CIA-RDP96-00787R000200130005-3

Approved For Relea # 2003/09/04 : CIA-RDP96-00787R000200130005-3
A DYNAMIC PK EXPERIMENT WITH INGO SWANN
Edwin C. May-
and Charles Honorton
Maimonides Medical Center
We report a preliminary study of psychokinetic influence on a
noise-driven binary random generator with Ingo Swann.
Mr .. Swann
has produced significant psi effects in a variety of controlled
tests in three different laboratories. He may, without exaggeration,
be called an applied specialist in this area. Our primary objective,
therefore, was not merely to provide another demonstration of PK, or
of Mr. Swann's abilities, but rather, to introduce him to this form
of psychokinetic task in order to assess the feasibility of develop-
ing a systematic research effort that would utilize Mr. Swann's in-
sight and ability to maximum advantage.
One of us (ECM) had described the instrument, which we call
PSIFI, in detail to Mr. Swann and it appeared appropriate that we
should take a "physics" approach to the PK task. Before the experi-
ment began, we explained the internal workings of the instrument and
the basic physics principles involved to Mr. Swann.
We partially
dismantled the sensitive sections of the instrument for his inspection,
pointing out internal data paths and the source of the random noise.
1
Present address:
Physics Dept., City College of San Francisco.
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3
1

Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3
May and Honorton/2
These were marked with tape. Further, we explained two possible
physics-type phenomena that would dramatically alter the behavior
of the instrument.
While Mr. Swann frequently peered into the open top of the
instrument and took frequent notes, we took data for runs of 10
3
trials at a generation rate of 50/sec. Mr. Swann operated the
instrument by manually depressing the reset button at the beginning
of each run. Feedback was provided via a scaler reading of the
run score. We informed Mr. Swann that chance expectation was 500/
run and that a significant influence on the instrument would be
reflected by consistent deviations from this mean.
Overall, 29 runs of 103
trials were completed.
The mean run
score of 493 is statistically significant (t = 2.71, 28 df, p = . 011,
2-tailed) . The mean run score for the first 10 runs was 495.4 (n.s.);
for the second set of 9 runs the mean was 493.1 (n.s.). The last
set of 10 runs was taken while Mr. Swann's EEG was being monitored.
During these runs, he was in a sound-attenuated room, adjacent to
the instrument room. These runs were independently significant,
with a mean of 491.8 (t = 2.63, 9 df, p = . 027) .
One hour after Mr. Swann's departure from the laboratory, C.H.
ran a control series of 3 x 10?
trials (i.e., 30 runs of 10"
4
trials) .
The control runs showed excellent approximation to theoretical
chance expectation, with a mean of 499.7 (t = 0.11, 29 df, p = . 91) .
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3
1

Approved For Releas 2003/09/04 : CIA-RDP96-00787R000200130005-3
May and Honorton/3
We conclude that Mr. Swann was successful in exerting a psycho-
kinetic influence on the instrument. We also regard the following
trends as provocative and highly suggestive of directions for
further collaborative research with Mr. Swann:
· Performance improved (and was independently significant)
with increased distance from the instrument. During the
last 10 runs, Mr. Swann was separated from the instrument
by distance and by two double steel walls.
· Performance showed steady improvement from the beginning
to end of session.
1
· Variability of performance decreased from the first to
the last 10 runs of the session (F10 10= 1.86, p = . 16) .
These latter two trends, while not in themselves significant,
suggest the advisability of further research relating to Mr. Swann's
belief that his performance in psi tasks is susceptible to learning.
1
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3

