by I particularly enjoyed reading http://dl.dropbox.com/u/1018886/Bem6.pdf
This excerpt from p5 is particularly illustrative:-
To appreciate how unlikely the existence of psi really is, consider the facts that (a) casinos make profit, and (b) casinos feature the game of French roulette. French roulette features 37 numbers, 18 colored black, 18 colored red, and the special number 0. The situation we consider here is where gamblers bet on the color indicated by the roulette ball. Betting on the wrong color results in a loss of your stake, and betting on the right color will double your stake. Because of the special number 0, the house holds a small advantage over the gambler; the probability of the house winning
is 19=37. Consider now the possibility that the gambler could use psi to bet on the color that will shortly come up, that is, the color that will bring great wealth in the immediate future. In this context, even small effects of psi result in substantial payoffs. For instance, suppose a player with psi can anticipate the correct color in 53.1% of cases—the mean percentage correct across participants for the erotic pictures in Bem’s Experiment 1. Assume that this psi-player starts with only 100 euros, and bets 10 euro every time. The gambling stops whenever the psi-player is out of money (in which case the casino wins) or the psi-player has accumulated one million euros. After accounting for the house advantage, what is the probability that the psi-player will win one million euros? This probability, easily calculated from random walk theory (e.g., Feller, 1970, 1971) equals 48.6%. This means that, in this case, the expected profit for a psychic’s night out at the casino equals $485,900. If Bem’s psychic plays the game all year round, never raises the stakes, and always quits at a profit of a million dollars, the expected return is $177,353,500. Clearly, Bem’s psychic could bankrupt all casinos on the planet before anybody realized what was going on. This analysis leaves us with two possibilities. The first possibility is that, for whatever reason, the psi effects are not operative in casinos, but they are operative in psychological experiments on erotic pictures. The second possibility is that the psi effects are either nonexistent, or else so small that they cannot overcome the house advantage. Note that in the latter case, all of Bem’s experiments overestimate the effect.
It did also cross my mind to ask if anyone has formulated a coherent model that would generate these kind of results as a statistical artifact without resorting to the psi hypothesis?
For example, a simple (and totally unsubstantiated) hypothetical alternative explanation. Suppose that on some level the negative stimulus in the Bem experiment (3 I think) where it is flashed up briefly after a response is regarded as a "wrong" answer by the participant. At some point, let's say half way through, the participant evaluates their performance so far - if they are doing "well" and frequently picking the image which isn't followed by negative imagery they continue to respond at the same speed. If they are doing "badly" and frequently picking the image which is followed by negative imagery they slow down their response times globally.
If this were the case (and I assume something like this could be tested for - measuring participants average response times as a function of how far through the test they had progressed), then it could alter the average response time for people who had done badly to be more weighted towards their responses in the second half of the test, where people who had done well would continue at the same speed and give equal weighting to the responses in both halves.
Of course that's a bit simplistic (and possibly horribly explained)...
So, um any thoughts?
