The scientific method

[ronthepon]

I came across a so called scientific method today.

 

Here it went:

1- Observe a new phenomena.

 

2- Develop as many hypotheses as possible.

 

3- Consider each hypothesis separately to get any predictions the hypothesis makes.

 

4- Experimentally test the hypotheses.

 

5- Get down to the most likely seeming hypothesis.

 

6- Develop an appropriate theory.

 

Any opinions?

[HappytheStripper]

Yup .. now all you need to do is choose your phenomena .. Good Luck !!

 

Any thoughts ??

[Tormod]

2- Develop as many hypotheses as possible.

 

Sounds a bit strange to me. You wouldn't develop as many hypotheses as possible. Rather you would come up with plausible explanations and start researching the most likely of those.

[Eclogite]

In the modern implementation of the scientific method peer review is also central.

At stage 1 the observations need to be confirmed by others, or to be gathered by techniques that have been well validated.

At stage 2 and 3 the hypotheses must be reviewed and informally assessed by colleagues.

The experimentation in stage 4 needs to be validated by colleagues

The report on the intial hypotheses and experimental results must be reviewed by the editorial bodies of the journal publishing the results.

The published research must be assessed by others working in the field

The work must be further reviewed, assessed, modified and validated or rejected by other parties, some of whom must necessarily repeat the experiments.

[Southtown]

In the modern implementation of the scientific method peer review is also central.

At stage 1 the observations need to be confirmed by others, or to be gathered by techniques that have been well validated.

At stage 2 and 3 the hypotheses must be reviewed and informally assessed by colleagues.

The experimentation in stage 4 needs to be validated by colleagues

The report on the intial hypotheses and experimental results must be reviewed by the editorial bodies of the journal publishing the results.

The published research must be assessed by others working in the field

The work must be further reviewed, assessed, modified and validated or rejected by other parties, some of whom must necessarily repeat the experiments.

In theory...

 

http://www.sciencenews.org/articles/20060506/note11ref.asp

http://jama.ama-assn.org/cgi/content/abstract/295/14/1675?maxtoshow=&HITS=10&hits=10&RESULTFORMAT=&fulltext=bias+peer+review&searchid=1&FIRSTINDEX=0&resourcetype=HWCIT

http://naturalscience.com/ns/articles/01-08/ns_slee.html

http://www.springerlink.com/(pk2ak045sixryl45abvqsqax)/app/home/contribution.asp?referrer=parent&backto=issue,5,5;journal,161,162;linkingpublicationresults,1:101589,1

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1114539

[Eclogite]

Nowhere, Southtown, in my post do I claim the scientific method is foolproof at the outset.

The method is, however, far superior to one in which no peer review at all occurs.

The method works over time. The peer review that leads to the downfall of a concept may occur a couple of decades after publication.

 

The links you have provided demonstrate the viability of the method by showing how it is self critical, a cornerstone of any system that strives for objectivity.

[Turtle]

I came across a so called scientific method today.

 

Here it went:

1- Observe a new phenomena.

 

2- Develop as many hypotheses as possible.

 

3- Consider each hypothesis separately to get any predictions the hypothesis makes.

 

4- Experimentally test the hypotheses.

 

5- Get down to the most likely seeming hypothesis.

 

6- Develop an appropriate theory.

 

Any opinions?

 

We are currently demonstrating it in the Strange Numbers thread. The Scientific Method in Action. Use it or loose it. :hihi:

[Southtown]

Nowhere, Southtown, in my post do I claim the scientific method is foolproof at the outset.

The method is, however, far superior to one in which no peer review at all occurs.

The method works over time. The peer review that leads to the downfall of a concept may occur a couple of decades after publication.

 

The links you have provided demonstrate the viability of the method by showing how it is self critical, a cornerstone of any system that strives for objectivity.

It is inanimate. A toy to be manipulated in anyway that the weilder sees fit.

 

The only way to be objective is to make ourselves be objective. The behavior of others is out of our control.

[Eclogite]

The only way to be objective is to make ourselves be objective. The behavior of others is out of our control.

We cannot make ourselves be objective. By providing multiple checks and balances as described in my previous posts we can approach objectivity asymptotically.

