How Light Regains its Velocity After Passing Through a Denser Medium?

You are confusing the fact that just because a hypothesis within a theory is proven false that does not invalidate the theory.

Like I said, tens of thousands of experiments have been found to support Evolution as a theory as it exists today.

Let me put it this way. What if I flipped a coin 10,000 times, and it came up heads 9900 times. Now pretend somebody came up to you and said "I flipped this coin 20 times and it came up heads 9 times, pick tails". Now lets pretend there is a psalm in the bible that basically says "always pick tails, god loves tails". You're picking tails, right? I mean, based on your whole evolution argument.

The hypothesis of a theory is the theory, the mathematical expression is just the language of the proof, there is no theory if the hypothesis is invalidated, jsut a bunch of experimental data. I do not understand why something so fundamental to the concept of science gets so convoluted when it comes to biology. 1. you observe something in nature and conceive of an idea of how the mechanics underlying the system function, 2. you develop a hypothesis to explain the mechanics and a proof of the hypothesis using a logical language, the hypothesis must provide for the ability to predict with some accuracy and precision the response of a system based on the hypothesize mechanics, 3. your proof is peer reviewed for logical errors, 4. it is tested by experimentation to disprove the theory. Biological theories fall short frequently never getting past step 1 then jumping to a psuedo version of step 3, and frequently some sort of inverted interpretation of step 4 (they believe experiments can support or prove a theory). Quantum mechanics went through the 4 steps shown above, as did relativity, and generally all relatively modern Physical sciences theories. Evolution could be correct, but it lacks the foundations to be much more than a written discussion of observations and relevence until the other 3 steps are properly completed. Just because we like the idea presented in step one should not mean we automatically accept it as a "proven" scientific theory. Because as a child I used to put teeth under my pillow, and wake up to find it replaced by money. I never saw any person place that money there, thus it spontaneously appeared from some reaction of the tooth. I could go to numerous children who have observed this happening without ever seeing any person switch the tooth for money, but it only takes the one child who caught someone to disprove it.

No, a theory is a theory. A theory can consist of many hypotheses.

For instance, one hypothesis, based on the theory of evolution, was that if we evolved from apes, apes should have DNA that closely resembles our own. This was found through experiment to be true. There are 10's of thousands of examples of other experimentally proven hypothesis that support the Theory of Evolution. That is why it is considered well established correct scientific theory.

The Scientific Method is controlled learning through induction which means that many minor observations add up to a probable whole.

The technical term for what you're doing is called Observational Bias.

However, you will believe what you want to believe and there's really nothing that I can say to convince you otherwise, so what I'm doing is basically wasting both of our time.

And there you again demonstrated previous point by statement one "hypothesis based on..". You have actually started a process for developing that can lead to another sub-theory. A hypothesis is the basis of development of a theory. From your hypothesis based on..." you could proceed to develop the rational argument, proof, to justify and support the hypothesis, then have the proof peer reviewed, then have experoiments conducted based on the accurate and precise prediction that could be derived directly from that theory. However, all theories must start with some initial hypothesis. the idea that a lot of observations compiled together as the basis for a theory is not scientific method that is just the basic method of developing any common theory, that is how they develop theories about blood letting, electroshock therapy as a solution to repair all mental illness, the four elements, and any number of very poorly developed ideas. Without rigorous testing processes specifically designed to disprove a theory and a strong rational proof, you just have a hypothesis or concept. You always have to test for a failure in a theory, not where it works for the limited body of evidence collected to this point (which is how the legal system works, and it is never wrong).

Observational bias is only one form of corruption of the experimental testing process, not the entire process. However, many soft "science" do utilize a system where if they find any evidence that supports their hyothesis they claim it a theory, and this is why at some point in the future even though in some cases people may not really look, someone discovers a case that disproves the "theory" and it is modified. this is also why soft "science" theories are written in a ambiguous language, much like legal decisions, so they can be modified at a later date to account for any evidence that may disprove them. If one experiment showed that the path of light from a distant source did not bend as predicted when passing in close proximity to a strong gravitational field relativity would be invalidated. You can predict the what is supposed to happen based on a theory.

Qualitative terms like closeness are relative. Frequently they are also quantified after they get results to fit their findings, technically all life on the planet is relatively close based DNA analysis. How close were they predicted they should have been prior to testing, what was an insignificant difference? How do you know a theory is disproven, if the line is mobile to fit the theory. Bear in mind a theory can be reasonable and not be a scientific theory (though absolute terms like correct do seem to used quite frequently and sternly in soft sciences for some reason), simply because it did not go through the proper process.

Let's not use the terms hypothesis or theory. Lets just describe the process.

The Idea

Let's say I observe nature for 30 years and read nature books and get an idea. My idea is that animals are occasionally born with or otherwise have their DNA altered to produce an inheritable defect which in turn gives the defected animal a better ability to survive. When the environment is such that the owner of the birth defect has a better chance to survive, the defect spreads among the animal population through higher odds of reproduction (since it helps it's owner survive longer on average). Given enough time, if the environmental conditions that favored the defect remain the same, the defect is spread through breeding and is no longer a defect but rather the norm. Now I say that, given billion of years, all the variety of animals (and other organisms for that matter) are the result of this defect selection process.

