Validity of Statistics

[AdvRoboticsE529]

What would you say is the validity of statistics?

 

Also, if you're going to support your argument with definitions please don't, as I think you must ask yourself why you have so much confidence in the definitions in the 1st place

 

I believe the uncertainty just encourages people to not search for the *real* relationship or variables / constants, and it also encourages pointless studies that emphasizes on correlation when the true relationship (if existent at all) is *not* proven.

 

This also goes for probability, however, I heard that probability is quite applicable in quantum physics.

[studiot]

 

Posted Today, 04:21 PM

What would you say is the validity of statistics?

 

Are you asking this as someone who understands proper statistics and its proper uses?

 

Statistics makes for a better, safer world by providing quality control tools.

 

There not even a sniff of a definition.

 

Edited March 31, 2014 by studiot

[AdvRoboticsE529]

Would you be open to the idea that if the true variables and relationships are found as searched for after the abandonment of statics, this would make an better and even safer world by providing better quality control tools?

 

Have you ever had the problem that you can't prove statistics in concepts as you can for the power rule of differentiation, for example, of the area of the circle by integration, that's some pretty *concrete* proofs in pure maths.

 

Also, the argument from authority is bad, especially if the authority is ambiguous.

 

Further, statics have encouraged many studies, inflated by media to mislead people who wish for scientific literacy, such as studies which claims "public speaking is the number 1 fear"...

[studiot]

 

Would you be open to the idea that if the true variables and relationships are found as searched for after the abandonment of statics, this would make an better and even safer world by providing better quality control tools?

 

 

And if I could prove the opposite, without mathematics?

[John Cuthber]

"Would you be open to the idea that if the true variables and relationships are found as searched for after the abandonment of statics, this would make an better and even safer world by providing better quality control tools?"

Not reliably.

For example, it might take a very long time to do so and, in that time, a lot of planes might crash.

In the mean time, stats does quite a good job.

 

Als, I'm not sure that ara is very well defined in pure maths.

 

"Further, statics have encouraged many studies, inflated by media to mislead people who wish for scientific literacy, such as studies which claims "public speaking is the number 1 fear"..."

Nope, stats is a tool like any other.

You can use it to inform people and help them understand things, or you can use it to lie with (strictly- you have to misuse it).

What you are talking about there is bad journalism, not statistics.

[AdvRoboticsE529]

Consider this, I can, with the application of mechanics, determine with 100% accuracy of an object's acceleration given several known variables, which initially might seem difficult to search for, yet in the process of creating methods and tools to calculate and determine such unknown variables in order to result in acceleration, can also improve technology, examples include laser technology.

 

Or, I can set up a practical, whereby I try to create most constants, limiting to several variables, in the meanwhile, unknown to myself that there are still existent variables that I have not made into constants. I throw an object for 'x' amount, and from statistics determine the statistical false acceleration of the object.

 

Perhaps it might not be as obvious in this context, however, in grand economics, it seems extensively obvious to me the in the application of statistics with heavy emphasis on correlation in an attempt to establish a relationship, when relationship exist between your correlated variable it is usually an accident, and if the unknown variable with true relationships are altered, you are oblivious to the reason why the correlation no longer applies.

 

Consider the paper: "On Default Correlation. A Copula Function Approach."

Agreed upon by most statisticians initially, difficult to apply practically and have failed on occasions.

 

Why would we want to encourage uncertainty?

[studiot]

I like to try to keep things simple and address one point at a time.

 

I asked a very simple question in my post 4.

 

What is a very simple answer, please.

[AdvRoboticsE529]

What do you mean in post 4?

[studiot]

 

What do you mean in post 4?

 

 

You made definitive statement, I addressed it by asking a question.

Edited March 31, 2014 by studiot

[AdvRoboticsE529]

Which statement.

If you are referring to the 1st statement, I want to see if people who defend validity of statistics are open to such propositions.

 

If you can prove that statistics provides superior quality control as contrast to the control of quality by determining true relationships, then it would make me question and begin to reform maths, when I reach university this will be a priority.

As you may know, 0^0 is undefined, we need complex numbers to amend previous errors, the typical in pure maths which shows how the system could be improved upon, if you can prove that uncertainty prevails over direct relationship more work would need to go back into the basics.

[studiot]

 

Which statement.

 

Perhaps you are new to discussion forums.

So this is meant to be helpful.

 

Each time I have extracted the part of your post I wished to discuss and put it in a light blue quotes box.

Exactly as I have done above these words.

 

Then I have made my point about this extract.

 

 

If you can prove that statistics provides superior quality control as contrast to the control of quality by determining true relationships, then it would make me question and begin to reform maths, when I reach university this will be a priority.

 

Consider the following:

 

A manufacturer of medicinal pills requires to have just exactly the correct amount of the correct drug in each pill.

 

So what to do?

 

There is only one certain way to ensure that every pill has the correct amount of the correct drug and that is to test each and every pill.

Unfortunately testing destroys the pill.

 

So having tested his pills to perfection the manufacturer finds his patients dying for lack of pills, since they were all destroyed in the testing.

 

 

 

Alternatively the maufacturer can test some of the pills and hope the rest are correct.

