[Febs]

Quote:

Originally Posted by earwicker7 /img/forum/go_quote.gif No, what I'm saying is that I wouldn't recommend using someone who's only checked out engine schematics on wiki as your mechanic. If I said "I am too a sound engineer, I read wikipedia" you'd justifiably laugh at me. All I'm saying is that there is a lot more to the scientific method than what you would expect if you used Head-Fi wisdom as your measurement; most (not all) people's understanding of the scientific method here is vague at best, as they are quoting a malleable website. I like wikipedia, it serves its purpose, but it is the single most unscientific source I can think of.

So then, in fact, Xenu's reference to Wikipedia does not does not prove any point that you are purporting to make, and was nothing but a cheap shot on your part. Don't you ever get tired of doing that?

 

[earwicker7]

Quote:

Originally Posted by Febs /img/forum/go_quote.gif So then, in fact, Xenu's reference to Wikipedia does not does not prove any point that you are purporting to make, and was nothing but a cheap shot on your part. Don't you ever get tired of doing that?

You know, I have better stuff to do than dance around this subject, it's like explaining colors to the blind

 

[Febs]

So apparently the answer to my last question is "no."

 

[cosmopragma]

earwicker7, I've just typed a reply to your rage but I had second thoughts and deleted it.
It's no fun to kick a man that has already knocked himself out and is lying on the ground.

 

[bigshot]

Quote:

Originally Posted by earwicker7 /img/forum/go_quote.gif Yeah, I guess graduating Phi Beta Kappa with a Bachelor of, oh what was that, oh yeah, Science doesn't give me any knowledge of the scientific method. Duh.

Motto!

See ya
Steve

 

[bigshot]

Quote:

Originally Posted by hciman77 /img/forum/go_quote.gif Really, my impression was that we are a relatively well educated bunch, we have several academics, researchers, doctors and a decent number of PhDs onboard. We have loads of people with formal training in EE and related disciplines.

Some of us just have good ol' American common sense.

See ya
Steve

 

[hciman77]

Quote:

Originally Posted by earwicker7 /img/forum/go_quote.gif Anyone who has ever been involved in an actual scientific experiment (which I have) knows that you do everything possible to prove your point,

Woah ! - I think what you meant to say was "you do everything possible to make sure that the experiment is conducted properly so as to make sure that you control all extraneous variables and to not let your biases influence the results"

"everything possible to prove your point"

sounds a lot like lets cook the results or ignore some data points to make sure the results fit our theory - not terribly sound from a scientific point of view.

 

[Hello]

I have to agree with earwicker's sentiment that many people who claim to be proponents of the scientific method really aren't very scientific in their reasoning.

A common mistake is to take the word of a scientist, thinking that amounts to the scientific method! But, actually, there's a subtle distinction. The scientific method is not about taking people's word for things -- its about experiments and verifiability of those experiments. And, sometimes experiments which 'prove' things are faulty, sometimes people fudge data, sometimes people just outright lie, etc. So, it's always a bit risky to just take someone's word for something, even if it's from a well-respected scientist.

There's also a lot of misuse statistics. Strictly speaking, if a person is trying to identify cable A from cable B, and gets it right say 50 out of 100 times, that doesn't prove no one could tell a difference. It doesn't even prove the test subject couldn't tell a difference! The conclusion is that there is no statistical evidence to support the hypothesis that the test subject could hear a difference, which is not the same thing as saying there's statistical evidence which proves the test subject could not tell a difference.

Now, as a matter of 'common sense' -- and common sense is actually quite unscientific -- I wouldn't pay thousands of dollars for a cable that I could only identify 50 out of 100 times, in comparison to some much cheaper cable. But, this is a personal decision that's different than trying to prove a more universal phenomena.

I think earwicker's sentiment about passing the buck in burden of proof has to do with the confusion of not having evidence to support a claim being the same as having evidence to deny a claim.

 

[hciman77]

Quote:

Originally Posted by Hello /img/forum/go_quote.gif There's also a lot of misuse statistics. Strictly speaking, if a person is trying to identify cable A from cable B, and gets it right say 50 out of 100 times, that doesn't prove no one could tell a difference. It doesn't even prove the test subject couldn't tell a difference! The conclusion is that there is no statistical evidence to support the hypothesis that the test subject could hear a difference, which is not the same thing as saying there's statistical evidence which proves the test subject could not tell a difference.

In part this is a question of semantics and in part it is to do with how you construct your hypotheses.

One common approach is the null hypothesis. You posit that there is no effect of changing a condition. So for example you posit that subjects will perform a dichotic listening task no differently if they are standing knee-deep in a vat of custard or not. If when the results come in the mean, mode, median and standard deviation are all the same and a T-test or one-way ANOVA or whatever doesnt reveal a difference then you:

"Fail to reject the null hypothesis"

If there is a significant difference ( p < 0.01 or P < 0.05) you "reject the null hypothesis"

Alternatively you might say

H1 (or RQ1) - Standing knee-deep in a vat of custard will impair dichotic listening performance

So you have a positive hypothesis

If there is no difference you say
there is no support for H1 or
H1 is not suppported

If there is a difference you say
there is support for H1 or
H1 is supported

Then you or others may repeat it varying

Subjects
Amount of custard (kneee-high, waist-high, neck-high)
Temperature of custard (cold, warm, hot)

While philosphically you may never prove that standing in a vat of custard (never) impairs performance eventually if you keep getting negative results you give up because it becomes those awful words generally accepted or you are bored with it or you cant get any more NSF money for replication studies.

 

[Hello]

And "Fail to reject the null hypothesis" does not mean "Accept the null hypothesis".

It's a very simple logical statement that A implies B is not equivalent to not A implies not B.

There are all sorts of reasons why one may not get statistical evidence to reject the null hypothesis when the null hypothesis is truly incorrect. Perhaps the differences in performance is very small, which would require larger sample sizes. Perhaps the wrong thing is being measured.

How you get to conclude 'not B' is my point. You're not using hypothesis testing. In your post, we get to conclude custard does not 'impair dichotic listening performance' because you got bored or because NSF funding dried up (after a bunch of negative results).

Now, I'm not saying you do this, but if after running a few experiments to see whether custard has an impairment on hearing, you don't get any evidence -- to me, the proper statement is to say you don't have evidence to support the claim for an impairment of hearing, and not to say "it's a scientific impossibility" or "it's statistical improbability".

If someone chooses not to believe something that is not scientifically proven, that's fine. But, something does not need to be scientifically proven to be true. Most things that people think they know are not scientifically proven. In fact, the very statement that something should be scientifically proven to be considered true is not scientific!

 

[Hello]

There's a lot of dishonesty in experiments and reporting statistical results. Even if cables did not make a difference, it would not be that hard to get a so-called statistically significant result. If you run an experiment a 100 separate times, you're going to have a pretty good chance that at least one of those experiments will produce a 'statistically significant' difference at say the alpha = .05 level. Of course, it is incorrect to throw away the negative results and keep the positive results -- but that's not to say companies don't do it.

The fact that no cable manufacturer has proclaimed a rigorous study that shows cable makes a difference to me means that cable manufacturers simply don't do the studies. Given the expense, the time, and the risk of performing such studies, it's no wonder that companies don't perform them. And, it doesn't seem like anyone else has the incentive to really pursue acquiring scientific evidence that cables can make a difference in sound.