• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Patricia Putt, Psychic

Pup

Philosopher
Pup
Joined
Dec 21, 2004
Messages
6,679
Concerning the proposed test here...

Ms. Putt will be able to see the volunteers' hands for skin color, wrinkles and callouses, and the basic outline of their height and body shape. Not to mention their shoes, unless they're covered by the gown when seated. How direct can the readings be? I'm assuming she can't just say: "You're African-American," "You're over sixty," "You're very tall," "You have calloused hands," etc.

But still, one could use the same clues for readings like "You've often had to deal with racial prejudice" or "You don't think young people today are like they used to be," or "You're self-conscious about your height."

Of course she wouldn't know what characteristic would stand out. For example, the first person at 6'3" might get the height reading, only for her to discover that everyone else is 6'5" or taller. But still, would guessing what characteristics stand out from the norm in the average population and offering short readings based on those, tend to beat 5/10 odds?
 
Pup,

This is a good point. I will write into the protocol that she cannot read on the basis of physical appearance.

Thanks!

~Remie
 
I suppose this is obvious, but I didn't noticed it explicitly mentioned in the protocol: she should not be allowed to mention or comment in any way on the subject's number in the reading.
 
Here is another test that can be done. This is to ensure the test cannot faked. Before the test is accepted by jref have a practice run using the current protocol. The only difference is that instead of using Ms. Putt some other person is used who has been told to cheat in any way possible. None of the people there will know that it is not Ms. Putt doing the test. If the testers say "Ms. Putt" has passed the test then there is a problem somewhere.

The second problem is what is the probability of getting 5 or more out of 10 right by chance? I suspect the answer is more than 0.001. Can some maths person please answer this question.
 
A pair of mittens might obviate the issue of hand identification.
 
I notice that both the testee and the subjects will know their number. This would allow the testee to include an indication of the number in the description. This could be done in several simple ways. For example, by using a key word list with each number one to ten corresponding to word to be used in the description. Or or by starting, say, the fifth word in each description with the corresponding letter, 1=A, 2=B and so on.

This problem could be eliminated by assuring that there is no prior contact between the subjects, and by not displaying the number. The number tag on the subject would be covered when he enters the room. The testee would fill out the description for each subject without knowing the number. After she completes the description the number can be uncovered and then written on the description sheet.

Robert Klaus
 
rjh01 said:
Here is another test that can be done. This is to ensure the test cannot faked. Before the test is accepted by jref have a practice run using the current protocol. The only difference is that instead of using Ms. Putt some other person is used who has been told to cheat in any way possible. None of the people there will know that it is not Ms. Putt doing the test. If the testers say "Ms. Putt" has passed the test then there is a problem somewhere.

The second problem is what is the probability of getting 5 or more out of 10 right by chance? I suspect the answer is more than 0.001. Can some maths person please answer this question.
Click to expand...

It's 0.0016
 
It may be covered by the no cheating clause at the top but, I think the detailed description should include the fact that Ms. Putt is not allowed to speak or make any gestures that might give away what she is writing while the subject is in the room.

The fact that the subjects can see and hear Ms. Putt bothers me. If there was any collusion between the subjects and the her, there are any number of ways she could signal what whe was writing. The protocol does not state where the subjects will come from, I presume they will be supplied by JREF with no prior knowledge or contact with Ms. Putt, but this still worries me when a $1M is at stake, I would not risk my money on this protocol.

IXP
 
I've read my suggestion over, and it still leaves a security hole. If the subjects know what order they are entering the room the same indicators in the descriptions could be used. The only solution I can think of is to make absolutely sure there has been no collusion.
 
Startz:

How are you calculating the probability?

Are you using the binomial?

Shouldn't it be this:

(10!/(5!*5!))*(0.1^5)*(0.9^5)=0.001488035

(I can't figure out how to insert an equation, so I wrote it in Excel-speak)
 
Two security improvements come to mind.

Forget the mask, mittens, graduation gown, etc., and put the subject behind a screen. Then the subjects and Ms. Putt cannot see each other, so they cannot transmit any cues visually.

Randomize the copies of the readings given to the subjects afterward to evaluate. This takes care of them knowing where they were in the sequence. Additionally, they could be assigned random hex labels to begin with, e.g. R572JK8. They have to drape the label tag over the screen before they sit down, and retrieve it after Ms. Putt leaves the room.
 
Answered my own question.

The equation

(10!/(5!*5!))*(0.1^5)*(0.9^5)=0.001488035

calculates the probability of exactly 5 correct guesses.

The probability of 5 or more is 0.001634937.

Forgive the intrusion.
 
I don't know if Ms Putt is reading this but if you are, I think you should just try something similar to the protocol described above... pick truly random people and don't give them any hints. I think you'll find you'll struggle to get 5 right.

If I'm wrong, please tell me!
 
Manduca said:
Answered my own question.

The equation

(10!/(5!*5!))*(0.1^5)*(0.9^5)=0.001488035

calculates the probability of exactly 5 correct guesses.

The probability of 5 or more is 0.001634937.

Forgive the intrusion.
Click to expand...

No problem. I'm always grateful for someone checking my math.
 
I'm not sure if I missed it, but shouldn't the bundles of readings that are handed back for identification be mixed up first?
 
Manduca said:
Forget the mask, mittens, graduation gown, etc., and put the subject behind a screen. Then the subjects and Ms. Putt cannot see each other, so they cannot transmit any cues visually.
Click to expand...


I'm sure that the reason for the masks and gowns is that Ms. Putt has already stated that she cannot perform if the subjects are behind a screen.
 
Is Patrica Putt interacting on any forum?

Applicants often wander over to engage in the forums. I would love to swap messages with her and hear what she has to say to us.
 
I don´t really like that poetry part... (not that I don´t like Blake!) The subjects reading the lines are bound to give away some information about themselves (dialect, possibly foreign accent). If all Ms. Putt needs to connect with the "control" is a sample of the subject´s voice, she could do equally well with a list of random syllables (like "Ba, Ma, Ga..." etc.) Or, if the words have to have some meaning, they should be in an extremely foreign language (like, say, Yucagiric or Basque) - or even an extinct (Sumerian, Hittite) or artificial (Esperanto, Volapuk) !
 
You must log in or register to reply here.

Back
Top Bottom