The best scientific method (Bayesian form of H-D)

Percy,Popper, in this statement, was talking about the end of the process, not the beginning, as I was. The goal is to go from subjective to objective, as Bayes allows us to do. The effort to remain completely objective results in some very ugly behavior, apparently entirely subjective in nature, from so-called scientists.Do the Google on Hypothetico-deductive.Stephen

Holmes,Do the google on Hypothetico-deductive science. Learn something. And quit giving MN credit for what H-D has accomplished.Stephen

Percy, Now, I realize that this is anecdotal evidence, but your statement is evidence that you don't understand science. That is, it confirms my earlier postulation of the hypothesis that you just don't get it.But, maybe not. What do you mean by "scientific conclusions?" Does "absence of any scientific data whatsoever" apply to the fertility study? How is that "absent?" How about the Bible Codes data. Everyone agrees that the patterns in the 94 paper are real and improbable. Are "evidence" that supports to some degree the idea that there are demons. That study hasn't gone anywhere, why do you call it absent?Now, let's do this right. What is your prior plausibility to the hypothesis that demons exist? That is, how likely is it that the, oh, 90% of the human race that is convinced that demons exist and must be dealt with, sometimes expensively, are right? One chance in a million? What number, upper limit if you like, would you intuitively or subjectively give it? In formulating the idea as an hypothesis, should we take your or the lowest value proposed, or the average of everyone's estimate, or what? Remember, that in H-D science, nothing is impossible, so the value, zero (which won't fit into the Bayesian equation) cannot be used. It would be dogmatic opinionation, anyway.Also, what do you think is the best explanation for the various actions that others call evil, genocides, tortures, serial killings, child abuse, wife-beating, self-destructive behaviors, suicides, religious nuttiness, human sacrifice, etc. Demons are proposed as a critical part of the usual explanation for these things. What is your best idea. How plausible to you assess that hypothesis to be? A priori.You don't need any evidence to have an opinion on this, according to the Bayes theorem. Be as subjective as you like. The process will peel away that subjectivity, leaving you with an objective estimate.Based on the mad dog model, I see the madness of human as very plausibly parasitically based, and since I cannot find an organic parasite, I accept the possibility of a spiritual one. So, I put the prior plausibility at about .6. This was my starting point, when I began looking at prayer experiments, etc.But what's your number?Stephen

