@ auplater,
no, John Curl is absolutely right in his remark- although it is unlikely it could happen.
What he describes leads to an example for expectation bias of the participant.
If you don´t know for which exact number of items to look, then your expectation will come (nearly inevitably) into play.
Just an example for another bias mechanism which will not be disclosed only if you choose to take "peeking" out of the game.
Furthermore that is the reason why it makes a difference if you have to design a test for "things" or for human perception.
Wishes
P.S. That´s one of the reasons why i like the orginal idea more.
no, John Curl is absolutely right in his remark- although it is unlikely it could happen.
What he describes leads to an example for expectation bias of the participant.
If you don´t know for which exact number of items to look, then your expectation will come (nearly inevitably) into play.
Just an example for another bias mechanism which will not be disclosed only if you choose to take "peeking" out of the game.
Furthermore that is the reason why it makes a difference if you have to design a test for "things" or for human perception.
Wishes
P.S. That´s one of the reasons why i like the orginal idea more.
This makes crystal clear sense to me, in contrast to most of the last 20 pages. Why not stick with what works??@ auplater,
no, John Curl is absolutely right in his remark- although it is unlikely it could happen.
What he describes leads to an example for expectation bias of the participant.
If you don´t know for which exact number of items to look, then your expectation will come (nearly inevitably) into play.
Just an example for another bias mechanism which will not be disclosed only if you choose to take "peeking" out of the game.
Furthermore that is the reason why it makes a difference if you have to design a test for "things" or for human perception.
Wishes
P.S. That´s one of the reasons why i like the orginal idea more.
you've missed my point
of course it does... however, what good is accomplished taking cheap pot shots at attempts by others to ascertain existance of effects by "speaking in riddles" rather than displaying an understanding and constructively contributing to the design?
So what if 10 out of 10 aren't broken in? The criteria set should be able to account for this, right?
After all, he is one of the primary proponents of what is being tested, no? and refuses to participate implying the test is always fatally flawed. I guess the autism / immunization crowd was right all along, eh??
@ auplater,
no, John Curl is absolutely right in his remark- although it is unlikely it could happen.
What he describes leads to an example for expectation bias of the participant.
If you don´t know for which exact number of items to look, then your expectation will come (nearly inevitably) into play.
Just an example for another bias mechanism which will not be disclosed only if you choose to take "peeking" out of the game.
Furthermore that is the reason why it makes a difference if you have to design a test for "things" or for human perception.
Wishes
P.S. That´s one of the reasons why i like the orginal idea more.
of course it does... however, what good is accomplished taking cheap pot shots at attempts by others to ascertain existance of effects by "speaking in riddles" rather than displaying an understanding and constructively contributing to the design?
So what if 10 out of 10 aren't broken in? The criteria set should be able to account for this, right?
After all, he is one of the primary proponents of what is being tested, no? and refuses to participate implying the test is always fatally flawed. I guess the autism / immunization crowd was right all along, eh??
Evenly divided would be fine and takes the human factor a bit more into account.
Wishes
Maybe just another sort of human factor?
John Curl has unfortunately a lot of "cheap shots" to take in this forum and obviously has made a lot of bad experiences with experimenters more interested in confirming their own bias than in objective and valid test results.
I don´t see the criteria set being able to handle that. The only one who could learn to handle it would be Andre himself.
As said above, i think the "5 out of 10" idea is a bit more constructive.
After all, he is one of the primary proponents of what is being tested, no? and refuses to participate implying the test is always fatally flawed. I guess the autism / immunization crowd was right all along, eh??
q.e.d. (reg. the cheap shots)
Wishes
Last edited:
Member
Joined 2002
Shenanigans. You've got the significance wrong. And what's the probability of no cables broken in? And with reference cables, what's the difference?
Face it, John, the only thing that's changed is that you're now trying to convince other people, using fallacious arguments, to run away as fast as you did.
Evenly divided would be fine and takes the human factor a bit more into account.
Wishes
I have no idea what that means, but there's no question that revealing the number of the burned versus raw shifts the statistics from choice to choice. Remember the famous "Let's Make a Deal" paradox that was a consequence of Bayes' Theorem?
At this stage it is more important to get two sets of cables, one set burned in, the other not. I would prefer to get them marked as A and B, I will decide if I can hear a difference between them, if so I will just for fun tell which set I think is burned-in, then I wan't to know if I was right or wrong of course. If I can't hear a difference we have a problem.
There are a few things that I want from my side, no matter how stupid it may sound. Not wanting to tell jleaman how to make his cables but I want all cables to be made 'directional'. I guess there are something written on the cable or else it should be marked as it come from the spool. The cables that are selected for burn-in must all be done in parallel and connected in the same direction. Since I don't know the cables that are used and the fact that it is silver, I suggest a burn-in period of at least two weeks. Obviously all joints should be made with the same solder, preferably with some silver content. The cables must be packed in such a way to prevent movement and short bends.
We must still discuss how the burn-in will be done.
Regarding the debate about how many cable sets should be burned in, I suggest a random selection with the condition that at least one set should be different, just to serve as a confirmation, for me at least, that the burn-in process of the second batch were effective and comparable with the first cables.
Comments?
There are a few things that I want from my side, no matter how stupid it may sound. Not wanting to tell jleaman how to make his cables but I want all cables to be made 'directional'. I guess there are something written on the cable or else it should be marked as it come from the spool. The cables that are selected for burn-in must all be done in parallel and connected in the same direction. Since I don't know the cables that are used and the fact that it is silver, I suggest a burn-in period of at least two weeks. Obviously all joints should be made with the same solder, preferably with some silver content. The cables must be packed in such a way to prevent movement and short bends.
We must still discuss how the burn-in will be done.
Regarding the debate about how many cable sets should be burned in, I suggest a random selection with the condition that at least one set should be different, just to serve as a confirmation, for me at least, that the burn-in process of the second batch were effective and comparable with the first cables.
Comments?
That's why I want two test sets to try first.
Just mark them from C to L and flip a coin for each set to decide their fate.
The burner-inner can use this:
RANDOM.ORG - Coin Flipper
Choose the number of coins to equal the number of wires. Assign burned-in to "heads," and not-burned-in to tails. Easy!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.