I was searching the early edition of PNAS for the abstract of yet another sloppy “science by press release” that didn’t bother to give the the title of the paper or the DOI, and came across this paper, so it wasn’t a wasted effort.
Steve McIntyre recently mentioned:
Mann rose to prominence by supposedly being able to detect “faint” signals using “advanced” statistical methods. Lewandowsky has taken this to a new level: using lew-statistics, lew-scientists can deduce properties of population with no members.
Josh (N=0) humor aside, this new paper makes me wonder how many climate science findings would fail evidence thresholds under this new proposed standard?
Revised standards for statistical evidence
Valen E. Johnson
Significance
The lack of reproducibility of scientific research undermines public confidence in science and leads to the misuse of resources when researchers attempt to replicate and extend fallacious research findings. Using recent developments in Bayesian hypothesis testing, a root cause of nonreproducibility is traced to the conduct of significance tests at inappropriately high levels of significance. Modifications of common standards of evidence are proposed to reduce the rate of nonreproducibility of scientific research by a factor of 5 or greater.
Abstract
Recent advances in Bayesian hypothesis testing have led to the development of uniformly most powerful Bayesian tests, which represent an objective, default class of Bayesian hypothesis tests that have the same rejection regions as classical significance tests. Based on the correspondence between these two classes of tests, it is possible to equate the size of classical hypothesis tests with evidence thresholds in Bayesian tests, and to equate P values with Bayes factors. An examination of these connections suggest that recent concerns over the lack of reproducibility of scientific studies can be attributed largely to the conduct of significance tests at unjustifiably high levels of significance. To correct this problem, evidence thresholds required for the declaration of a significant finding should be increased to 25–50:1, and to 100–200:1 for the declaration of a highly significant finding. In terms of classical hypothesis tests, these evidence standards mandate the conduct of tests at the 0.005 or 0.001 level of significance.
From the discussion:
The correspondence between P values and Bayes factors based on UMPBTs suggest that commonly used thresholds for statistical significance represent only moderate evidence against null hypotheses. Although it is difficult to assess the proportion of all tested null hypotheses that are actually true, if one assumes that this proportion is approximately one-half, then these results suggest that between 17% and 25% of marginally significant scientific findings are false. This range of false positives is consistent with nonreproducibility rates reported by others (e.g., ref.5). If the proportion of true null hypotheses is greater than one-half, then the proportion of false positives reported in the scientific literature, and thus the proportion of scientific studies that would fail to replicate, is even higher.
In addition, this estimate of the nonreproducibility rate of scientific findings is based on the use of UMPBTs to establish the rejection regions of Bayesian tests. In general, the use of other default Bayesian methods to model effect sizes results in even higher assignments of posterior probability to rejected null hypotheses, and thus to even higher estimates of false-positive rates.
This phenomenon is discussed further in SI Text, where Bayes factors obtained using several other default Bayesian procedures are compared with UMPBTs (seeFig. S1). These analyses suggest that the range 17–25% underestimates the actual proportion of marginally significant scientific findings that are false.
Finally, it is important to note that this high rate of nonreproducibility is not the result of scientific misconduct, publication bias, file drawer biases, or flawed statistical designs; it is simply the consequence of using evidence thresholds that do not represent sufficiently strong evidence in favor of hypothesized effects.
=================================================================
The full paper is here: http://www.pnas.org/content/early/2013/10/28/1313476110.full.pdf
The SI is here: Download Supporting Information (PDF)
For our layman readers who might be a bit behind on statistics, here is a primer on statistical significance and P-values as it relates to weight loss/nutrition, which is something that you can easily get your mind around.
Gross failure of scientifical nutritional studies is another topic McIntyre recently discussed: A Scathing Indictment of Federally-Funded Nutrition Research
So, while some dicey science findings might simply be low threshold problems, there are real human conduct problems in science too.
Oops, my bad. Forgot to reply to your citation of Cairns/Foster/Hall on adaptive mutation (AM). As I’ve commented repeatedly above, it has long been known that stresses can increase the mutation rate.
