< Back to Volume 12, Number 3
Volume 12: Number 3: Article 4
Evidence that Anomalous Statistical Influence Depends on the Details of the Random Process
Michael Ibison, Princeton Engineering Anomalies Research, Princeton
University, Princeton, NJ 08544
Within the field of anomalous statistical influence it has become widely
accepted that the susceptibility to human influence of a (non-pseudo)
random event generator (REG) is independent of the details of its construction.
This view was formalized by Schmidt, after obtaining similar experimental
results from very different REGs, in the Equivalence Hypothesis that
all REGs are equally susceptible. Since then, a number of models of
anomalous influence have built upon this hypothesis to predict a scaling
of the anomalous statistical yield with other factors, assuming the
REG details can be ignored. For example, "time-normalized," and Decision
Augmentation Theory (DAT) models predict that the anomalous Z-score
scales as the square-root of, respectively, the time invested by the
human operator, and the number of operator initiations of the random
process (button pushes).
This paper is a report of an experiment to test the equivalence hypothesis,
and, by implication, the validity of any derivative model. Human operators
were invited to affect "trials" formed from accumulations of binary
random events wherein the method of accumulation was randomly switched
between two different modes. In one mode, a trial equal to the sum of
200 bits was presented to the operator as the target of intention (to
elevate or depress its value). In the other mode, the target of intention
was a number distilled from the sum of 2,000,000 bits, having the same
chance statistics as the 200 bit trial. By keeping the display and time
between trials identical for both modes, both operator and experimenter
were blind to the randomly alternating method of computing the trial
value.
A surprising and important finding from our results (p = 0.00037)
from 140,000 such trials is that anomalous influence of a random process
does strongly depend on the method of generation. As a consequence,
these results significantly refute the equivalence hypothesis, and therefore
the time-normalized and DAT models. Instead, it must be concluded that
"anomalous interaction" (Psychokinesis) is innately (partly or wholly)
a property of the machine, and therefore its description lies partly
or wholly in the domain of a (future) physics.
Keywords: anomalous influence, psychokinesis
FULL TEXT:
Evidence that Anomalous Statistical Influence Depends on the Details of the Random Process
To purchase back issues contact Allen Marketing & Management: 1-800-627-0629