Percy,Useful sites. Thanks. I had seen some, but some I had not. The last, that finally decides that Bayesian approaches are not really desirable, I don't actually agree with. He talks about the posterior plausibility wavering indefinately, and therefore never converging on "truth." But, this cannot happen, with strong inference. There, theory A's predictions are contrasted with theory B's. When one is not confirmed, reducing that theory's plausibility, the other is confirmed, with an increase. The ad hoc adjustment is perhaps to change the theory one holds as most plausible, or better, to merge the two somehow, especially if the more popular one was ahead, but was not confirmed in this trial. The new theory, with the advantages of both, will then be more consistently confirmed.Now, your complaints about subjectivity are always addressed concerning data being used to confirm or not a hypothesis, while my delight in subjectivity is prior to gathering or considering any data at all. On just gets an idea, and likes it, thinks it very well might be true. Is a hunch, as they say. But, when the data are gathered to test the predictions from the hunch, they are what they are, and subjectivity is not welcome. For example, in the fertility study, they made a prediction, gathered the data, and got a confirmation. But, many evolutionists here have a subjective revulsion at the idea that such a thing could be true, and immediately dive in to find something wrong with this published study. The objective, scientific thing, is to sit down at their desk and come up with a replication that would confirm it or not. Call in some peers to see if there are any flaws in their idea before doing the test, etc. Like Witztum did with the original Bible Codes study.I actually agree with most of your remarks about subjectivity, and wonder why you are making them. My current working hypothesis is that you know how subjectively you feel about all this, feel guilty about that, project it onto me, and start judging away. What Yeshua called worrying about the splinter in your brother's eye, when you have a log in your own. I struggle to understand why you feel that it is relevant that since the discovery of the benzene ring idea, which came in a dream about snakes, we have learned of sophisticated ways to detect that cloud of electrons. Sure we have. But, we got started with a dream, a subjective impression, for which there was no precedent.Now, take Asgara's quote to heart, and examine your life. This remark, for example: Now, I impugn scientists along with Kuhn, who in his study of history of science, simply observed that most are maintaining an accepted paradigm, instead of looking for the anomalous data, the reason to improve the theory. This is how I was trained to think. Sneer review, so much gas, cowardly comfortable gas-bags, all around the real scientists. MacArthur kept advising me, to hope that my ideas are proved wrong, and forced to be modified before ten years pass. Because that's good science! The ideas will still be useful, honored, but improved and improving. Of course, we all had "physics envy" looking at Newton, proved wrong but still regarded as great. (He was big into Bible Codes, by the way.)As to subjective impressions as scientific data, good Lord, no way! Who would even think of such a thing. Ok, the guys here who see some wierdness in the fertility data, have a subjective impression that somebody cheated, and conclude that it must be thrown out. Or, "no one else here agrees with" me. Well, actually, I hear that there are over 2000 persons here, and using Solomon's one in a thousand estimate, I think that we might find one or two others. So what? That's the point. They disagree, but if you read their posts, they are very busy with ad hominems, meaning that they are having problems finding what is logically wrong with what I am saying. So they have to attack the one saying it. Or stop using their subjective impressions as data.or this "for" demons, not "of" demons. Because the people praying were Christians, going in some measure we suppose by the rules for praying against demons (deliver us from the evil one) is in the Lord's prayer. But, more importantly, it confirms Jehovah's reality, is evidence for, not of, Jehovah, who in turns warns us about demons. But try to understand my H-D point. Let's say that my estimate of the plausibility for demons was .6 before the study and was .61 after. Then the study is evidence for demons, but weak evidence. I've said over and over that it is only in combination with so much other evidence for, not of, demons, that the plausibility is now so high in my mind. Bayesians can even use anecdotal evidence, which is just common sense.Enough. Thanks again for the research and the sites. That last guy is really cool, the way he handles the truth that Bayesian methods are practically impossible to apply specifically, just exist as a ideal that shapes our thinking. But I smelled a dogmatic opinionated mind-set. An inability to cope with ideas that have plausibilities in the middle ranges.So, say a trial prayer. Since you couldn't make yourself think of an number you'd accept for the plausibility for the idea of demons, and even cannot grasp what on earth everybody who talks about the beasts means, you could pray for understanding. I believe that you'd get it, in a way that would scare you. Don't worry about unbelief. Just pretend you want to ask, "Daddy, why?" E.O. Wilson does it, just for fun. Try it.Stephen

Crash, refering to evidence for vs evidence of. Evidence of is evidence that assigns a plausibility near one to an event. Evidence for is evidence that raises the plausibility a measurable degree, but could leave it moderately low.For example, the remarkable "statistically significant" correlation between exceptionally good weather and the Princeton Spring Alumni gathering, which is earnestly wished for and prayed for by potential attendees, is evidence for an effect of the prayers/wishes on weather. But, is hardly evidence of such a causative relationship.S.

P.See, I cannot fathom why you would say this: I have clearly and repeatedly asserted that evidence is to be evaluated objectively, and that subjectivity has its only place in the construction and evaluation of the prior plausibility of your hypothesis.But you keep hearing me say that subjective evidence is useful in evaluting the posterior plausibility of the hypothesis, as compared to the prior plausibility. What's the matter with your comprehension? Have you done your prayer experiment yet? You seem afflicted with a demon called wishful thinking. I have done all of these things, to my, and (Hello, God? What do You think?) Jehovah's satisfaction. Now the comprehending of what I have said is up to you. Take lots of vitamin C (mineral ascorbates), say "I choose life." and pray agnostically that Jehovah would keep demons of confusion out of your mind while you try to understand these things.S.

