or the possibilities of collusion.
Another experiment would then be necessary, and the arguments would begin all over again. On this question I am in agreement with Schiller, and I favor a quite different method of approach.
The main obstacle to the acceptance of parapsychological phenomena is the apparent rarity of the people who can produce them under even reasonable conditions of control. Now this rarity I believe to be apparent rather than real. We do not know the signs by which to distinguish these exceptional card-guessers and so we waste time and effort in testing the wrong kind of people. There is increasing reason to believe that we shall not discover them in university populations and that it is a waste of time to experiment with students. But experience of the last few months has indicated that it is among the less sophisticated types that we should pursue our search -- especially among children living in rural communities or in backward countries.
I think there is little doubt that with an increasing number of such high-scoring subjects much of the prejudice of ordinary scientific workers will disappear. When more and more competent Experimenters report on cases of high-scoring subjects, the hypothesis of collusion will become as extinct as the dodo. While it is, in the last resort, possible to suggest that two or three Experimenters have faked their results, this will not be possible when scores of competent investigators produce their reports on similar cases. I suggest to Price, therefore, that efforts should be directed toward the discovery of the personality characteristics of these people who make averages of 8 or 10 hits per 25 over considerable periods, the sort of communities in which they may be successfully found, and so on. In other words we should aim at repeatability by more and more investigators.