The supposedly anti-darwinian “magic” of AM didn’t survive long, as is usual with such supposed mysteries. It was solved in 2000:
http://www.pnas.org/content/97/12/6646.full
“Abstract
Upon starvation some Escherichia coli cells undergo a transient, genome-wide hypermutation (called adaptive mutation) that is recombination-dependent and appears to be a response to a stressful environment. Adaptive mutation may reflect an inducible mechanism that generates genetic variability in times of stress. Previously, however, the regulatory components and signal transduction pathways controlling adaptive mutation were unknown. Here we show that adaptive mutation is regulated by the SOS response, a complex, graded response to DNA damage that includes induction of gene products blocking cell division and promoting mutation, recombination, and DNA repair. We find that SOS-induced levels of proteins other than RecA are needed for adaptive mutation. We report a requirement of RecF for efficient adaptive mutation and provide evidence that the role of RecF in mutation is to allow SOS induction. We also report the discovery of an SOS-controlled inhibitor of adaptive mutation, PsiB. These results indicate that adaptive mutation is a tightly regulated response, controlled both positively and negatively by the SOS system.”
If you were really interested in learning biology rather than promoting a teleological philosophy, it wouldn’t have taken you long to find this reference & others explaining how yet again the supposed ID magic is entirely explicable by biology. Same goes for the flagella of bacteria, archaea & eukaryotes. ID is just old creationist vinegar in a new bottle, using the same tired arguments of 19th century opponents of Darwin.
@ur momisugly milodonharlani
You wrote: “Ohno was wrong about that. He didn’t know then what we know now, & was already shown wrong even before we found out as much as we have about genomics.”
I quote here from the Wiki-bloody-pedia because it’s in electronic rather than paper form:
This reflects what I discussed with a friend who was far more into genomics than I am when I was in that philosophy of biology class in 2005. I do note that Ohno’s hypothesis regarding whole genome duplication (R2) appears to have fallen, but the part that interested me appears to still be currently accepted contra your statement above.
@ur momisugly milodonharlani
I don’t believe that I have ever stated that the Modern Synthesis is false. I have stated that Dawkins’ claims of diverging species being unable to ever exchange genomic information, i.e. HGT can’t happen, is false and that’s not the same thing. I have also stated that while ongoing ID might be possible, there is no evidence for it and that I have no intention of getting into the stoush between the creationists and biologists. Hardly a ringing endorsement for ID. And implying that I endorse ID in the very limited sense that you use the term seems tantamount to calling me a creationist.
I have said that there are some legs in an argument for the universe being created by an intelligence, but that’s entirely separate from any arguments about God fiddling with his creation.
And that I hope is that…
@ur momisugly milodonharlani
So now John Cairns, ? Foster and Barry Hall are “anti-darwinian” creationist magicians, too. [sigh] At least I’m in good company…
Dawkins does not say that about HGT. Search and find out what he really does say.
“If you were really interested in learning biology rather than promoting a teleological philosophy”
From my notes ca. 2005:
Biologists, despite frequently claiming to eschew teleological language, use such quite frequently. So it goes…
Mr. Git:
Cairns et al never made the philosophical leap you did. They knew there would be a natural explanation unlike your leap of faith.
Your false examples are straight out of Answers in Genesis. I see you also doubt plate tectonics. Can you blame people for thinking you’re a creationist?
You should study biology rather than philosophy thereof.
I read about this thread on 350.org. Cannot say I was surprised to find Intelligent Design supporters here.
You really cannot see the distinction between recognizing that adaptive behavior has a survival enhancing purpose so was selected for and believing that a supernatural power directs evolution in general toward some purposeful end such as humanity?
Some philosopher! Pure sophistry!
Diego Diaz said @ur momisugly November 15, 2013 at 12:48 pm
From The River Out of Eden 1995 pp 7-8
Emphasis mine. If species are irrevocably separate as Dawkins states, then there is no possibility of HGT.
