic framework, because the measurement devices were developed with a theory in mind. I don't agree with philosophers of science who say this makes all measurements entirely subjective, because I think they underestimate the amount of possible theories that would lead to the same methodology and measurements without changing the outcome (a bit like solving systems of linear equations - error bars tend to give you a certain amount of room before the class of the system changes, and these critical points can be quantified).

However, the moment you start to draw conclusions from data, you start to apply hypotheses and weigh results. Just think of global temperature measurements: if we took all the measurement stations and weighed them equally, the results would be skewed toward European and perhaps American areas, simply because that's where the majority of measurement stations are. Thus we have to weigh measurements from other places more heavily, even though this increases the uncertainty.

Of course, while that's an example where something is obviously wrong, climate modelling is much more subtle. Since you're fitting your models to observed data, you have to be very careful you're not overfitting. Climate models tend to be so ridiculously complex that it becomes very difficult to figure out what their assumptions are. Meta studies help to regain some measure of generality, but they won't defeat a bad assumption if it is implicit in every model.

So modelling itself is far from objective, and that's without even trying to interpret the results. But politicians (who tend to have an education in the Arts rather than the Sciences, with a law or economics degree at best) aren't going to look through scientists' models and draw their own conclusions. They wouldn't know where to start. So it comes down to scientists to put the different sources of information together, using the most sensible weighing they can think of, and present 'executive summaries' of their work to politicians. That's fine, as long as their choices are carefully reviewed, but it's by no means objective.

Rod
Rod
Joined: 3 Jan 06
Posts: 4,396
Credit: 811,266
RAC: 0

I think policy regarding

I think policy regarding climate change is still being run by the policy makers and the mob. The only change is scientist's creditability is being questioned by the policy makers and the mob.

There are some who can live without wild things and some who cannot. - Aldo Leopold

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6,076
Credit: 116,687,831
RAC: 47,451

RE: not being able to see

Quote:
not being able to see the forest because those trees keep getting in the way!


Oddly, regarding the CRU matter and the divergence b/w tree ring data and the more 'exact' ( as some would have it ) temperature measurements - that the trees were removed! By that I mean it would seem that the trees 'know' something that we don't, in the sense that their growth patterns went the other way from the 'official' sensors. So why are they doing one thing when we think they ought be doing another? That's really my central premise in this thread, when you start to ignore big hints like this then you're thinking becomes orthogonal to reality pretty quick slick. Presumably the climate scientists wish to close the circle at some point and give predictions ie. when temperature does this -> the eco-sphere will do that .... but you'll fall over at that point 'cos you disconnected the eco-sphere relevance early on.

Now here's a bit if a laugh, if you will. I've just returned today from Far North Queensland. Beautiful place, lovely sights and experiences, alot worth preserving, and no shortage of Snappy the Crocodile! I visited a 'rainforest centre' where you get to view the glorious habitat on elevated walkways. I strongly recommend it for any coming down this way ie. Daintree. Alas it had an 'education centre' - for school groups really so you expect they'd shoot stuff at a lower level - but a couple of clangers really stood out.

Firstly : one panel described the Big Bang as 'a theory about an early thermal event that some some scientists believe in'. Hmmmm, what an odd way to say that. I guess we can tread on some socio-political toes but not others, eh? :-)

Secondly : each entry fee was described as 'reducing global warming'. Bland statement, literal, no fill. Couldn't find any little asterisks or notes to join the dots on that one. I wasn't aware that we could buy our way out of GW, but hey if we can I now know where to spend it! :-)

Thirdly : one display wrote of 'global ramming'. I await, with tension, the outcome of that!

Fourthly : there was a Porsche, a Merc, and an Evo-9 in the staff car park. All nearby a sign that said they were a carbon neutral establishment! But they probably belonged to some rich visitors, right? :-)

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter. Blaise Pascal

Matt Giwer
Matt Giwer
Joined: 12 Dec 05
Posts: 144
Credit: 6,891,649
RAC: 0

RE: Every scientific

Quote:
Every scientific measurement you make is embedded in a theoretic framework, because the measurement devices were developed with a theory in mind. I don't agree with philosophers of science who say this makes all measurements entirely subjective, because I think they underestimate the amount of possible theories that would lead to the same methodology and measurements without changing the outcome (a bit like solving systems of linear equations - error bars tend to give you a certain amount of room before the class of the system changes, and these critical points can be quantified).

It has been my experience that these self-proclaimed philosophers of science are not scientists and as such their opinions are without foundation.