It's a nice idea. How do I scientifically confirm it?

Well, it's not the sort of idea I can prove through deduction. One will have to use induction. Now one can never prove something through induction, one can only prove that something probably is. However one can prove that something probably is to a pretty high degree of certainty. Let's try to do that.

So how do I prove the idea above to a high degree of certainty inductively? Well, we should test the idea. Of course the idea above has too many aspects to test all together all at once, so we will have to test it piece by piece. For instance:

The Tests

Is the DNA of animals with like traits closer than those with unlike traits? For example, are mammals' dna similar to each others than to reptiles? Are plants dna closer to each others than animals? Experiments must be done to find out. There are millions of potential tests here. Do any of the tests prove my idea? Nope. But they could support it. After many thousands of tested examples, the test is shown to support my idea. Awesome. But that is not enough, that is just one aspect of my idea, I must test others.

Is the Earth old enough to account for the variety of life? We do a multitude of experiments and find it is! Does this confirm my idea? Nope, but my idea seems to based on good assumptions so far. Let's keep testing.

If my idea is correct, then traits we observe should have a traceable line in the animals we observe them in. In other words, if man evolved from earlier monkeys, than a trait such as walking upright, which requires specialized hips, should be observable gradually over time in the fossil record. There are literally millions of examples among different animals that we can test to see if this is true. Do these unseen yet predicted fossils, when found, prove evolution? Nope. But combined with the other tests above it's starting to feel like we are on the right track.

So I continue this process. I think of ways to test my idea. Over and over I do it. Each on a different aspect of my idea. A multitude of times. My friends come up with ideas and test them. My enemies come up with ideas with respect to my idea and test them. The vast majority of tests confirm my idea.

So we say my idea is pretty much proven through induction. That if it were wrong, then among the tests above that wrongness would have been exposed through sizable occurrences of contradictions.

Now, after literally 10s of thousands or different types of tests of my theory, a theory tested literally over a million ways (because each test could consist of thousands of varieties of that type of test (monkey to man, dinosaur to bird, etc.). Now a guy comes along.

The Guy

The guy doesn't like my idea. The guy doesn't look at all the tests done, the guy looks at only the tests that seem to contradict my idea. Now there are only a few of those tests and they have been thrown out for a multitude of reasons. The experiment was poorly done, the test was poorly constructed (too many variables), human error, the results when retested are not reproduceable, etc.

The guy says "AH HA!, you are ignoring this evidence that contradicts your theory". I respond, there have been millions of tests that confirm what I say and only a thousand that seem to contradict it. Those that contradict it can be explained as statistical aberrations of testing, yet you insist in using only that data and ignoring the vast ocean of data that supports my idea. I tell the guy he has an observational bias.

I use the following example to illustrate the mistake the guy is making:

The Example

I tell the guy "imagine 100 people flip a coin 10 times each". Now imagine you add up the results of all the people. You now have 1000 coin flips. Let's say that 907 of the coin flips come up heads and 93 come up tails. The conclusion is the coin is biased towards heads to roughly 9/10 odds.

The guy doesn't like biased coins. He chooses the coin flips three out of the 100 flippers because they seem to contradict the idea of the biased coin. They have results of A=(5H, 5T), B=(8T, 2H), and C=(6T, 4H). The guy says, look at B's flips. How can you say that the coin is favored by 9/10 odds for heads when he got 8 tails in 10 throws. I point out that it is statistically possible for a biased coin favoring heads 9/10 to come up tails 8 times in 10 throws. That is why we do so many throws. The more throws you do, the clearer the true bias.

The guy smiles and says, you are ignoring the data of thrower B. I say, oh no, we've included the data, it just is marginalized by the rest of the data. I tell the guy he's using observational bias. He doesn't agree.

Am I wrong?

Nice job! (I wanted to give a GA but I ran out.)

I too have had cases in the past in which I could not award GAs/OTs, nor even post replies; but not recently. The language is, "You or someone else on the network...."

I have suggested that the number of allowable votes/replies be increased, but I am not sure if that is where the problem lies. (No one has explained this, so I remain in the dark.)

I was able to give Roger a GA for post 99.

Re: The language is, "You or someone else on the network...."

Yes, thanks!

I would like to know how many GAs/OTs we are entitled to per day, so I could judge whether it is me or somebody else on (my) network. Yesterday I meant to count, and then lost track.

I would echo the request for more (I meant to make that request--haven't figured out the best place to make it yet).

Until then, when I find something I particularly think is a GA, I'll do what I did above--that is, post something saying I would have given a GA.

Sometimes I give GAs and OTs just to do what I think part of the spirit of them was intended--that is, to guide future readers toward good answers and away from things that would waste their time.

I don't really care about evolution being right or wrong, just that the approach is not sufficient to form a scientific theory at this point. As you stated ityou can never prove it, but you stated earlier it was correct. How could you know it is correct yet unproven? Scientific theories are not totally fluid, and should not be or you have no capacity for predicting outcomes based on the mechanics of a system. Quite simply , and clearly stated above, you appear to believe that if hypothesis or idea has more observations reported that support them than those that refute them they must be accepted scientific theories, which is incorrect. Also as indicted it is a idea based on apparently good asumptions, that alone doesn't make a scientific theory. A proof is a relatively straight forward process to bring a hypothesis into a theory.

note: inductive reasoning while reasonably valid, afterall it is the approach most used by engineers, is not the model for the modern scientific method, deductive reasoning is (this could probably be found nearly anywhere even a simple internet search though a better more detailed definition might be found through a search of APS or ACS resources).