Brad McFall,The energy to do science is basically subjective, a desire for certain pleasures, or a fear of certain hurts. Some of the pleasures include the "AHA" experience, or the satisfaction of seeing a easily missed pattern in nature, or getting an "Attaboy" from someone you respect. But, as we know from history, both personal and published, desiring pleasures has a dark side which can muddy up life, including the life of science. The trick is to get the pleasure without the mud.Now, sophisticated techniques, especially mathmatical technigues, are impressive, generating some of the pleasures noted above. But the dark side, to me at least, is getting so deep into them that I neglect their real purpose, to enrich the data base, and ultimately, our total human experience. It's like sports. Yeah, a great professional performance makes me cheer. But, when the athlete goes on to be a jerk, somehow their great prowess sours, an least to me a spectator. I guess the other athletes still enjoy playing with them.So, in working with really brilliant math minds, I always keep asking for predictions about unknown patterns or experimental outcomes. From what I have seen, there is so much we don't know about ecology, so many many variables entering into the error term in an unknown way, that it is easy to enter into mathematical overkill. Simple math often predicts justs as well as complex mathematical structures, so we invoke a sort of Occham's razor and stay with the simplist math that gets today's job done. H-D methodology will eventually get you to theories that require more sophisticated mathmatical modelling.When I was doing this, I was very conscious that I was boring both the mathematicians and the naturalists, who were very sophisticated in their respective areas of expertise. But who were failing to find patterns or predictions in the data base we all were exposed to. It annoyed them to have me finding remarkable correlations in nature, based on "simplistic" math and the "crude" data sources. I recall taking MacArthur and R. Levins' two simplistic early math models of niche packing, which made simple but opposing predictions of which species were lost going from rich to poor communities, based on whether the resources were discrete or continuous. I tested the predictions using Christmas Bird Count data, and found patterns in the distribution and abundance of birds that had never before been noticed. All simple and crude, but the patterns were real, and eventually confirmed. Created, I thought, a job for some sophisticated math modellers and subtle naturalists.What I am saying is this: the deep joy of understanding marvelously complex, rich math arguments has a place with other mathmaticians, as math. To bring it into science, especially H-D science, however, stay as simple as possible, to generate predictions about whatever data can be gathered. Trust that you will get attaboys from non-math types, that your math colleagues will scorn. Trust that your much more sophisticated skills helped you find the simple tools that worked. Trust that, down the road, as the theory developes, you will actually find yourself challenged with a mathmatical puzzle the solution of which will amaze your once-scornful math colleagues.Stephen

S.You assert, When someone jumps off of a bridge, and moves in space to the, say, river below, tell me:Were they drawn to the river by a rubber-band like force contracting and drawing them down? (Newton's theory), orWere they traveling inertially through space, following the straight line, which happened to be bent so that it led to the river below? (Einstein's theory), orWere they pushed towards the river by Ether-pressure, generated because there was more ether on the outside of them and the river, then there was between them and the river? (Zero-point energy theory)And you, do you exist in your present form because:You were created by Someone, to accomplish some of Their goals, by artificial selection and genetic engineering, (Theory of Creation) or,You are an accident of matter, that purposesly obeys certain physical and chemical laws, such that accidental reproductive entities occurred and whenever the accidents produced more reproductive units, they increased, until here you are. (Theory of Evolution)?S.PS, I understand that it is the policy of this form to discourage unsupported assertions. Since you present no authority for your other remarks, it seems best for me to ignore them.

Quetzal,Please forgive my insults to you. They were inexcusable.Stephen

P. No, still the same free-spirited guy. Why not consider the possibility that you have something to learn from me? As you must have seen from earlier posts, since my "terrible brain injury or disease," I've had a great life, made other even greater discoveries (food chain dynamics, for example). And, when I learned that Newton, my idol, was an early scientific bible codes scholar, it encouraged me to dive right in.S.

P.Ok, here's the application of my analogy. You say, Now, I distinquish the words "of" and "for" and "confirmed" and "prove." But, you seem unable to do this. I agree that there is no evidence that "proves" the existence of demons (although the Bible Codes data gets awfully close. Be interesting to ask Witztum or Gans what they would say to that.) Or that is evidence "of" demons. But, all the evidence I cite, and the controversial nature of it means nothing to me, since the very best studies in the history of science were controversial in the same way, all the evidence I cite is evidence for, confirming demons. Now, Mammothus is getting this, I'm sure you can too, so I'll try again.I go into my back yard after a light snowfall, and see my trash can tipped over. "Drat," I say, "probably those damn raccoons again." Tipped over trash can evidence of raccoons in neighborhood. So, I look around, and Viola'! I find tracks in the snow, racoon tracks. Now I have evidence of racoons. If I hadn't looked for tracks, or there was no snow, and my neighbor came by and said, "You know, I thought I saw a racoon in our neighborhood!" I would have answered, "Yes, I have some evidence for racoons being around. My trash can was tipped over, but it could have been dogs, although in this town, there are rarely stray dogs around." In any case, in his mind and in mine, the plausibility of racoons was increased. I might have gone out some night later with a flashlight to the outdoor catfood dish, to see if I could get more evidence of racoons.In Lawrence, we have blurry photographic evidence for the presence of cougars in town. Then someone found a scat, which DNA tells us was a cougar scat. Now, we are close to evidence of cougars, although there may be suspicion that the scat was planted. The blurry photograph confirmed many people's suspicions that a cougar was around, but was not proof.I'm just reporting that, to my trained scientific eye, the data from the prayer studies (and Loehr was a scientist, and presented his studies scientifically) were evidence for spiritual beings in general, including Jehovah and Satan, in particular. I then read CS Lewis' The Screwtape Letters to get a better idea of how demons were supposed to work, several books by people who reported anecdotes about deliverance, came up with some simple prayer experiments that I could run, and convinced myself that we probably are not alone. Discussing these experiments with colleagues discouraged me from any attempt to get them published, but I had Kuhn's book encouraging me to not take that limitation as any evidence that my evaluation of the plausibility of demons was way too high.You note elsewhere that it is my job to convince you. This I do not agree with. I was taught that we each and every one had to judge for ourselves, that persuading any one else was basically impossible. All that is possible is to give counsel, not persuasion. whether the quote at the bottom of Holmes posts applies to you or to me, will be decided, if at all, when and if our souls leave our bodies and we come to judgement day. Between now and then, we each have to decide for ourselves what to do the reported experience of others.Stephen[This message has been edited by Stephen ben Yeshua, 02-03-2004]