Science is what scientists do. NEVER has any "measurement device" been conceived outside of a physical framework. As such there is no way to consider a measurement method an abstract endeavor making such observations nonsense -- most charitably they are made in ignorance of science. If the Large Hadron Collider or LIGO is an abstraction then so is a yardstick and for exactly the same "philosophical" reasons.

Quote:
However, the moment you start to draw conclusions from data, you start to apply hypotheses and weigh results. Just think of global temperature measurements: if we took all the measurement stations and weighed them equally, the results would be skewed toward European and perhaps American areas, simply because that's where the majority of measurement stations are. Thus we have to weigh measurements from other places more heavily, even though this increases the uncertainty.

In the real world one increases the measurements in places where there are gaps such as in the LHC rather than arbitrarily and deterministically changing the data set of well understood regions. Rather than do the hard work equivalent to an LHC are changing what is known. Fact: Anything which changes the raw data makes statistics inapplicable because of the foundation of statistical theory.l

Weighting the data means changing the data making statistics inapplicable. It does not increase uncertainty as that is a statistical concept and is thus inapplicable. It does in fact make the data set useless for any statistical analysis.

Quote:
Of course, while that's an example where something is obviously wrong, climate modelling is much more subtle. Since you're fitting your models to observed data, you have to be very careful you're not overfitting. Climate models tend to be so ridiculously complex that it becomes very difficult to figure out what their assumptions are. Meta studies help to regain some measure of generality, but they won't defeat a bad assumption if it is implicit in every model.

If the data set is modified in any way then the model is being fit to the modification not the data.

Quote:
So modelling itself is far from objective, and that's without even trying to interpret the results. But politicians (who tend to have an education in the Arts rather than the Sciences, with a law or economics degree at best) aren't going to look through scientists' models and draw their own conclusions. They wouldn't know where to start. So it comes down to scientists to put the different sources of information together, using the most sensible weighing they can think of, and present 'executive summaries' of their work to politicians. That's fine, as long as their choices are carefully reviewed, but it's by no means objective.

However any executive summary which differs in style or content from the synopsis of any professional publication is no longer science. The summary is no more comprehensible than the paper itself.

These discussions are little more than trying to find an excuse to do politics instead of science by means of distorting science.

For example distorting a data set is sort of framed in terms of "it is the best we can do." So? It does not mean you can do science with a distorted data set. Reality does not change regardless of how heartfelt the pleading of inability. If you can't do it then you can't do it. Inability does not justify pretending it can be done. Reality does not make exceptions based upon what human perceive to be their needs.

mikey
mikey
Joined: 22 Jan 05
Posts: 5,477
Credit: 531,814,983
RAC: 180,883

RE: These discussions are

Quote:
These discussions are little more than trying to find an excuse to do politics instead of science by means of distorting science.

Actually in this case the Scientists distorted the data set making it fit their own agenda! So while it was intended to be a Scientific paper it ended up being a Political paper instead.

Quote:
For example distorting a data set is sort of framed in terms of "it is the best we can do." So? It does not mean you can do science with a distorted data set. Reality does not change regardless of how heartfelt the pleading of inability. If you can't do it then you can't do it. Inability does not justify pretending it can be done. Reality does not make exceptions based upon what human perceive to be their needs.

I think we humans will always try to make the data fit the way they think it should until it just won't, then we humans will come up with another theory that does fit. When we see a news story of someone being hurt in an accident by another driver, most of us will jump to the conclusion that the other driver was at fault. BUT that is not always the case, sometimes the person that was driving did something dumb causing the accident to occur leaving the other driver as just an innocent bystander caught up in the moment. My example could be someone hurt at an intersection caused by them running the light and the person in the 2nd car being unable to avoid the person that got hurt! We always jump to conclusions based on our past experiences, sometimes they are right and they are wrong! But how we deal with the data makes us who we are and how we will handle future data.

Ver Greeneyes
Ver Greeneyes
Joined: 26 Mar 09
Posts: 140
Credit: 9,562,235
RAC: 0

RE: In the real world one

Quote:
In the real world one increases the measurements in places where there are gaps such as in the LHC rather than arbitrarily and deterministically changing the data set of well understood regions. Rather than do the hard work equivalent to an LHC are changing what is known. Fact: Anything which changes the raw data makes statistics inapplicable because of the foundation of statistical theory.


If you really think that then there's no way I can convince you of anything beyond it. But I don't think you realize just how many branches of science that excludes.