It is fully possible that the complexities of the systems analyzed are beyond the analytical capacity of those studying them (afterall, this is not what biologists are trained to do) and thus they can not take that extra step to move from a hypothesis or idea to a theory, and if the theory is correct it should be irrefutable, if it is flawed then some reproducable experiment properly designed will disprove it.

What specifically with respect to my comment (#99) do you object to? Where is the flaw in my process described in (#99)? Please express the flaw that you see in (#99)plainly so that I might respond.

For instance, I don't use the terms "theory" "hypothesis" or "scientific method", yet your response to my comment #99 talks about all three as if I did. It makes me feel like you weren't responding to #99 at all, but rather just continuing your line of thought.

So please respond to #99 specifically. What is the flaw in what I present in #99?

Oh sorry, I guess I will simplying previous discussion to line item responses.

1.Error in this premise is, if you can not use deductive reasoning to develop a theory, then it doesn't not conform to the scientific method since deductive reasoning is the basis for the scientific method. A theory not developed through application of the scientific method is not a scientific theory, it is just a theory.

You are talking about the Scientific Method, theories, etc. I did not use the term "Scientific Method" or "theory" in my comment #99. Can you be specific about what it was in comment #99 that you object to? What part of the process I described in comment #99 to determine the accuracy of my idea was wrong? Perhaps you feel my process in #99 was ok? If so please say so. If not please describe what specifically was wrong with it.

3. If any evidence that can not be discounted for the error identified above refutes a idea mean then that it can not be a valid scientific theory, at least within the range of applicability tested. No matter how many tests have been conducted to support your idea. Toss a coin 15 times it lands on heads and 1 time on tails, doesn't validate a idea that coins only land on heads, it doesn't even validate that the coin has a 15:1 probability of landing on heads. The 1 tail invalidates the idea.

Experiments (Tests) are performed by people and people are fallible. An experiment can be hidden variables, measuring error, etc. If out of 100 experiments 98 result and 2 produce a conflicting result, it is unreasonable throw out the hypothesis.

The flaw to your argument is your assumption that experiments are always preformed perfectly. They are not. If they were, what you are saying is valid. Since they are not, one must do each experiment many times to achieve statistical certainty.

However, throwing out results with out finding some flaw in the experiments and justifying the discounting of those results simply because they appear to be stantistical outliers that may refute a hypothesis is exactly consistent with the scientific method, which must be employed to develop a scientific theory.

Any test results that refute a theory and can not be shown to have specific errors, simple refute that theory and invalidate it at least within some range of applicability. If you identify an error in a experiment then you can justify its removal but must maintain some reference to the existance of such experiments and the reason for discounting them. Statistical favorability alone does not justify something as a scientific theory, though it can and does justify other fields theories frequently. This is just a standard process for accepting scientific theories as reasonable. This is also why a theory must be unambiguous, otherwise any experiment could be interpreted to test an ambiguous theory. Use of languages such as english allow for individual interpetations and very poor expression of the logic needed for a rational argument free of substantial ambiguity. (There are some cases in math that could be ambiguous, though far less than english). We conduct multiple experiments to remove perceived random errors due to things like limitations on our analytical caapcity, rounding erros and such, not to remove human errors. We require experiments be reproducible for two parts, to identify possible recurrent or human errors that may have occurred, and to verify that the experiment truly represented the mechanics of the system (if it did the deviation from the theory would recur under the same experimental conditions). Statistics are applicable to each test itself solely to remove random errors. If 98 test produce positive results out of 98 attempts then maybe the theory is acceptable, or not, depending upon if the experiments were truly devised to test the theory for failure and if 98 out of attempts of anything actually represented the absolute final results. Any 2 failing results indicate a flawed hypothesis, unless you can show where the experiments were flawed.

Please point out the flaw in the following reasoning:

1. Man is not perfect
2. Experiments are performed by man, either directly or indirectly

Given 1 and 2, Experiments are imperfect.

3. Experiments test hypotheses
4. Experiments are imperfect

Given 3 and 4 No single experiment can test a hypothesis perfectly.

5. No single experiment can test a hypothesis perfectly
6. An experiment contradicts a hypothesis

Given 5 and 6, No Single Experiment that contradicts a hypothesis invalidates the hypothesis perfectly.

Thus many experiments must be done and the results of the experiments must be viewed statistically in order to confirm or reject a hypothesis.

Do you agree, RCE? If not, what is the flaw in my reasoning above?

Roger,

Our flaw is thinking that certainty/uncertainty has something to do with statistics. What we apparently need are infallible absolute truths that come from.... I almost wandered into dangerous territory, forgive me.

That really is what it comes down to.

I couldn't resist.

RCE appears to understand English. However he doesn't answer our questions or acknowledge a reasoned argument. This contradicts our theory that he understands English. So RCE does not understand English at all.