M. When I read your opening to this thread, you asserted that I claimed that something, demons I think, could be proven or were proven by H-D science. This despite my repeated and emphasized statement that science, especially H-D science, cannot prove anything. In the Bayesian analysis, you cannot have a plausibility equal to one, but you can get asymptotically close to one. But, even if you do, the theory can be and often has been, shown to be basically untrue. Newton's theory of gravity a case in point. Untrue, but basically still useful, a good approximation. It's so beautiful, really. H-D science is never ending without being dangerous. I don't know about you, but frankly, I love doing science more than I love all the discoveries. It's the discovering process. And it never ends, no matter how plausible some theory gets, it can still be proved wrong, we still can get a whole different view. What fun. Meanwhile, if the engineers have picked up our almost certainly true theory, it will work for them, even if its not true. And after it is disproven! What a great method!The only problem is, you have to have the imagination to play with semi-plausible ideas. Like Demons. Or Bible Codes. I'm sure you are making a mistake taking the codes critics seriously. It's not that difficult!Stephen

Brad,Interesting post. One of your comments was, The Kaufmann's can be convinced, but the way to do it is to put all the math in a mental briefcase, and look into the "everybody knows" literature. As in, "everybody knows" that there are fewer species on islands. Or that Sycamore leaves are bigger than maple leaves. Or that sycamores have peeling bark, but cottonwoods do not. Ask the naturalists, how do account for that? When they answer, take their fuzzy verbal reasoning, and start digging in your briefcase for the mathematics that translates those words into symbols. Then play with the math, and see what sorts of logical deductions, especially counter-intuitive deductions, you can come up with. Go back to your naturalist friends, and ask them to help you translate these deductions into predictions about measurables. That will impress a few, and make permanent enemies out of most.Getting more Kaufmanns into the scientific community depends on processing some psychology hypotheses pertaining to denial, until you know what can be done and how to do it. Meanwhile, as MacArthur showed me by his life, it's best to anchor your social and personal reward system on something besides their reactions. Look up my article on MarArthur's life in the Annual Review of Systematics and Ecology, 1973, I think.Stephen

M.I distinquish confirm from prove, the first being the verb for evidence increasing the plausibility of a hypothesis, the last being the verb for increasing that plausibility beyond reasonable doubt.As to Bible Codes,I have studied the critics way more than you have examined the original paper and the replies to the critics.I grant that there could be clearer studies with results more heavily dependent on the hypothesis that there are demons. But any such research has to operate from the postulated ontological nature of the demons. Since they are supposed to be anti-truth, especially anti-science, this poses problems, which can best be overcome by involving God.But, I daresay that if you asked people who "believe" in demons what happened to the plausibility in their minds for demon's existence when they heard of the studies, many would say that it increased. Be an interesting study.S.

P.You ask, Evidence for something deals only with something quite uncertain, that is at least a little less uncertain because of the evidence at hand. It is suggestive, but not persuasive. It confirms, is consistent with the idea, and is itself implausible enough to not occur unless the idea under consideration, or something like it, is true. This is all in the bayes theorem. P, plausibility, given E, is higher than P given not E. If P given E is still fairly low, this state is addressed with "evidence for." If P given E is close to one, we say "evidence of."Hopes this helps.S.