 

But what does he do if they are not?

And how certain can he be if the test pills are OK, that the rest are OK.

 

That is precisely the situation I meant in my first post for the proper use of proper statistics.

[AdvRoboticsE529]

So what would you say?

 

Would you think that it is preferable to determine the validity of statistics... with statistics? : )

 

Or would you not be satisfied with the absence of the level of proofs found in poor maths, and that the definitions are susceptible to being amended. For example, classical definition of variance, initially requires the summation of the difference of squared values minus the mean, depending on the person you speak with, they have different reasons for why the value is squared, inflates value, results in positive value (excluding application of complex numbers), for some people, subjectively, apparently gives a nicer set of mathematical description etc.

[studiot]

Perhaps you have a poor maths (statistics) teacher.

 

A good strategy would then be ask lots of questions about the whys and wherefores, rather than branching out on your own with what appears to be an attack on statistics itself.

 

I would agree that there is a great deal of poor or even false statistics about. There are even books, radio and television programmes about that very observation.

 

But neither poor teaching nor poor practice invalidated good teaching and practice.

[AdvRoboticsE529]

Oh yes, testing a specific statistical set of pills in order to determine 100% of the pills.

 

In previous decades, check toxic capsules, example of social phenomenon includes Chicago Tylenol Murders, tempered capsules that is unaccounted for in statistical models.

 

Some methods can destroy the pills, so how about create new methods and technology in the process to result in a mass-confirmation of properties of pills? Instead of living in uncertainty?

 

If it is admitted, then perhaps we can move and improve upon previous methods.

 

I for one, intend to make identification of specific set of diseases quickly by the percentage of active anti-bodies / antigens, not the statistical approach, but the basically a fraction approach, or, you can take the statistical approach to result in the possibility of incorrect identification.

[davidivad]

statistics must be interpreted.

like art, it requires pattern recognition which is subjective.

this is why it is the job of the scientist to be as objective as possible.

in a paper that has statistics, and is well written, the author does not quantify the data for you. he or she merely shows you the data.

it is up to the professionals of the field to decide if it is useful.

since right and wrong are subjective ideas in themselves, we pick the interpretations of those best able to pick the least of evils.

truth be told the concept is effective enough to have measurable results.

[Acme]

What would you say is the validity of statistics?

 

Also, if you're going to support your argument with definitions please don't, as I think you must ask yourself why you have so much confidence in the definitions in the 1st place

 

I believe the uncertainty just encourages people to not search for the *real* relationship or variables / constants, and it also encourages pointless studies that emphasizes on correlation when the true relationship (if existent at all) is *not* proven.

 

This also goes for probability, however, I heard that probability is quite applicable in quantum physics.

If you don't like statistics, then don't use them. Likewise if you don't like conventions, ignore them. In any case, nothing you present here is going to change/stop the teaching and use of statistics and that's a *real* relationship you can test.

Edited March 31, 2014 by Acme

[davidivad]

biased results?

[Greg H.]

There are three kinds of lies: Lies, Damned Lies, and Statistics.

That said, statistics are very useful in a number of fields provided that:

A) They are collected correctly, with a proper methodology, and

B) They are interpreted correctly by the reader.

 

Unfortauntely, both A and B fail with amazing regularity through no fault of the math itself.

[studiot]

 

statistics must be interpreted.

like art, it requires pattern recognition which is subjective.

 

In research, yes.

 

But with the greatest respect, in the quality control situation we have been discussing there is no room for art or pattern recognition.

 

A batch of product either falls inside or outside of the quality control limits, as calculated in a particular way.

 

 

 

as calculated in a particular way

 

This twist in the tail is more important than might at first appear.

[AdvRoboticsE529]

So, what would you guys think would be the concrete answer to the validity of statistics? Or perhaps limited thesis is written upon such and have not been considered sufficiently?

[davidivad]

statistics itself is not valid. it is a matter of successful results.

[studiot]

 

So, what would you guys think would be the concrete answer to the validity of statistics?

 

Since you mention concrete, this is a good subject about which to discuss statistics.

 

Let us say you are in charge of a concrete production plant outputting 2500 tonnes of concrete per day.

 

How do you make sure the concrete is up to specification?

 

I should tell you further that the concrete is particularly high specification, because it is to provide the runways for jumbo jets to land on and failure of substandard concrete could lead to landing accidents with subsequent loss of life and runway functionality.

 

This is a real world problem I have faced, and met with success, in the past, with proper use of statistics to help me.

Edited March 31, 2014 by studiot

[davidivad]

 

Since you mention concrete, this is a good subject about which to discuss statistics.

 

Let us say you are in charge of a concrete production plant outputting 2500 tonnes of concrete per day.

 

How do you make sure the concrete is up to specification?

 

this may not be a good comparison as concrete is mixed specificly to the application.

however, you make a good argument for it's usefulness.

Edited March 31, 2014 by davidivad

[studiot]

 

this may not be a good comparison as concrete is mixed specificly to the application.

 

Go , how?

[davidivad]

you place an order for concrete.

you ask for six bag.

you get six bag.

 

not all concrete is the same stuff either.

Edited March 31, 2014 by davidivad