_{Now I'm out of here.}

If an experiment is flawed that is why the experiments must be reproducably, i.e. cold fusion. If you conduct an experiment, it can not be thrown out as an outlier becasue it refutes a theory that most other experiments didn't refute. You can attempt to reproduce the experiemnt to determine if the results differ such that you refute the experiment, and thus it is not appliable to the theory. Also, statistics does not refute non-random errors due to man in scientific experiments, it is used to eliminate random errors.If the errors are due to soemthing consistent in man conducting the experiments then you just get a non-random bias in the results, and have to have many different men conduct the "reproducable " experiment to see if they get the same results. If the error is due to somethign about all mankind, then it is a bais that statistics can not remove or even identify. It doesn't, however, validate a theory because some experiments represent statistical oultiers while other experimental results appear to fit the theory. repoducibility of experiments is a means to refute flawed experiments.

My question was is there a flaw in my logic of comment #113. Please answer my question directly. Is there a flaw in my logic? Or do you agree that no one single experiment can refute a hypothesis perfectly?

Yes, it is possible for one experiment to refute a theory, if reproduction of that experiment doesn't refute the experimental findings.

You are contradicting yourself.

You say that it is possible for one experiment to refute a theory if it is repeatable (ie, produces the same contradicting result). How does one know if an experiment is repeatable without doing it more than once? If one does an experiment more than once, than it cannot be called "one experiment".

Do you agree?

I didn't say you had to repeat the experiment to refute the theory. It must be possible to repeat the experiment though to refute the experiment, this is not a contradiction. The experiment could refute the theory in and of itself. It must be able to be repeated, not necessarily repeated. Repeating a experiment is for people to test the veracity of the experiment and potentially refute the experiment if they do not accept the experimental results. However, no matter how many time you repeat the experiment to verify it, the first experiment refuted the theory.

I could pose the hypothesis that all pigs fly, you would only have to conduct 1 valid experiment to refute that hypothesis. Now someone may want to challenge your experimental results, and they should be able to reproduce the experiment to test its veracity. I suspect most people would not question the veracity of your experimental results. However, in the end it was only the initial experiment that was necessary to refute the hypothesis.

I will ask as simply as I know how.

Can a single experiment alone refute a hypothesis perfectly?

Please answer yes or no.

Yes, as stated very clearly above, a single experiment can refute a hypothesis. There is a difference between an experiment and a data point. One single valid experiment can refute a hypothesis. This is fairly common understanding any physics or chemistry major learns. I am guessing you must have learnd this in college. The question isn't that a experiment can refute a theory, but whether the experiment is valid.

How can you know whether an experiment is valid or not without repeating it?

Any additional experiments to validate the intiial findings do not further refute the theory, they just verify that the findings are consistent. So the first experiment still refuted the theory, it is just a very good practice to independantly verify the experiments findings. Repeating the experiment doesn't have to do with refuting the theory, it has to do with the question of the validity of the experiment.

How can something that you are not sure is valid refute anything. Must it not be verified first?

verifying the experiment doesn't mean the initally experiment could not refute the theory, it just means that the intial experiment was valid. Any subsequent experiments are not testing the theory, they are testing the initial experiment. It is the findings of the initial experiment that already refute the theory, all subsequent findings just verify the initial experiment by not refuting the findings of the experiment.

How about this.

A human being can never know for sure whether or not a hypothesis is refuted using only one experiment.

Can you agree with that statement?

4. Observational bias occurs where you look for positive results to support their idea, or for ease of testing, not recording or discounting the negative results without substantiating the basis for discounting the results due to a logical error in the test, a improper assumption (even this would be problematic to discount), or a mistake made during testing (such that it didn't conform to good experimental practice and quality control). This would be contradictory to a position of throwing out any number of test results because you had substantially more positive test results.

So I guess you would be wrong to assume this would be an applicable methodology to developing a scientific theory.

You are incorrect. Observational bias is when you selectively choose a specific subset of data rather than using the whole of the data available. This method uses statistical anomalies (my 8 tails 2 heads example) to incorrectly invalidate a hypothesis. Statistical anomalies are present in all data, so this method can be used to invalidate all theories. F=ma, Relativity, QED, you name it.

I dread entering back into this because I still believe that you are not acting rationally but I feel that I must.

Where do you get the idea that the scientific method does not use inductive reasoning. In mathematics inductive reasoning is frowned upon but the scientific method exists in the first place because of inductive reasoning. It is true that inductive reasoning is rarely used alone and that deductive reasoning is often used to validate an inductively reasoned theory, but the existence of inductive reasoning in the chain does not negate any scientific value. The very concept of an overlying theory of anything comes from inductive reasoning. One must first observe something that appears to be repeatable or even a trend in order to say that there can be a theory for how it or anything works. To not accept inductive reasoning is to accept only chaos.

Inductive reasoning however is not the basis for the method. Again it is pretty simple to research the scientific method. He indicated that when you hypothesize something to complicated to use deductive reasoning to develop a theory you use inductive reasoning then to develop a theory, which is incorrect. This concept of employing deductive reasoning to develop a theory is fundamental to the scientific method.

