H. Schmidt, RANDOM GENERATORS AND LIVING SYSTEMS AS TARGETS IN RETRO-PK EXPERIMENTS, Journal of the American Society for Psychical Research, 91(1), 1997, pp. 1-13
PK efforts by one subject on two kinds of prerecorded time intervals w
ere compared. One kind was provided by the durations of successive bre
athing cycles of a human subject. The other kind was computer generate
d, with the help of a true random number generator, such as to match t
he breathing cycles with regard to average length and variance. After
the prerecording of the interval lengths, each interval received a ran
dom target assignment, long or short, specifying the direction of the
PK effort. An audio feedback had a built-in inversion provision, howev
er, so that the subject listened to tones of varying durations with th
e goal of always extending the durations of the tones. The subject suc
ceeded in this task by purposeful PK missing by means of getting menta
lly and physically tense and then desperately attempting to shorten th
e tone intervals. With both kinds of intervals 20 sessions were held,
each session containing 40 time intervals. The subject could not consc
iously distinguish between the two kinds. The PK-effect on the actual
breathing intervals was significant with z = 3.17 (p = .00076, one-tai
led). The scoring on the other intervals was lower, z = 1.68 (p = 046
one-tailed), but the difference is not statistically significant.