What is a good sample?

[Philosophy Explorer]

This question has been burning me for a long time as it relates to knowledge and discovery.

In chemistry, when one synthesizes a purported new compound, a second derivative is made and if its properties (e.g. melting and boiling points, density, viscosity, etc.) matches up with what is expected, then you (the chemist) have most likely created a new compound.

To be specific, in studies involving groups of animals/people, how many are needed for a study to be statistically valid? Often new studies are done with differing results due in part that larger groups have been used.

What have you to say about this? Is statistics a reliable science?

PhilX

[Blaggard]

It's quite mathematical ie arbitrary, to make a good mean of anything takes fairly simple equations. The more people you have in general the better your exposition. That said of course if you open it up to peer pressure you then have exponentially more maths, and the whole thing tends to resolve itself. I could of course write down the arbitrary equations of standard deviation and statistical analysis, I wont though because maths is boring. Suffice to say the more work "you" or people do the more science thrives. The more you vacillate and armwave about a hypothesis the more it dies.


Suffice to say statistics are not science but then neither is science without statISTICS, assuming they have evidential backing.

[Philosophy Explorer]

It's quite mathematical ie arbitrary, to make a good mean of anything takes fairly simple equations. The more people you have in general the better your exposition. That said of course if you open it up to peer pressure you then have exponentially more maths, and the whole thing tends to resolve itself. I could of course write down the arbitrary equations of standard deviation and statistical analysis, I wont though because maths is boring. Suffice to say the more work "you" or people do the more science thrives. The more you vacillate and armwave about a hypothesis the more it dies.


Suffice to say statistics are not science but then neither is science without statISTICS, assuming they have evidential backing.

Due to its axioms, I do regard math as a science and as a handmaiden to the other sciences.

I'm sorry that you find math to be boring as I find you to be knowledgeable and we both know that math has many branches and aspects to it.

PhilX

[Blaggard]

I don't find maths boring but other people do, when you start using statistcal symbols, most people who don't like maths change the TV station. I've done physics and shit, I know some maths...

σ it is not interestingly friendly.

And thanks for that I find you quite interesting too, as long as you don't mind a bit of banter.

I don't think maths is a science, but I do think it is fundamental to any science. You can't get a Nobel prize for maths, you can however get hearty handshake from those who use your maths in science.

[Philosophy Explorer]

I don't find maths boring but other people do, when you start using statistcal symbols, most people who don't like maths change the TV station. I've done physics and shit, I know some maths...

σ it is not interestingly friendly.

And thanks for that I find you quite interesting too, as long as you don't mind a bit of banter.

I don't think maths is a science, but I do think it is fundamental to any science. You can't get a Nobel prize for maths, you can however get hearty handshake from those who use your maths in science.

It's too bad Newton can't get a Nobel prize in math. He was a very good mathematician.

PhilX

[Blaggard]

It's too bad Leibniz didn't either, but both of those two developed calculus for either physics or biology. Not I hasten to add for the sake of it. Evidence is king, your symbols on a page are little more than that. Prove it or be damned.

Incidentally I know a little about maths, but I am not a big fan of pure maths receiving prizes in science.

[Philosophy Explorer]

Not all math is boring to people. Proof of that are the Sudoku puzzles found on supermarket shelves and in papers and magazines (Sudoku is based on 3 x 3 magic squares).

PhilX

[Blaggard]

No I know that, but I think most people don't want to hear why the differential is rise over run on a slope of change. And hence how exactly you can close in on a perfect differential which leads to a integral.


https://www.youtube.com/watch?v=-r6NY4Kl8Ms

Numberwang night, I suggest you watch it. You will have no idea what any of it means, but then that is typical.

It does send up soduku though. Soduku has little to do with maths, it is basically trial and error, with some excited plebs thinking they done maths. Which is of course good for those who disliked maths at school. The rest of us though, it's just mental masturbation for x not y.

It's good to have people believe logic is maths, but it is not. For a start logic is true and maths is not.

