roger
Penultimate Amazing
- Joined
- May 22, 2002
- Messages
- 11,466
1. Obviously her plea to remain anonymous received a "no" 
2. I don't understand why JREF feels that a success rate of <50% constitutes performance less than or equal to chance.
For example, suppose the test is to guess a number in the range 1...1000. I get it right 20% of the time. That is fantastically better than chance, and I should win the million. On the other hand, if the range is 1..2, and I get it right 51% of the time, I wouldn't win the challenge. The % success rate by itself tells us nothing.
It is not clear from her letter exactly what the criteria are for success and failure, but if she can perform even slightly better than a control (no psychic attempt to bend the flame), then she is doing what she claims, even if she does it only 1% of the time, if the control exhibits 'success' 0.5% of the time.
I acknowledge that probably there is more correspondance, and perhaps it makes clear why a success rate of 20-30% would be no better than chance, but based on the information in thread I would say she may have a legitimate test protocol.
edited to add link to the thread I am discussing.
2. I don't understand why JREF feels that a success rate of <50% constitutes performance less than or equal to chance.
For example, suppose the test is to guess a number in the range 1...1000. I get it right 20% of the time. That is fantastically better than chance, and I should win the million. On the other hand, if the range is 1..2, and I get it right 51% of the time, I wouldn't win the challenge. The % success rate by itself tells us nothing.
It is not clear from her letter exactly what the criteria are for success and failure, but if she can perform even slightly better than a control (no psychic attempt to bend the flame), then she is doing what she claims, even if she does it only 1% of the time, if the control exhibits 'success' 0.5% of the time.
I acknowledge that probably there is more correspondance, and perhaps it makes clear why a success rate of 20-30% would be no better than chance, but based on the information in thread I would say she may have a legitimate test protocol.
edited to add link to the thread I am discussing.