Diego Diaz said @ur momisugly November 15, 2013 at 1:09 pm
Since I have never read Answers in Genesis, I have no idea whether your statement is true or not. Please quote my words and provide a link. FWIW I am an agnostic on many things, most especially those I have never investigated. This is the very opposite of the faith you are alleging I possess.
You did not search. Afraid of truth?
He has written and spoken much on HGT. Your conclusion from out of context citation is unwarranted.
Some philosopher! No logician!
Diego Diaz said @ur momisugly November 15, 2013 at 1:17 pm
Please quote my words where you claim that I said I believe that a supernatural power directs evolution.
I did write above “The [God done it explanation] has, as you correctly point out, zero supporting evidence and any Bayesian argument requires ‘Bayesian prior assumptions that make the theory globally inconsistent and absurdly unlikely from the point of view of any sort of axiom of parsimony’.” Emphasis added.
AiG misuses Cairns, same as you, without citing SOS research.
1. It’s all just down to an amazing long streak of lucky events. We Know The Truth. Live with it. [neoDarwinists]
2. This is not the explanation for evolution. Evolution involves whole nucleotide sequences (horizontal gene transfer). [Panspermia advocates/Margulis et alia]
3. There is some undiscovered mechanism/mechanisms operating to skew the odds. That is, the process is not random at all. [Prigogine et alia]
4. God done it.
This list is hardly exhaustive, and it isn’t clear at all how much luck is needed. I’m a bit of an expert on evolutionary algorithms, which are important in all sorts of optimization theories and search algorithms because they are literally the most efficient search algorithm we’ve been able to find for high dimensional spaces. Also (although my previous reply may have given the wrong impression) I’m not advocating only mutation plus natural selection, although that is a rather powerful process. There could be all sorts of ways that the dice are loaded, if you will. The world’s most successful species very probably evolved to evolve better than that — the algorithm itself is subject to tuning.
I’m very fond of Prigogine’s general work, not specifically in the context of evolution. But it is indeed the case that there can be all sorts of things that skew the odds while still remaining “random”. They simply aren’t random on the original oversimplistic picture of mutation plus natural selection. As a single example, it is now known that species hybridization is not rare or nearly impossible, it is commonplace. Hybridization plus mutation plus selection is one of many possible skewing mechanisms.
None of this affects the direct evidence we have of evolution in the form of a fossil record that — highly incomplete as it is (the odds of a dying animal being fossilized are poor indeed) — paints a clear and compelling picture of the evolution of species from common progenitor species, happening over and over and over again for nearly every species we look at. In our own family tree we can quite literally see where a break occurred in a chromosome comparing human and gorilla DNA, along with many other homologies:
http://en.wikipedia.org/wiki/Human_evolutionary_genetics
These genes are pure evidence for evolution. The leave lots of room (and the article discusses some of that room) for different mechanism for the changes, but there is little doubt that discrete changes occurred, and not much doubt when they occurred, to split the genetic line that became modern humans off from the line that became modern gorillas. There appear to be multiple mechanisms at work here — small changes that gradually are selected out according to benefit, and large changes that occur “all at once” (or in a comparatively short period of time), but both are arguably evolution, and neither one invokes an external intelligence making changes in Earth’s species over some 3 to 4 billion years.
This is the critical issue, not the particular mechanism(s) of evolution. Is evolution the result of reasoned activity by an external entity or is evolution a naturally occurring process of genetic change due to all mechanisms plus the inevitable sorting out of living creatures on the basis some mix of luck and fitness to survive and reproduce (and yes, I’m well aware that “reproduce” is broader than just the individual as well, while even at the tribal or familial level it is still evolution)?
The fossil record is absolutely overwhelming evidence for evolution, whether or not the evolution is precisely Darwin’s theory of evolution. Evolution best fits the observations. Evolution can be and has been observed in the lab. We use evolution as a tool — humans have evolved improved crops and animals for husbandry for thousands of years if not much longer by simply becoming an agent of natural selection, picking the biggest and best of animals to breed, picking the animals that were the safest to work with and best behaved according to our needs. Humans created the dog out of a wolf ancestor by means of pure evolution — breeding most the animals that exhibited (and continue to exhibit) the most desired traits. Perhaps, as you assert, mutation is too slow a mechanism to be able to create radical changes in appearance and behavior rapidly, but if so the differences between chihuahua and great dane are difficult to account for.