[Philosophy Explorer]

Leibniz was said to be good at logic too.

When I do a study, I like to get confirmation through a second, independent way which helps to really convince me (I did this in a study that helped to get prospects to act to get carpet deodorizer using a special technique that quadrupled my sales, raising my level to triple the quota).

PhilX

[Lev Muishkin]

This question has been burning me for a long time as it relates to knowledge and discovery.

In chemistry, when one synthesizes a purported new compound, a second derivative is made and if its properties (e.g. melting and boiling points, density, viscosity, etc.) matches up with what is expected, then you (the chemist) have most likely created a new compound.

To be specific, in studies involving groups of animals/people, how many are needed for a study to be statistically valid? Often new studies are done with differing results due in part that larger groups have been used.

What have you to say about this? Is statistics a reliable science?

PhilX

You are comparing a handful of elements with groups of complex organisms, interacting in a diverse ecosystem.
A single complex organism consists of a long list of elements in a bewildering list of diverse compounds and biological structures.
You might take a 99% sample and still get an answer that would not predict the next action of he last remaining 1%.

[Philosophy Explorer]

This question has been burning me for a long time as it relates to knowledge and discovery.

In chemistry, when one synthesizes a purported new compound, a second derivative is made and if its properties (e.g. melting and boiling points, density, viscosity, etc.) matches up with what is expected, then you (the chemist) have most likely created a new compound.

To be specific, in studies involving groups of animals/people, how many are needed for a study to be statistically valid? Often new studies are done with differing results due in part that larger groups have been used.

What have you to say about this? Is statistics a reliable science?

PhilX

You are comparing a handful of elements with groups of complex organisms, interacting in a diverse ecosystem.
A single complex organism consists of a long list of elements in a bewildering list of diverse compounds and biological structures.
You might take a 99% sample and still get an answer that would not predict the next action of he last remaining 1%.[/]

Correction LM,

I wasn't comparong.

PhilX

[Blaggard]

Leibniz was said to be good at logic too.

When I do a study, I like to get confirmation through a second, independent way which helps to really convince me (I did this in a study that helped to get prospects to act to get carpet deodorizer using a special technique that quadrupled my sales, raising my level to triple the quota).

PhilX

No idea but if you like a second opinion you might be on the right track. Might being the operative word.

[Lev Muishkin]

This question has been burning me for a long time as it relates to knowledge and discovery.

In chemistry, when one synthesizes a purported new compound, a second derivative is made and if its properties (e.g. melting and boiling points, density, viscosity, etc.) matches up with what is expected, then you (the chemist) have most likely created a new compound.

To be specific, in studies involving groups of animals/people, how many are needed for a study to be statistically valid? Often new studies are done with differing results due in part that larger groups have been used.

What have you to say about this? Is statistics a reliable science?

PhilX

You are comparing a handful of elements with groups of complex organisms, interacting in a diverse ecosystem.
A single complex organism consists of a long list of elements in a bewildering list of diverse compounds and biological structures.
You might take a 99% sample and still get an answer that would not predict the next action of he last remaining 1%.[/]

Correction LM,

I wasn't comparong.

PhilX

If you were not "comparong", then why did you mention social complexity and chemicals in the same thread? Have you thought about what you are really asking here?

[Philosophy Explorer]

LM asked:

"If you were not 'comparing', then why did you mention social complexity and chemicals in the same thread? Have you thought about what you are really asking here?"

Why? Because I'm talking about sampling in two different scenarios.

(btw did you know that certain physics laws such as Boyle's law is really based on sampling?)

PhilX

[Lev Muishkin]

LM asked:

"If you were not 'comparing', then why did you mention social complexity and chemicals in the same thread? Have you thought about what you are really asking here?"

Why? Because I'm talking about sampling in two different scenarios.

(btw did you know that certain physics laws such as Boyle's law is really based on sampling?)

PhilX

Boyle's law is based on the examination of phenomena in a controlled and closed system. The results of which are replicable, and demonstrable. That's what makes it a law.
Any questions?