Even when I give you links that show that inductive reasoning must be part of the scientific process you still refuse to budge. You are not rational and unresponsive to my question. Bye Bye.

You're incorrect. You use inductive reasoning to develop the theory, then you use deductive reasoning to determine tests to confirm or not confirm the theory, then you do experiments (tests) and get results. Based on these results you revise your theory / hypothesis. The scientific method combines both induction and deduction.

http://atheism.about.com/od/philosophyofscience/a/ScientificMethod.htm

Yes some older scientific principles exist that have some basis in inductive reasoning (generalizations). I beleive your article actually identifies why use of inductive reasoning is inappropriate, even if there are cases where it has been employed in the past. The scientific method, however, is the idea of how scientific theories should be developed, not the commonly accepted practice of science by the general public. The use of soley inductive reasoning to develop a theory under the scientific method employing the modern models is inappropriate, but the public is not so discerning as long as the results work for now in most situations they are familiar with. Also, while inductive reasoning may sometimes employed initially to develop a hypothesis, it is not necessarily required, deduction on the other hand is a requirement to transforming a hypothesis into a theory, predicting results based on the theory, and testing the theory. A very simple and broad example of the process is shown in the attached link.

http://en.wikipedia.org/wiki/Hypothetico-deductive_model

In practice, the scientific method in many cases has very little to do with what the general public concieves of as science, and many "scientist" also. This is what creates the ambiguity between what defines the practice of science. The attempt should always be made to meet at least the broadest most general terms of the method. All of this leads to how you define a scientist then, beyond just someone who practices soem form of science.

Here are some other links on induction in the scientific method.

http://atheism.about.com/od/philosophyofscience/a/ScientificMethod.htm (The scientific method is a combination of induction and deduction feeding back upon the other)

http://en.wikipedia.org/wiki/Philosophy_of_science (It is not possible for scientists to have tested every incidence of an action, and found a reaction. How is it, then, that they can assert, for example, that Newton's Third Law is universally true? They have, of course, tested many, many actions, and in each one have been able to find the corresponding reaction. But can we be sure that the next time we test the Third Law, it will be found to hold true? One solution to this problem is to rely on the notion of induction. Inductive reasoning maintains that if a situation holds in all observed cases, then the situation holds in all cases. So, after completing a series of experiments that support the Third Law, one is justified in maintaining that the Law holds in all cases.)

http://citeseerx.ist.psu.edu/viewdoc/downloaddoi=10.1.1.131.2694&rep=rep1&type=pdf

(The first is that induction is an integral part of the practice of science and Propper and Medewar, therefore, in spite of their membership in the class of intellectual giants, are not only talking nonsense about induction having no place in science, but are committing a logical heresy by doing so.)

Actually you are looking at it backwards, they have tested it numerous times and found that it hasn't failed. You do not conduct experiments to support the theory, you conduct experiments to disprove the theory, and if they continue to fail to disprove the theory then it remains applicable. Deductive reasoning is why it is the negative, to avoid inductive reasoning. This is why any single valid experimental finding can refute a theory. That is really what is being tested, the case that there is one situation that exist within the applicable boundaries or parameters of the theory where the theory is not applicable. If you assume a theory is universally applicable within a defined set of boundaries, then it must be applicable everywhere within those boundaries. By finding one theoretical prediction that doesn't match the experimental results within those boundaries, you have demonstated that the theory is not valid. It isn't that the 3rd laws is universally true, but rather that it is universally valid until some valid experiemntal findings disproves it. You always assume a theory is valid when a rational proof of the theory is provided, until experimental data disproves it (then run a bunch of experiments to test it from every conceivable approach).

2. Tests should never be thrown out, you can identify logical errors or inconsistencies in the assumptions used to develop a test of the theory, or even flaws in the testing process and discount the findings applicability, but not throw them out.

Tests are not thrown out, they are marginalized through statistics. You problem is you don't want the experiments that disagree to be given equal weight as experiments that agree. You claim that one suggestions invalidates the hypothesis. But as I stated earlier, this assumes that the experiment was perfect and done perfectly, an unreasonable assumption. It is precisely because of the inherent error involved in experiment that a statistical approach must be used.

You wrote:"Tests should never be thrown out"

So if I'm measuring current and do an experiment where I forgot to turn on my current sensor, and get zero current for the length of the experiment, you are saying that that data MUST not be thrown out? I don't think that's what you mean, though it's' what you said.