One more thing though: I wasn't talking about models that work on modified data, I was talking about models that work based on implicit assumptions and ideas about modelling strategies - which includes just about every model out there. Of course, whichever way you look at it, the fact remains that you're really just 'playing with data' to draw your conclusions. That's true for even the simplest or weakest (most generalized) models. My point is that this doesn't make them invalid, as long as they have predictive value (which things like psycho-analysis don't).

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6,076
Credit: 116,687,831
RAC: 47,451

RE: Fact: Anything which

Quote:
Fact: Anything which changes the raw data makes statistics inapplicable because of the foundation of statistical theory.


I think what you may mean is that attempting to reliably deduce population statistics/trends from sample statistics is invalidated when there is a non-random selection of data points ( which is what the CRU matter demonstrated as a deliberate ploy, it was inbuilt within their research to depth - including long standing coding mechanisms ). But I reckon you are right in a sense, if you change the data set you have to re-do the analysis. Is that what you had in mind?

BTW - has anyone else here actually reviewed the primary ( online ) revealed log from CRU ( that is : the un-edited literal words of the researchers within the context that they themselves endow )? I don't mean derivative spin commentaries and whatnot, but the complete and actual words from the horse's mouth? This should, in my opinion, lay to rest any worries about anti-AGW conspiracies arising from this matter. It really was all their own work. But heck, I beg you not to take my word for a damn thing! Read for yourself .... you have an un-rivalled opportunity! :-)

It is interesting that the LHC is mentioned. The entire field of high energy physics is littered with excellent lessons upon how to approach statistics, sampling problems, measurement error, experimental theory dynamic and good old fashion luck & stuff-ups! Also Einstein was appointed to the Swiss patent office partly on his ability and experience with the problems of direct physical measurements - as then mentioned in what we would now call his 'resume'.

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter. Blaise Pascal

Rod
Rod
Joined: 3 Jan 06
Posts: 4,396
Credit: 811,266
RAC: 0

A Scientist (among many) has

A Scientist (among many) has passed on

There are some who can live without wild things and some who cannot. - Aldo Leopold

Rod
Rod
Joined: 3 Jan 06
Posts: 4,396
Credit: 811,266
RAC: 0

My Apologies for this hijack.

My Apologies for this hijack. I would like to broaden the question and state maybe we should stop relying on scientists to solve our problems:-)
From IEEE Spectrum:

Why Engineers Must try to Save The World

_______________________________
I am just throwing this out there. I really don't have an opinion on anything lately.. :-)

There are some who can live without wild things and some who cannot. - Aldo Leopold

mikey
mikey
Joined: 22 Jan 05
Posts: 5,477
Credit: 531,814,983
RAC: 180,883

RE: My Apologies for this

Quote:

My Apologies for this hijack. I would like to broaden the question and state maybe we should stop relying on scientists to solve our problems:-)
From IEEE Spectrum:

Why Engineers Must try to Save The World
_____________________________
I am just throwing this out there. I really don't have an opinion on anything lately.. :-)

So what you, and the article, are saying is don't invent anything because it might be used as a military item?! So the 'inventor' of the wheel should be taken out and shot and we should go back to before it? The wheel let man doing everything from expand his horizons to producing electricity so you and I can discuss this! It is also used in the military by trucks, jeeps, cars and even tanks!! Would Oppenheimer have thought the same way if he had seen nuclear submarines, ships or power plants in action? there are people in this World that without Dr. Oppenheimer would not be living up to 21st century standards without him and his 'experiment'! Dolly the sheep was the first publicized cloning of a higher level animal, but it won't be the last! One day someone WILL clone a human, if they haven't already. It is something we will deal with as a human race that 50 years ago was unthinkable, as was 'the bomb' 50 years before its development! No to stop development of something because someone thought 'outside the box' is silly and dumb, IMO. How about the the Mars, Lunar and other space craft that have expanded out exploration of our own Solar system and beyond? Without nuclear power powering those things that would not have been possible!! Heck the Russians drove a Lunar Rover around on the moon for over a year, TWICE, on nuclear powered rovers!! No Dr. Oppenheimer saw the worst, hopefully, use of nuclear power, there are many, MANY examples of the same power being used for the good of mankind!! Remember our moms said, 'you have to take the bad with the good', this is one of those times!! And Forrest Gump said 'life is like a box of chocolates, you never know what you are going to get'! Dr. Oppenheimer opened the box, we are starting to learn how to harness what is in there and to use it for more than blowing things up!

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.