[G Anthony]

Let us talk about the scientific method and why it is so important.

This is really a philosophical discussion pertaining to the philosophy of science. There is already a philosophy forum. So, we can avoid going off the reservation by sticking to nuts and bolts.

The key to my essay is the point made about the Null Hypothesis. Most scientific results are expressed statistically and the Null Hypothesis is supposed to be an expression of how apparently positive statistical tests could have been fooled by random noise or just plain chance. Then if a major source of noise is found or if chance is confirmed, Null is true and the alternate hypothesis is false. This is just insurance against bias: looking at the other side of the coin.

All scientists are biased, but only a few admit it. Yet, if they hew to the scientific method, their conclusions may still be trusted. Motivation is irrelevant. Even biased scientists can remain ethical.

The scientific method is a framework or skeleton upon which may be hung the musculature of such ethics.

[tomgwyther]

Is not the concept of a 'Null hypothesis' already part of the scientific method?
More commonly referred to as a counter hypothesis. For example: I could hypothesize that if I let go of an object I'm holding, it will hover in the air. The corresponding counter-hypothesis could be that it will fall to the ground. The enormous amount of experimental data available shows us that the Counter hypothesis is true. in this instance, hypothesis and counter-hypothesis are interchangeable.
More recently, scientists at CERN have data which suggests sub-atomic particles traveling faster than the speed of light, when current scientific understanding says that this is impossible. The hypothesis that stuff is traveling at this speed would indeed be considered a counter - or null - hypothesis to the universal constant.
Scientists from all over the world are studying the data to see of noise or chance could be involved.
Null hypotheses are being investigated all the time.

[G Anthony]

pantheory, on 1 January 2012 - 05:05 AM, said:

G Anthony my friend,


We aren't? It doesn't? For me they are synonyms

But seriously I think your general idea has merit in that I also think that all theories should be required to formulate null-hypothesis, at least for them to be considered a valid theory. But who in the mainstream will ever agree to this great idea? Only those most enlightened who are very rare individuals.

I agree that nearly all theories today do not meet this criteria. Although no theories can be easily proven, I try diligently to put together clear ways by which my own theories might be disproved. I believe this is one clear distinction between a well proposed theory and today's hypothesis. Realize that a mal-formed theory will probably have a much greater longevity if they do not have a null-hypothesis, and if they are not making any new consensus predictions.

Some poor examples of theory IMHO are a Big Bang beginning, dark matter, dark energy, warped space, since none of these, like astrology, can ever be proved or disproved -- IMO. All instead might be considered valid hypothesis until something better comes along

But in the same way, and for the same reasons I think you should form a null-hypothesis concerning religion for your own sake so that you may know the sad truth before we do not meet our maker. ................... In this vain don't count on Paskal's wager. Enlightenment by one's own mind in contradiction to whatever stands in the way of logic, is far above all else regardless of the multitudes that profess otherwise IMHO. Please excuse the little bit of philosophy thrown in here with good intent. --------Happy New year to all. It's only 10:30 PM here Dec. 31, on the west coast of the North American colonies -- upon this writing

with regards, Forrest Noble

Hi Forrest,

Let's keep to the thread! So the ability to formulate a Null Hypothesis is the key to crafting a theory according to the rigor of the scientific method. Not all of our thoughts require this rigor. Only the ones that we intend to publish so that others can figure out what we mean and duplicate our actions and thoughts in order to come to the same conclusions (to within some tolerance, I guess). Much that passes for science today fails when we can find no evidence that a proper Null was ever cobbled together. When it is suggested that the scientific method should be suspended in order to accept Dark Energy, I get worried.

[Klaynos]

!

Moderator Note

Moving to general science

[Ophiolite]

G Anthony, on 29 January 2012 - 02:44 PM, said:

Let us talk about the scientific method and why it is so important.

Let's do that. Why do you think it is important?

G Anthony, on 29 January 2012 - 02:44 PM, said:

Most scientific results are expressed statistically

You think? Would you provide evidence that substantiates this claim please.

G Anthony, on 29 January 2012 - 02:44 PM, said:

Then if a major source of noise is found or if chance is confirmed, Null is true and the alternate hypothesis is false.

Am I missing something? You seem to be saying that in some circumstances the null hypotheis can be proven. My understanding is that it can be disproven, or rejected, or not rejected only.

G Anthony, on 29 January 2012 - 02:44 PM, said:

All scientists are biased, but only a few admit it. Yet, if they hew to the scientific method, their conclusions may still be trusted. Motivation is irrelevant. Even biased scientists can remain ethical.
The scientific method is a framework or skeleton upon which may be hung the musculature of such ethics.

Why are the ethics of the matter important? I fail to see what you are driving at. The scientific method is blind, therefore only conclusions reasched via the scientific method need t be considered.

[CaptainPanic]

G Anthony, on 29 January 2012 - 02:44 PM, said:

Most scientific results are expressed statistically.

Ophiolite, on 31 January 2012 - 08:33 AM, said:

You think? Would you provide evidence that substantiates this claim please.

I think it is not a bad assumption... although I doubt there are many statistics of the statistics to back up the claim. We can however say that:

1. Most measured data points will (or at least should) be expressed with a margin of error. That in itself is already a statistical statement, which means that the vendor of a piece of equipment, or the scientists themselves say: "We expect that with XX% certainty the true value is within so-and-so many units from this data point".

2. And if you read a conclusion of a paper, very often the results will be summarized. Average values, or a fit will be presented. Many scientists enjoy using a statistical fit of a data set, which makes it easy to interpolate (or even extrapolate!) data.

3. The scientific method almost always will duplicate an experiment, but in practice the data points will not exactly overlap, so it is a common practice to use average values.

I don't think anyone would dare to claim that all scientists apply statistics correctly. I have seen plenty of disasters in papers where statistics were used wrongly, or even abused, to fabricate some results. But it is used a LOT.

Anyway, 56% of all statistics are wrong - including this one.

[StringJunky]

CaptainPanic, on 31 January 2012 - 11:59 AM, said:

Anyway, 56% of all statistics are wrong - including this one.

65.7643% of statistics are made up on the spot.

This post has been edited by StringJunky: 31 January 2012 - 03:18 PM