Well really you would not have collected data, because you could not have taken any measurements. So unde such a mundane situation the measurement doesn't begin until you have the measuring device operational, you could just turn the device on, unless you contaminated the samples. What if the experiment is super sensitive to minute variations in current, and the the probe had disturbed the chemical/physical characteristics of the conductor enough to distort the results (you would then run a bad test or throw out the sample and start a new run). However, in general you should still should not thrown out any testing, document the test (or attempt at a test occurred) and indicate the error or a suspicion of error and the confidence in that suspicion, then the data can be discounted from the findings. The reason for this is very simple, it is a measure of data quality and of the whole data collection process quality. If you happen to have a recurring issue with perfoming tests where you forget to turn instruments on such that the results are obvious, what about measurements taken where you had to calibrate instruments properly or allow them to come to a specific temperature before initiating a test. All of the testing could be suspicious at that point if this was a recurring problem. also the equipment could be faulty and the next guy may not take notice, so when the test results are reviewed it could be identified that a faulty peice of equipment may have been a problem and those tests were discounted because the results may not have been accurate due to the equipment used in testing. The only reason to throw any information about an experiment out is to hide some error or data that is inconsistent with the expected or desired results. Plus as you said human make errors, so if all experimental data collected is perfect then there must be some question about the lack of occurence of human errors being documented. When I used to work as a chemist we would perform testing and then if we found we suspected we had done something wrong, like forgot to add vanadate to our molybdovanadate solution or maybe diluted the sulfuric acid twice, we would document the errors stop the procedure on that batch unless the samples were limited, and start a new batch with new codes. If i got a odd results for pH I would document it, get another pH meter test the sample, and check the questionable meter against calibration standards. That is just good laboratory QA/QC practice. Another example would be electro-optic courses in physics in college, we would frequently get bad IC chips and resistors, that would perform within range when cool but later would start heating up and start performing improperly. When this happens we documented the data if data had been measured and the chips for testing later, separated the chips (we just threw the resistors out), obtained a new chip and started again. There are so many reasons why you never throw experimental data and observations during measurents out.

I think you're confused. You're last response seems to explain why there are many reasons not to throw out data. But I never said there wasn't.

You however said that "tests should never be thrown out". And I provided an example where even you agree the data would be useless and the test could be thrown out. Thus I've proven your statement is incorrect.

Please answer the question I asked in comment #126

Actually what i said was in response to your implications in the prior post128 qustioning that data must not be thrown out. As thrown out implies no rational explanations and discussion in the findings relating to data discrepancies and why the may not have been used in the final findings. also, what I stated, failry clearly, is you never conducted any measurement and thus could not have collected data, since the instrument was not on. Until you turned the instrument on you would not have begun collecting data, unless your data was to observe the some effect of poking/clamping a probe into/on something. I would be surprised if you got a reading from a instrument that wasn't on. However, if you had say a a colorimetric test and forgot a reagent (or believe you forgot it), you would then have to document that test and note the discrepancy. Now in the process of developing the findings they could look at the discrepancy, document it and its implications (if it were not invalid) and then proceed to look at the findings if that observational results were not considered in the final analysis. Afterall, how do you know that the measurement was not actually correct. You then run confirmation testing to verify the discrepancy, unless the results are obvious such as getting absolutely no reading at all out of the instrument used.

Do you believe that an experiment can be conceived and executed perfectly?

Please answer yes or no.

Conceived perfectly, yes it is possible. Executed perfectly, probably not but may be possible. Execution is a hard one to respond yes of no to, simply because the possibility that a experiment might be conceived that is so simply and the subsequent findings so distinctive (with no intermediate or ambiguous results, as in the case of a yes or no response) that it is only executed correctly or incorrectly and there is no variation. The breadth of possibilities for designing an experiment are so broad, it is impossible to claim any absolute. However, failure to execute a experiment properly makes the experiment invalid, failure to execute a specific portion of a test properly invalidates the test results (which should have been identified during testing or subsequent analysis, and be discussed in the findings anyways). So you do something wrong in an experiment, simply put you tell the readers and explain that is the reason for not considering that invalid data in the findings. If the entire experiment is invalid, then you must perform it again, because the results would not be meaningful, document the bad experimental results (just in case there was something meaningful there that was overlooked), and provide findings related to the valid data collected. Now this process is risky, and could be subject to observational bias as I discussed before. As in, if you try to prove something you observe only what supports your hypothesis, and throw out (or don't observe) that which does not support the hypothesis without rational justification.

You presented a Non sequitur = loosely: nonsense sounding good. And I do not care for the teleological aspects.

Real science is FALSIFIABLE = 10,000 supporting tests do not count in the face of one undercutting any part of the support. That is not my opinion, but a foundation of science. That is the way science expands.

Look up in Wikipedia "blackbody radiation". The problem got solved by Max Planck by introducing quantum in the equation, laying the foundation to the quantum physics. A few years later Einsten won the Nobel prize explaining the photelectric effect on the very same basis.

A few years later Einstein became aware of a letter signed by some hundred scientists opposing one of his famous theories. His answer: why a hundred? A single one would suffice (to turn it over) with the right argument!

Then again, he was a real scientist.

That was a good century ago. Are we better now? That is quite debatable. A good decade ago an australian country doctor made the preposterous announcement, that stomach ulcers were caused by a bacterium. Ridiculous, as everybody "knew", that stress and spices cause it! He even infected and cured himself: no avail. for some years he and his academic sidekick were ridiculed. Then somebody got a novel idea, I am sure nobody stumbled upon in the annales of science: let's repeat the experiment!! Lo, and behold it was true!?! Within a few years, it became the most successful medical treatment. See Wikipedia: Helicobacter Pylory.

The science is functioning (or not) as before. Human nature is as before. And voting does not matter the cows methanic emissions (still).

