Posted by Science Nerd on October 19, 2001 at 17:27:22:
In Reply to: OK, I did it - update posted by js on October 18, 2001 at 23:50:59:
Okay, if you truly want to carry out an unbiased, objective test and get a more accurate analysis, you need:
More than one model of mouthpiece
To have a number of people to listen
A (different) group of people to play them.
You need a double- blind situation where no one who is playing or hearing the mouthpieces knows what they are listening to (including not knowing which models are part of the study.
You need a defined hypothesis to test, such as "the addition of mass to a tuba mouthpiece design will (or will not) affect the color/response/volume/etc (You pick one) of a given instrument when all other factors remain constant."
Then, you know what you are trying to prove or disprove.
You can then design the details of your study to isolate these factors.
You also should present your method to someone else for criticism. You should then give your raw data to a third party to see if they come to the same conclusions.
Basically I would really like to see some proper research conducted on this sort of thing so we can get back to playing some music once in while.