[Southtown]

Forgive me for being meticulous, I'm not trying to pick a fight. But it seems to me that "providing checks and balances" will do no more that give others an option to utilize them. If you meant 'we can approach objectivity by utilizing checks and balances' then I would say you see my point.

[Michaelangelica]

As I look at scientific method, clinical trials, evidence based medicine, reviews etc I am becoming increasingly sceptical of the whole process.

I despair that we will ever approach truth and fear we will always be prisoners of our own prejudice

 

 

Social scientists have tended to be wary of the reductionist approach of evidence-based medicine and have wanted a much broader range of information to be admissible. Evidence-based medicine has been at its most confident when evaluating drug treatments, but many interventions in health care are far more complex than simply prescribing a drug. Insisting on randomized trials to evaluate these interventions may not only be inappropriate, but also misleading. Interventions may be stamped “ineffective” by the hardliners when they actually might offer substantial benefits.

 

Then there is the constant confusion between “evidence of absence of effectiveness” with “absence of evidence of effectiveness”—two very different things.

 

Finally, even some of the strongest proponents of evidence-based medicine have become uneasy, as we have increasing evidence that drug companies have managed to manipulate data. In the heartland of evidence-based medicine—drug trials—the “evidence” may be unreliable and misleading.

 

All this doesn’t mean that evidence-based medicine should be abandoned. It means, rather, that we must never forget the complex relationship between evidence and truth.

 

<br style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0px; padding-top: 0px; padding-right: 0px; padding-bottom: 0px; padding-left: 0px; "><br style="margin-top: 0px; margin-right: 0px; margin-bottom: 0px; margin-left: 0px; padding-top: 0px; padding-right: 0px; padding-bottom: 0px; padding-left: 0px; ">Read more: Evidence: A Seductive but Slippery Concept - The Scientist - Magazine of the Life Sciences http://www.the-scien.../#ixzz17idKUrSE

http://www.the-scien.../display/57832/

 

SEE Also

http://www.the-scientist.com/blog/display/55671/

http://www.the-scientist.com/blog/print/55878/

http://www.the-scientist.com/careers/article/display/55042/

http://www.the-scientist.com/blog/display/53667/

[Ken]

In theory...

 

http://www.sciencenews.org/articles/20060506/note11ref.asp

http://jama.ama-assn.org/cgi/content/abstract/295/14/1675?maxtoshow=&HITS=10&hits=10&RESULTFORMAT=&fulltext=bias+peer+review&searchid=1&FIRSTINDEX=0&resourcetype=HWCIT

http://naturalscience.com/ns/articles/01-08/ns_slee.html

http://www.springerlink.com/(pk2ak045sixryl45abvqsqax)/app/home/contribution.asp?referrer=parent&backto=issue,5,5;journal,161,162;linkingpublicationresults,1:101589,1

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1114539

 

:unsure: In my discipline I have never heard of non-blind peer review. Why would anyone follow such a procedure when on the face of it there exists a possibility of bias?

 

It is essential that the person be separated from the question, the method, the actual data, and the final interpretation. And that approach works well, at least to those who understand it. :)

[HydrogenBond]

Let me give an example of how an empirical system can fall short of reality, unless reason has the final say. This is just an example for demonstration purposes and is not a theory. Although insisting it is a theory, for the political appeal, will be more conducive to getting resources for the illusion.

 

If our modern culture people are getting heavier. The theory is, chocolate is a gateway food to overweight. If we do this empirically, I am banking on the fact that chocolate is very tasty and because it does tastes so good and is readily available, most people would have tried chocolate at least once. I have logically inferred this stacked result, but added an empirical twist. If we do a blind cross sample of the overweight people, I would bet that chocolate was tried at least once, by people who are overweight (80% or better).

 

At the rational level, one can see this is being clever. But if we put aside reason and stick to blind empirical testing, it is a good empirical theory that can be verified.

 

If we add the buzz word "risk", it works even better. The fear of risk can be used to narrow the ability to reason, so the rational clever is harder to see.