Approved For Release 2665/09/04" CIA RDP96-00787R000200130005-3
for the New York Meeting of the
American Physical Society
2 - 5 February 1976
Physics and Astronomy
Classification Scheme
Number 05
Bulletin Subject Heading in
Which Paper Should Be Placed:
General Physics
Possible Detection of Experimenter-Instrument
Interaction with a Selected Subject. E. C. MAY and
C. HONORTON, Maimonides Medical Center- An experiment
was performed with an individual who claims to influ-
ence remote physical systems by non-physical means.
Prior experimental tests in 3 other laboratories pro-
vided encouraging results. In the present experiment
the subject was tested with a noise driven binary ran-
dom event generator in which each event is compared to
a "target" alternating between '0' and 'l' at a rate
of 50 events/second. The number of events and the num-
ber of matches (events corresponding to targets) were
recorded. Data were taken for 29 runs of 103 events
during which the subject attempted to influence the
binary output of events. The results corresponded to a
probability against chance of p<0.011. Independently
significant results (p<0.027) were obtained during the
last 10 runs in which the subject was separated from
the instrument by 12 feet and by two double steel
walls. A control experiment was run one hour after the
subject's departure from the laboratory. Control data
were taken for 30 runs of 104 events and showed good
approximation to chance expectation (p<0.91). The
experimental results suggest that some anomalous in-
fluence on the generator occurred.
Submitted by
E. C. May
Signature of APS Member
-
E. C. MAY
Same name typewritten
Maimonides Medical Center
Address
Division of Parapsychology
4802 -- 10th Avenue
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3
Brooklyn, New York
11219

Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3
VOLITIONAL CONTROL IN A PSYCHOKINETIC TASK
WITH AUDITORY AND VISUAL FEEDBACK
Charles Honorton and Edwin C. May"
1
Maimonides Medical Center
During the last five years, 15 experiments have been reported
·
dealing with psychokinetic influence on electronic random generators ..
Twelve of these studies (80%) showed statistically significant (p <. 05)
evidence of PK influence. Even if we were to assume that for each re-
ported study there are five unreported, nonsignificant studies, this
result is still highly significant (p <1 x 10-3)
The purpose of the present experiment was to replicate and extend
the random generator (dynamic) PK work reported principally by Helmut
Schmidt and his associates. Specifically, we were interested in the
question of volitional control of dynamic psychokinetic influence.
The random generator used in this experiment has been described
by May (1975) . In order to totally exclude the possibility of generat-
or bias, all runs were performed in an automatic mode such that the
target ( 'heads' or 'tails') was alternated on each trial.
The generat-
ion rate was 10 bits/sec., with a run length of 100 bits (i.e. 10 sec.) .
We decided in advance to have 10 Ss complete 5 high-aim runs, try-
ing to enforce scores above 50%, and 5 low-aim runs, trying to enforce
scores below 50%.
Order of conditions (high-/low-aim) was alternated by
S.
1
Present address :
Physics Department, City College of San Francisco. .
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3

2
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3
Honorton and May
Two modes of feedback were used. Ss received auditory feedback
(tone) whenever they averaged +1.26 0 over the last 10 trials. Ss
received visual feedback in the form of a digital chart recording
·
which registered hits as deflections above baseline and misses as de-
flections below baseline (Fig. 1) .
All 10 Ss were experienced in some form of internal state explor-
ation.
Nine had participated in previous psi research in our laboratory ;
six had prior experience with dynamic PK tasks; six had prior experience
with EEG/EMG biofeedback training; four were meditators.
During the session, S stood or sat in front of the polygraph with
headphones for auditory feedback. For the high-aim condition, S was in-
structed to maintain the pen deflection above baseline as much as possible.
S was informed that when the pen was above baseline and the auditory feed-
back was on, s/he was scoring at a rate of 70% or higher. Similarly for
the low-aim condition, S was instructed to maintain the pen deflection
below baseline as much as possible, and was informed that when the pen
was below baseline and the auditory feedback was on, s/he was scoring at
1
a rate of 30% or lower.
The running time for each S was less than 3 min.
The mean run score for Ss in the high-aim condition was 51.92 (t =
2.89, 9 df, p = . 009, 1-tailed) . Of the 50 high-aim runs, 8 were signifi-
cantly (p <. 05) above chance (exact binomial p for 8/50 at p≤. 05 = . 0032) .
Only one of the 50 high-aim runs was significantly below chance.
The mean run score for Ss in the low-aim condition was 49.22 (t =
-1.11, 9 df, p = . 15) . Of the 50 low-aim runs, 4 were significantly
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3

