Is this for real?

[GaryProtein]

In science, the observations and data come first. That is the basis for a model. The hypothesis and then the model comes from analyzing the data and observations.

Making a model first, is placing the cart before the horse.

I'm just being pedantic here.

 

[opus111]

Indeed - and the results of listening are data to which a model is to be fitted. Soundminded's doing it the other way around - saying Myles' data doesn't fit his (SM's) own model. Doing science would involve rejecting a model which doesn't fit data, not rejecting data that won't fit a particular model.

 

[MylesBAstor]

In science, the observations and data come first. That is the basis for a model. The hypothesis and then the model comes from analyzing the data and observations.

Making a model first, is placing the cart before the horse.

I'm just being pedantic here.

How do you carry out an experiment without having a thesis? You're saying you just blindly do an experiment. Not when I did research.

No you have a hypothesis that you're going to test, do the experiments, collect and analyze your data and see what the data shows. Does it or doesn't it make sense? For instance, I want to see if the presence of intracellular thiol compounds affects a cells response to radiation and/or drugs. That's where a grant proposal starts. Then I work with different agents, biochemical pathways, free radicals, etc to see if these combinations increase (additively or synergeistically), decrease or have no effect upon cellular survival.

 

[MylesBAstor]

Indeed - and the results of listening are data to which a model is to be fitted. Soundminded's doing it the other way around - saying Myles' data doesn't fit his (SM's) own model. Doing science would involve rejecting a model which doesn't fit data, not rejecting data that won't fit a particular model.

Yes.

 

[GaryProtein]

Actually, the model should be made from the data not made to fit the data.

The observations are primary.

 

[A.wayne]

What 10K speakers cables have you compared to the lower priced alternatives?

What lower priced speaker cable have you compared to your 10k speaker cables ..?

Did you change your binding post when inserting said 10k speaker cable, solid copper post makes as big a difference as the cables...

Regards .

 

[GaryProtein]

How do you carry out an experiment without having a thesis? You're saying you just blindly do an experiment. Not when I did research.

No you have a hypothesis that you're going to test, do the experiments, collect and analyze your data and see what the data shows. Does it or doesn't it make sense? For instance, I want to see if the presence of intracellular thiol compounds affects a cells response to radiation and/or drugs. That's where a grant proposal starts. Then I work with different agents, biochemical pathways, free radicals, etc to see if these combinations increase (additively or synergeistically), decrease or have no effect upon cellular survival.

Having a thesis (idea) does not require a model. The model is more advanced than the thesis. Having a thesis comes from initial observations that make you say to yourself, "hey, I saw a, b, and c in my prior experiences/experiments. Based on that, let's see if x, y, and z follows." The observations of a, b, and c then directs you to create an experiment to create your model. This isn't blindly experimenting.

Maybe I'm too caught up in semantics here.

 

[opus111]

I can't see a reason why a model shouldn't be made to fit the data.That would be making the data primary and the model contingent on the data. The alternative to that is the data are made to fit the model surely?

 

[GaryProtein]

I mean to say that the model should come from (be created from) the data.

A preconceived model should not be morphed to fit data.

 

[MylesBAstor]

I mean to say that the model should come from (be created from) the data.

A preconceived model should not be morphed to fit data.

Gary,

Do you think that everything happens in a vacuum? There are reams and reams of published data and publications for which to draw upon for ones experiments.

Also have you ever written a government grant proposal? You better have an idea to test on page one, some data to back up your thinking or else you won't get your foot in the door of a study section.

As far as preconceived models, you are thinking in black and white terms. It's how do you adapt your model to fit the exception that exist in every realm. Second, there can be very good models developed that are also tweaked to fit all the data. And then there's arguments such as whether a linear or linear-quadratic model fits some experimental model because of the statistical uncertainty at low levels.

 

[Robh3606]

To me it's you develop a theory based on observation. Basically this is what I think is happening. Then you design experiments that prove the theory.

Rob

 

[MylesBAstor]

What lower priced speaker cable have you compared to your 10k speaker cables ..?

Did you change your binding post when inserting said 10k speaker cable, solid copper post makes as big a difference as the cables...

Regards .

Did you?

And not always, as there are even more factors to consider. And I've played with and listened to many binding posts in my day, especially when modding my Maggies. BTW, there really is not such thing as a "solid" copper binding post that I know of. Every one that I know (maybe with the exception of Furutech?) is doped with some material such as Tellurium to harden the copper and make the posts easier to machine; even those so called copper posts from Edison Price. And the biggest problem that I found with these "solid" copper posts such as the Edison Price were that they were still too malleable and stripped easily.

As far as cables, the cheapest that I tried were the Radio Shack Phoenix Gold. The A320 (IC and speaker cable) as I wrote about were an absolute steal for $20 or $30. (and no the other Phoenix Gold cables weren't as good nor are the newer ones the same.) I've also tried 14 gauge Dayton wire (I think it's around $2-3/ft) and it's pretty good too. But the better the system, the more differences become apparent between these entry level and the better cables that I've listened to on the market including WW, MIT, Transparent, Cardas, Kubala Sosna, Cello Strings, Cable Research Labs, XLO, etc.

 

[GaryProtein]

Gary,

Do you think that everything happens in a vacuum? There are reams and reams of published data and publications for which to draw upon for ones experiments.

Also have you ever written a government grant proposal? You better have an idea to test on page one, some data to back up your thinking or else you won't get your foot in the door of a study section.

As far as preconceived models, you are thinking in black and white terms. It's how do you adapt your model to fit the exception that exist in every realm. Second, there can be very good models developed that are also tweaked to fit all the data. And then there's arguments such as whether a linear or linear-quadratic model fits some experimental model because of the statistical uncertainty at low levels.

I basically agree with all your points. Nothing happens in a vacuum and at this point in our civilization was have lots of data to draw from, but from a purist standpoint (the black and white you mentioned), the model should come after the observations. I admit I was getting too philosophical.

Regarding paper writing, as you stated you do need an idea with data to back it up when you write a proposal and in my experience virtually paper writers have an agenda because they have to follow the grant proposal they got their money for. They know beforehand what their data has to show, and unfortunately, often make sure their data supports what they wanted the grant to study because if it doesn't they won't easily get another grant. Practically no one ever publishes negative results. Those studies get buried because negative results can be the death knell for a researcher. No grantor wants to dole out grant money to researchers who don't get positive results.

 

[MylesBAstor]

I basically agree with all your points. Nothing happens in a vacuum and at this point in our civilization was have lots of data to draw from, but from a purist standpoint (the black and white you mentioned), the model should come after the observations. I admit I was getting too philosophical.

Regarding paper writing, as you stated you do need an idea with data to back it up when you write a proposal and in my experience virtually paper writers have an agenda because they have to follow the grant proposal they got their money for. They know beforehand what their data has to show, and unfortunately, often make sure their data supports what they wanted the grant to study because if it doesn't they won't easily get another grant. Practically no one ever publishes negative results. Those studies get buried because negative results can be the death knell for a researcher. No grantor wants to dole out grant money to researchers who don't get positive results.

Gary:

It's a Catch-22. You can write something very original and untested and the study section will say it's a great idea but collect more data so we can consider the grant application. The bottom line is that you're writing the grant proposal to get money so can carry on the research.

I agree and though not widespread, new ideas often get trashed. Look at what happened to the person who question the then "cholesterol" dogma and was ostracized for 20 years and banished to the back labs of some third rate hospital. But he kept going and proved that cholesterol isn't the only factor contributing to CHD.