In parapsychology studies, closed decks skew results
I’ve seen this in myself when working with a deck of playing cards. My best solution has been to shuffle well and then just take a handful of the cards so I don’t know how many of each there will be.
Closed deck=”a deck with a fixed number of each of the alternative cards.”
My speculation is that the more extended training intensified a flaw in Honorton’s and McCallam’s training procedure, viz., that they used closed decks (five of each of five symbols) for the feedback conditions. It has long been known that you cannot legitimately test for ESP when giving immediate feedback with closed decks. By keeping track (consciously or unconsciously) of what has already turned up, the subject can optimize his guesses near the end, i.e., if he knows all five stars have already come up he will no longer guess star, and thus will increase his hit probability on the remaining cards. I suspect that in the more extended training with closed decks, McCallam’s and Honorton’s subjects began improving their memories, not their ESP abilities. This probably didn’t happen much in the short (three run) training groups because of the subjects’ focus on the ESP task, but it would have become the winning strategy in the longer training.