First off that would not pose a scientific theory to describe a theory on the basis of the experiment meant to test the theory. That is inappropriate. If a the hypothesis that a theory is based on is proven wrong then the theory must be wrong at a rudimentary level. If a hypothesis to theory is intergral or woven throughout a theory and proven wrong then all of the theory is questionable at a minimum.

Second, you could conduct 9,900 experiments that confirm the earth is flat, only takes one good experiment to proven the flat earth theory wrong. for most applications a flat earth theory might be useful. In land surveying they use a flat earth perspective when conducting local optical surveying, the curvature just is not considered (they don't even bother to look at it, for all purposes they cosndier the earth flat within about 1 mile). This however doesn't mean that surveyors believe the earth is flat. Though some will continue to treat the earth as flat well beyond the range where the curvature starts to affect their results. 9,990 times out of 10,000 these guys get proper results, because there are so few projects that test this incorrect hypothesis to a level where there is a significant difference.

Could you be a little more obtuse, please. I almost understood what you wrote. I did have to make several assumptions about several of your undefined pronouns to make sense of your writing though to come to a plausible conclusion.

If I did understand your surveying example, you have a fundamental flaw in your argument that you are conveniently ignoring. Depending on what a surveyor is measuring and to what level of precision the data is required to be taken, the curvature of the Earth may or may not be relevant to the measurement tolerances. When I bought my home, the surveyor was contracted to identify the boundary line to within a 1/4 inch of elevation by the bank. (An absurdly tight specification IMHO but that is what the bank asked and paid to be done.) Now if some one tried to extrapolate the curvature of the Earth from my property survey results, then they should hide their head in shame. The values from my property survey does not indicate any curve to the Earth at all. However, at work with focused X-ray light traveling 60 meters of transit along our longest beam lines the microns of displacement measured by our surveyors do reflect the curvature of the Earth.

A single, repeatable experiment does not necessarily refute a scientific theory. This can be because the experiment is flawed, the analysis is flawed, the experiment was designed to deceive (Penn and Teller) or just that communication about what the experiment did and measured has become corrupted. At the same time, many times a repeatable experiment has revealed a previously misunderstood nuance of an established scientific theory that needs further exploration and explanation. Insisting that a scientific theory must remain immutable and true for all known and unknown circumstances is a level of dogma that should be reserved for religion, not science.

I'm confused. At least with respect to the question of surveying (on a curved earth) both you and RCE seem to be saying the same thing--essentially, that for short distances, the curvature of the earth is not significant and may not even be observed.

What am I missing?

Because it may not be observed for 9,990 surveys (as an example of experiments to disprove a flat earth theory), as most are performed at a relatively small scale the results would appear to support a flat earth theory. However, a very few would appear to disprove it. Thus the argument that 9,900 out of 10,000 makes a theory valid is not sufficient. It only takes one good experiment to disprove a bad theory.

With regards to the comment about high standards being a dogma more consistent with religious beliefs, I would tend to disagree. Religions do not require any rational arguments and can be based solely on a limited amount of evidence (e.g. miracles). It would not be representative of religion to hold their theories to a high standard of validation. Religion on the other hand depends on faith and a common belief (this would be based on experience and observations by the believers that appear to support their beliefs, neglecting those few that appear to disprove their beliefs as outliers or any lack of observations as supporting evidence). It is a bit disturbing that in some peoples perception the standard applied to acceptance of scientific theories should somehow be less rigorous than that applied to acceptance of religious beliefs.

Well actually; within the methodology of surveying the curvature of the earth is present and recognized. There is a thread on that somewhere ref setting up a check system for a theodolite. I.e. "They" do "bother to look at it".

As a former chemist I can tell you it very disconcerting when some biologists attempt to use DNA to prove a connection between anything. DNA supports inheritance, but not really evolution. They frequently try to use evolution as an explanation for DNA results they do not understand, but it is sort of like the dark energy argument when the data doesn't appear to match the predictions. It is the case of I know the theory is right therefore the data must be flawed and we need to add some cheat in to make it match the theory.

You can twist any DNA results to both support and refute evolutionary theory, unlike inheritance theory. When all it really can do is refute a specific genetic connection between the test subjects, which is based on inheritance/hybridization theory, not evolutionary theory.

You should really contemplate what evolutionary theory states, it is different than inheritance theory. What would that theory predict and what would be a valid test of that theory. If you find tests that do not correlate with the theory, then the theory is flawed and a new theory is required. A new theory may be based on some of the ideas hypothesize in the previous theory, but it is new. For instance how would you test a physical sciences theory prediction. If a physical sciences theory predicted that something would occur relative to a time term, such that there was a rate of occurrence, you would develop a test to verify that the results were consistent with the predicted occurence in some term of time. If you required some input for the theory to predict some output result would occur, you would apply that input and see if the predicted result occurred.

Neither one of you know anything about recent genetic testing and the results from the human genome sequencing. You both wish to repeat your ancient myths and beliefs exactly as your religious dogma demands. You both refuse to accept anything that resides outside of your preconceptions. Since you both refuse to listen to anything outside of your preconceived ideas I know that anything that I cite, quote, pontificate, explain or try to get.... (I promised myself that I'd hold my tongue this time, excuse me) Well this is for others who will listen.