Where I teach in the summers at the Duke Marine lab I can look out of my window and see the results of a mere 400 years of evolution. The so-called “Shackleford Ponies” live in comparative abundance on the outer bank islands just outside of my back door. They are the distant descendents of domesticated horses that made it to shore when ships foundered offshore in the sixteenth and seventeenth centuries. In a mere 400+ years, they have evolved to be much smaller (the islands have a severely constrained food supply), shaggier (they have to survive the winter), and most interestingly, they have evolved the ability to survive on brackish water, as the islands have almost no fresh water and what they have is inevitably far saltier than actual fresh water. The horse’s kidneys have significantly drifted due to mutation plus natural selection in only a few hundred years. The horses have other interesting traits — they have learned to read the tides, and know how to move between the islands when it is safe to do so, for example — but they are again direct evidence of natural evolution occurring on a fairly rapid timescale even in a small population.
In the laboratory, every physician faces the challenge of successful evolution almost every day. Bacteria and other parasites are evolving resistances to the drugs and antibiotics we use to combat them. New genetic variants of diseases are constantly emerging and the successful ones can at any moment become the next pandemic. Bacteria or viruses that are not diseases mutate and become new species that are diseases — another source of potential pandemic. We observe evolution at every scale from microscopic to macroscopic, we use it as a tool, we use evolutionary algorithms in computational optimization theory, we have immensely powerful computer programs that can do statistical genetic analysis. While scientists are hardly a uniform group and are apt to disagree about all sorts of theories or the meaning of this or that bit of evidence, you will have a very difficult time finding credible scientists who don’t believe in evolution as the explanation for the diversity of species and as the best explanation for the natural history of the biosphere, even as I’m sure you can find that those same scientists are not at all in agreement on the precise details of the mechanisms involved (and could even all be right as there could be many mechanisms).
God, of course, is a universal explanation, as are invisible fairies, solipcism, sufficiently advanced (and very, very patient) space aliens, pink unicorns, etc. One can always just make something up and explain away any inconsistencies by adding exceptions. This is not very reasonable — the more exceptions and inconsistencies you add, the lower the probability is that your “theory” is correct, especially without either supporting evidence or the ability to falsify a claim on the basis of evidence. As such it can never by purely logically or empirically refuted, any more than I can logically or empirically prove that I’m not a power unit in the matrix so that everything I think I know about the world is false. However, neither can it ever be logically supported until one comes up with direct observational evidence, and so far Gods and Pink Unicorns alike are scarce as hen’s teeth. So are super-intelligent space aliens or any other non-diety “intelligences” that ID advocates will offer up to prove that they don’t really mean God, it could be any old superhuman intelligence.
This also applies to your semi-telelogical argument about the tuning of the Universe to “life” if not more specifically “human life”. There are a number of problems with this argument, as I’m sure you are aware. The first is that given direct knowledge only of the spacetime continuum in which we seem to live, we cannot speak of “probability”, as in “what is the probability of finding a Universe with the parameters our Universe has”. The answer is “unity”. Behold it. It Is. So far, we are unable to empirically demonstrate a hat full of Universes from which our Universe is drawn, nor can we postulate on the basis of any evidence a process that might have produced a “Universe tuned to life”. Worse, any attempt to do so will run into the teleological paradox. If you wish to assert that the Universe appears designed, and thereby infer a designer, that designer must be even more complex than the Universe, and hence must in an information theoretically precise sense be less likely. You are invoking a cause that is even less likely than the effect that you wish to claim is necessary because it is supposedly so very unlikely, in spite of the fact that the latter is manifestly real and the former pure supposition.
Who designed the designer?