 

Another problem with empirical science versus rational science, is the disproportionate amount of resources needed to do each type of science. With reason, one uses the brain, which is free. With empirical it is hard to do anything without access to the resources needed to conduct the needed experiments. As such, by simply shifting resources, one can shift the preponderance of the empirical data. As a modern example, if we do a research resource balance, for and against global warming, I would guess the acceptable science is where the money is; cart before the horse. If you shifted the resources to the other side, empirical science would also shift, simply because the rules of protocol, by being resource intensive, would mean the other side is better able to generate empirical data.

[Ken]

Let me give an example of how an empirical system can fall short of reality, unless reason has the final say. This is just an example for demonstration purposes and is not a theory. Although insisting it is a theory, for the political appeal, will be more conducive to getting resources for the illusion.

 

If our modern culture people are getting heavier. The theory is, chocolate is a gateway food to overweight. If we do this empirically, I am banking on the fact that chocolate is very tasty and because it does tastes so good and is readily available, most people would have tried chocolate at least once. I have logically inferred this stacked result, but added an empirical twist. If we do a blind cross sample of the overweight people, I would bet that chocolate was tried at least once, by people who are overweight (80% or better).

 

At the rational level, one can see this is being clever. But if we put aside reason and stick to blind empirical testing, it is a good empirical theory that can be verified.

 

If we add the buzz word "risk", it works even better. The fear of risk can be used to narrow the ability to reason, so the rational clever is harder to see.

 

Another problem with empirical science versus rational science, is the disproportionate amount of resources needed to do each type of science. With reason, one uses the brain, which is free. With empirical it is hard to do anything without access to the resources needed to conduct the needed experiments. As such, by simply shifting resources, one can shift the preponderance of the empirical data. As a modern example, if we do a research resource balance, for and against global warming, I would guess the acceptable science is where the money is; cart before the horse. If you shifted the resources to the other side, empirical science would also shift, simply because the rules of protocol, by being resource intensive, would mean the other side is better able to generate empirical data.

 

You present at least two independent issues, in my opinion.

 

The first essentially deals with what is commonly known as the third variable problem. Since no investigation could possibly include all pertinent, or potential, factors the methodology directs research to limit the number of variables studied at any one time. There are statistical techniques that allow for multi-variate analysis (from Chi-square tests of independence through Analysis of Variance to Multiple Regression Analysis and beyond) but most lay-people only are aware of uni-variate tests - zero amount of some variable compared to some finite amount. The notion of "control group" is broadly understood, the many ways in which variable-control is possible is not.

 

I should note that in your example you mis-label chocolate. Chocolate is at the very top of the food pyramid - it stands as one of the major food groups. :P

 

On your second point, I would reject "rational science" as a part of science. In my scheme this would fall into the method of Philosophical investigation not Science. The inescapable, fundamental, bed-rock of Science is the observation. Without data there can be no scientific investigation.

 

I agree that resource allocation plays an important role in where those investigations take place, but the situation is not nearly so bleak as you suggest. There are many sources for research funding. Often proposals that would not survive federal or large foundation grant review can gain support from smaller groups (who may have some vested interest in specific "results").

 

Science is a never-ending, self-correcting, process that looks very much like a mason building a wall. Each new course of blocks builds on the foundation of other blocks. You don't build a wall from the top down, you build it from the bottom up.

 

Using your example from the climate debate I would argue that the current broad consensus is more likely to be more correct than an opposing view - at least at this point in time. Remember, the goal of Science is to move from uncertainty to less uncertainty -- never to assume that certainty can be achieved.

 

The debate has become hopelessly polarized by politics among the general population. I generally like Al Gore but he has created a fire-storm of debate from the uninformed by his intrusion into a field for which he has no preparation or understanding.

 

Where do the opponents get their "evidence"? I find it comical that recent data is challenged based on data from the same people that identified such phenomena as climate cycles. The lay-man is incapable of sifting the evidence of past climate change - the only information to support their opposition is the evidence produced by climatologists.

 

The consensus, among those who actually know the origins of the data and have some sense of the value of the data and have applied expert knowledge to critique the data, strongly suggests that recent human activity has ADDED to the expected cycling of the climate. Scientific skepticism suggests, at this point, that the likelihood is correct -- but new data may change the consensus.

 

My take-away is that if you have to bet on the validity of the theory that the best bet is that the consensus is more likely to be correct than the opposition.