Humans cannot fabricate one of the important enzymes for healthy living, Vitamin C. Most animals on this planet synthesize this enzyme through a series of DNA/RNA controlled chemical reactions from Glucose. The human DNA sequence has a Vitamin C Mutation that prevents us from being able to make Vitamin C. We are not the only creatures on this planet though that cannot produce Vitamin C, notably most of the other primates and the Guinea pig. When the Guinea pig DNA sequence for Vitamin C was examined it had a mutation that prevented Vitamin C synthesis. However, the mutation was found in a different part of the sequence than humans. When the other primate's DNA was compared to humans though the mutation occurred in the same location. As one compared the change in the sequence across the other primates to our mutation sequence, the primates closer to us had much more similar errors. The chimpanzee and bonobo erroneous Vitamin C sequences are nearly identical to ours. A nice video that ties this all together.

The evolutionary theory predicted long ago that species with a common ancestor that survives with a mutation will carry that mutation. This prediction came long before DNA sequencing was possible. So the evolutionary theory has a prediction realized. This is just one of thousands of other predictions the evolutionary theory realizations.

So you are implying we evolved froma guinea pig, as we have a consistent mutation?

Also, I really have no religious dogma, and am aware of the results of the human genome project, as would any other chemist with a biochemistry emphasis. I actually just dislike the religious dogma associated with certain biological concepts, even if they seem like a reasonable idea. To some people a flat earth seems like a reasonable idea based on their experience and knowledge base.

I knew it. Even when I carefully craft precise complete English sentences and cite others work that reiterate my information you refuse to understand what I've said. You refuse to grasp information that disagrees with your preconceptions. I can accept the idea that you or anybody will disagree with a concept that I believe to be true. I can work and discuss ideas with people who understand my perspectives. But you cannot understand a concept that disagrees with your bias. You have reached the level of troll here.

Hmm, you brought up guinea pigs as if there was a more primitive relationship which has become divergent over time. (Of course that would not explain everything else that supposedly evolved in between). I did however note that you must resort to simple name calling to support your position, rather than finding the evidence that you seek with regards to evolution and not other theories that may be have a relationship. You can believe, i suspect that chemistry will eventually demonstrate to a scientific level of something along the line of evolution may be a contributory factor in speciation. however, as a former chemist, i can tell you the problems associated with the current theory typically mean chemist have to approach the subject matte from another direction and general try to avoid any specific involvement in the topic for research. On the otherhand, inheritance is a much more elegant theory that can be readily tested. You should definitely bear in mind there is a distinctive difference between theories of evolution and inheritance, but not between inheritance and genetics. Inheritance/genetics are like the maxwell's theories of biology. While evolutionary theory is more like alchemical concepts of transmuting lead to gold (bear in mind the alchemist were also actually valid in their belief that lead could be transmuted to gold).

Please learn to read. Reading involves following the meaning of the words on the page, not just being able to pronounce the sound of the words.

redfred said:

"However, the [guinea pig] mutation was found in a different part of the sequence than humans."

If you had understood the meaning of the words as you read, you could not possibly have come to your drawn conclusion.

redfred's response: "You have reached the level of troll here" was well founded.

Re: So you are implying we evolved froma guinea pig, as we have a consistent mutation?

From the post you are replying to: However, the mutation was found in a different part of the sequence than humans. When the other primate's DNA was compared to humans though the mutation occurred in the same location. As one compared the change in the sequence across the other primates to our mutation sequence, the primates closer to us had much more similar errors. The chimpanzee and bonobo erroneous Vitamin C sequences are nearly identical to ours. A nice video that ties this all together.

It sure doesn't sound like he's saying that. Nor is he saying we evolved from the chimpanzee or bonobo.

I don't know why the two of you (RCE and redfred) are seeming to argue--it seems you both believe in evolution (at least, for some definition of evolution).

Are either one of you arguing from a religious standpoint?

To be honest I do not believe in evolution, I just suspect it is the most plausible idea we currently have, someone just needs to take a unbiased scientific approach not tainted by others evolutionary suppositions and develop a theory. You are correct it does appear he is arguing simply for a simple religious position, the "belief" of or faith in evolution as the one answer. Some people believe so absolutely and keep adapting their beliefs to account for any major discrepancies that arise (or the interpretation of the discrepancies). This is much like geocentrist theories or the past, where when a discrepancy arose, they kept augmenting the theory to fit the data (of course they had a lot of evidence to support geocentist theories of the universe also).

The concept evolution today, is vastly different than original postulated. Not because of new evidence supporting it, but because of new evidence that refuted it. However, there is a certain belief amongst some beyond a shadow of a doubt that it can not be wrong. What I am suggesting is that there maybe any number of plausible alternatives while support inheritance/hybridization and genetics. Thus the alchemist example. Just because they believed (maybe not as zealously) that lead could be turned into gold, doesn't mean their theories based on that hypothesis were correct (even if the initial hypothesis was plausible) and apparently many of their theories for their time based on the knowledge available seemed plausible (they never seemed to work the way they predicted, so they always just amended it to account for the discrepancies and failures).

Remove the religious belief element from evolution, start fresh and maybe science will step in for existing biologists and provide a adequate rational proof with substantial valid experimental testing that doesn't invalidate the the theory.