If you postulate a designer that needed no designer, then surely it is far simpler to not bother, and postulate a Universe that needed no designer. You have the substantial advantage in the latter case that the Universe is not a postulate, it is an accomplished fact and every instant of your experience empirically affirms this. There is no evidence whatsoever for the undesigned designer.
rgb
Dr. RGB:
Glad to see a real scientist among the ignorant crackpots here, casting pearls before swine.
Good luck educating them.
Diego Diaz said @ur momisugly November 15, 2013 at 1:29 pm
I possess the following by Dawkins:
Unweaving the Rainbow, Allen Lane 1998
The Blind Watchmaker, Penguin 1998
Climbing Mount Improbable Penguin 1996
The Extended Phenotype Oxford 1989
It would seem I am unafraid of Dawkins 😉 Tell me where in those pages I may find Dawkins contradicting his statement that I quoted above.
In time it took you to comment, you could have searched “Dawkins horizontal gene transfer”. Why did you not? Scared of truth?
Oh, and I do know, BTW, that you are not arguing for either ID or God per se — my arguments are against teleological arguments in general. I didn’t even point out the usual answer to the assertion — there could even be a hat from which separated, discrete Universes (or rather, spacetime continua as I dislike using the term Universe for anything but “everything that actually objectively exists”, the existential Universe) are “drawn”, or bubble out of, and their physical parameters could be completely random but it is a simple matter of fact that there won’t be any sentient beings that evolve to wonder about all of this in any of those Universes except the ones where the parameters are just right for this to occur.
Indeed, this isn’t all that crazy an argument. Forget physical coupling constants. What is the probability that I exist, doing what I’m doing right now, in a Universe that is in exactly the state that this one appears to be (even with the right coupling constants)?
Mathematics lacks enough zeros to put before the first nonzero digit… or, unity. Personally, I prefer unity, unless and until we find the hat and some sort of random process. I might even prefer unity afterwards. What’s the probability that I just typed the phrase “What’s the probability that I just typed the phrase “”What’s the probability that I just typed the phrase “””What’s the probability….””” (recurse indefinitely)? Apparently, it was unity.
That’s the problem, those pesky Bayesian priors. Even with a hat, or urn, to draw from, you arrive at very different estimates for the probability of drawing a green ball from a large urn given only the evidence that you have drawn one green ball from it depending on how you think the urn was prepared! This makes e.g. Polya’s Urn type estimates so very interesting.
rgb
Dr. RGB:
Glad to see a real scientist among the ignorant crackpots here, casting pearls before swine.
Good luck educating them.
Dear Diego,
Real scientists try to avoid ad hominem, and there are quite a few non-crackpots on WUWT. That said, I’m sure there are indeed some crackpots, but they are more likely to be found on e.g. PSI than here. In fact, there it is more or less a sure thing. Pompous Git is actually a gentleman and a philosopher, and while we do not always agree, I would hardly call him ignorant or a crackpot. Can he be mistaken, or misinformed? Sure. So can I. So can you, if you’re honest.
So far, he has said little that I would fundamentally disagree with except create a list of evolutionary theories that I think is a bit too narrow and give perhaps a bit too little weight to the ongoing work of evolutionary biologists. We’re way past the point where evolution itself can be (sensibly) doubted, and sure, there is lots of ongoing research with all of the usual trappings of well-done science — theories, experiments, observations, simulations — directed towards a better understanding of all of the mechanisms whereby evolution works. I doubt that any competent evolutionary biologist these days adheres to the strict mutation plus natural selection mechanism, simply because there is direct evidence of a lot more mechanisms at work. It’s not my field, so I am not going to be the best one to list all of the ones that are current best belief based on the evidence, but I’m guessing any good, current textbook in introductory evolutionary biology would have a survey, and even wikipedia probably isn’t too bad.
As he pointed out, he most definitely did not advance the God hypothesis or any variant thereof, and if he did bring up one of the neo-teleological arguments, well, even I think that the observations are “interesting” at the same time that I recognize (and pointed out) that they cannot constitute statistical/empirical “evidence”.