3
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3
Honorton and May
below chance (exact binomial p = . 24) . Only one of the 50 low-aim runs
was significantly above chance.
The difference between high- and low-aim conditions was significant
(t = 2.05, 9 df, p = . 035) . Five of the 10 Ss individually obtained
significantly more hits in the high-aim condition than in the low-aim
condition.
The exact binomial p for 5/10 at p <. 05 is:
.00006. None of
the Ss obtained significantly more hits in the low-aim condition than in
the high-aim condition.
Despite the short duration of each experimental session, the Ss ex-
hibited a significant decline effect. Taking the first two runs of each
S, the mean run score for the high-aim condition was 54.3 (t = 3.03, 9 df,
p = . 007), and the mean run score for the low-aim condition was 48.2
(t = - 2.12, 9 df, p = . 031) . Thus, the high- vs. low-aim difference is
independently significant for the first two runs of each S (t = 2.90,
9 df, p = . 009). The last two runs of each S were associated with a mean
run score of 50.5 for both conditions. This decline in volitional perform-
ance was significant for the high-aim condition (t. = 2.94, 9 df, p = . 016,
2-tailed) .
The results of this study replicate Schmidt's findings and demonstrate
that Ss can exert volitional control of dynamic psychokinetic effects.
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3

Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3
HIGH AIM / LOW I
fig. 1
(
HIGH
LOW
10/10 -
5/10
-
Y AVE.
0/10 +
FEEDBACK
HITS = 60
HITS = 41
TRIALS = 100
TRIALS = 100
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3

Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3
MEAN/ RUN
t
P
N SIGNIF. RUNS
P
td
P
N SIGNIE SS
P
HIGH AIM
51.92
2.89
.009
8
.003
2.05
.035
5
0.00006
LOW AIM
49.22
-1.11|
.150
4
.240
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3

.
1
Approved For Release 2003/09(04 - CIA-RDP96-00787R000200130005-3
for the New York Meeting of the
American Physical Society
2 - 5 February 1976
Physics and Astronomy
Classification Scheme
Number
05
Bulletin Subject Heading in
Which Paper Should Be Placed:
General Physics
Possible Detection of Experimenter-Instrument
Interaction. C. HONORTON and E. C. MAY, Maimonides
Medical Center - Many experimenters have suspected
that there must be some anomalous influence upon their
equipment, when, after watching an experiment crash
into oblivion, the equipment resumes normal operation
just as the technician arrives. We decided to look for
a measurable effect. A noise driven random binary
event generator was constructed in such a way that
each event is compared to a "target" which alternated
between '0' and' '1'. The number of events and the num-
ber of matches (events corresponding to targets) were
recorded. In the control experiment, 20 runs of 105
events each were taken with no one present. The mean
score was 49980±139 matches which corresponds to a
probability against chance of p<0.57. Then 10 "experi-
menters" were asked to influence the generator to have
more matches than expected for the first 5 of 10 runs,
and less matches than expected for the last 5 runs.
The results, which corresponded to a probability
against chance of p<0.035, suggest that some anomalous
influence upon the generator occurred. In addition, 5
of the 10 people individually had significant results
at the p<0.05 level. This corresponds to a binomial
probability of p<0.00006.
AUTHORS' INSTRUCTIONS :
Please place these two papers
sequentially in the same session
with this one first.
Submitted by
E. C. may
Signature of APS Member
E. C. MAY
Same name typewritten
Maimonides Medical Center
Address
Division of Parapsychology
4802 -- 10th Avenue
Brooklyn, New York 11219
-
Approved For Release 2003/09/04 : CIA-RDP96-00787R000200130005-3