So please, be patient and polite. Calling people names is so very rarely productive. And bear in mind that the people who were pushing ID above are not, actually, particularly regular posters on WUWT. I don’t know what attracted them to this particular thread, but does it matter?
rgb
Lady with ID videos and man with literal interpretation of Genesis sound ignorant to me. I apologize but in Amazon Basin my work is hurt by fundamentalist missionaries.
I see this is not a good site for me.
Thank you for your excellent comments and you are right I was rude.
Goodbye.
@ur momisugly rgbatduke
We do not seem very far apart here. However, I think you may have missed my main point. Micro-evolution among dogs (for example) is manifestly not speciation, or at least not yet in any of the many accepted versions of biological species. Some speciation may be attributed to randomness, but this must take ever so much time given the size of the space of possibilities within one gene assuming one nucleotide substitution per million replications.
Now I have never invoked God as a serious explanation for novelty arising in evolution, though ever so many seem to want to attribute that to me. What I have stated ever so many times is that there are some laws in the universe to which we are currently oblivious. Any claim to a complete answer must be predicated on knowing what we do not yet know. And the only being I have ever heard of that might know what we do not yet know is God which is rather funny when you think about it.
Apropos the fine-tuning argument. It’s a bit much to attribute it to me, though I am flattered that you might think so 🙂 It’s an argument that’s actually been around a very long time, but really only gained its current status when Fred Hoyle did his work on nucleosynthesis in stars. I suspect that something akin to fine-tuning was in Einstein’s mind (for example), not just Paul Davies’. Yes, the probability that the universe has the properties it has is exactly one. But that’s IFF there are no mysterious other universes. I have an aversion to the concept of “other universes” when the universe is by definition “all that there is”. Invoking more than one “all that there is” would seem to violate simpliciter.
I have never come across any explanation of what might be counted as evidence for God. Hume’s On Miracles effectively dispatches most of the arguments believers believe are valid and true. Conversely, I have never come across any explanation of what might constitute evidence against God’s existence. The argument you present: “Who designed the designer? If you postulate a designer that needed no designer, then surely it is far simpler to not bother, and postulate a Universe that needed no designer.” is the closest we can get. But it’s argument to the best explanation for an atheist. The believers say, as did Aquinas and Aristotle, there is an uncreated creator and for them that is an argument to best explanation. IOW ABE arguments heavily depend on what one’s presuppositions are.
I find it far easier to state, as did Thomas Huxley: “I am agnostic.”
It would seem Gregory Chaitin has recently written a book: Proving Darwin: Making Biology Mathematical and I have downloaded it to my Kindle. However, the sun is shining in a clear sky for the fist time in many weeks and farms do not take care of themselves so it isn’t going to be read by me today.
http://books.google.com.au/books?hl=en&lr=&id=J8XG8Y5Ou5QC&oi=fnd&pg=PR7&dq=Proving+Darwin:+Making+Biology+Mathematical&ots=9NuHNx_Fl2&sig=SzF-NGUahndpPxZIUuk7jEKCflM#v=onepage&q=Proving%20Darwin%3A%20Making%20Biology%20Mathematical&f=false
@ur momisugly Diego Diaz
There are many different kinds of people who post here. If you stick around long enough you might learn something. I constantly do so. And if you mistook me for Janice… well, what can an overweight 62 year old male say 🙂
Hint: there are likely some cultural differences at play here. My “God done it” statement above is not good English grammar. It is in fact a probably far too well veiled reference to a Monty Python sketch in which God says in a very prissy voice: “He done it!” about a character who was murdering archbishops or some such. Silly I know, but I then am renowned for silliness…
I know you are not she nor ignorant prejudiced fundamentalist with no interest in truth like her or Genesis speaks of plasma and ions man but still this not a good site for me but thank you for saying “stick around”. I am Catholic but not crazy fundamentalist. My work is to help indigenous people and the nature of the Amazon, my home.
So goodbye to you too sir as to Dr RGB.
One other thing. I do not know the new maths book you have but there is much computer evolution already showing it works and even using evolutionary algorithms to make new software. Random more or less evolution is fact that can be demonstrated with maths.