FEEDBACK AND PRECOGNITION DEPENDENT REMOTE VIEWING EXPERIMENTS

• Document Type: CREST
• Collection: STARGATE
• Document Number (FOIA) /ESDN (CREST): CIA-RDP96-00789R002200200001-5
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 22
• Document Creation Date: November 4, 2016
• Document Release Date: May 17, 2000
• Sequence Number: 1
• Content Type: REPORT

Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
SECRET
FEEDBACK AND PRECOGNITION DEPENDENT
REMOTE VIEWING EXPERIMENTS (U)
PETER J. McNELIS, DSW
CONTRACTING OFFICER'S TECHNICAL REPRESENTATIVE
RESTRICTED DISSEMINATION TO THOSE WITH VERIFIED ACCESS
TO THE PROJECT
333 Ravenswood Avenue
Menlo Park, California 94025 U.S.A.
[phone]
Cable: SRI INTL MPK
TWX: 910-373-2046
! e
a tE-c d For Release 2000/08/08: CIA-RDP96-007 Rna2k Y001-5 GNOT
EN LEANENATIONALS
 Appro For Release 2000/08/08: CIA-RDP96-0077 PQ QDj 0001-5
9
Final Report- -Objective F, Tasks la and lb December 1987
Covering the Period 1 October 1986 to 30 September 1987
FEEDBACK AND PRECOGNITION DEPENDENT
REMOTE VIEWING EXPERIMENTS (U)
By: EDWIN C. MAY NEVIN D. LANTZ
Prepared for:
PETER J. McNELIS, DSW
CONTRACTING OFFICER'S TECHNICAL REPRESENTATIVE
WARNING NOTICE
RESTRICTED DISSEMINATION TO THOSE WITH VERIFIED ACCESS
TO THE PROJECT
MURRAY J. BARON, Director
Geoscience and Engineering Center
CLASSIFIED BY: HO, USAMRDC (SGRD-ZA)
Copy ....T.. of ...1J. Copies.
This document consists of / pages.
SRI /G F-0304
NOT RELEASABLE TO
SECRET FOREIGN NATIONALS
For ReleastV$0 YItf'~ bP e~1R. L? i8 Q040 373-2046
[phone]. able: I
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
ABSTRACT (U)
(S/NF) Two different precognition experiments were conducted during
FY 1987. The first of these involved a well-calibrated viewer (Subject
372) and used natural Bay Area sites as targets. Ten real-time and ten
precognitive trials (counterbalanced) yielded no statistical evidence for
remote viewing. In the second experiment, four viewers contributed
approximately 30 trials each in a similar counterbalanced real-time vs.
precognition protocol. In this experiment, however, the target material
were photographs from a national magazine. No statistical evidence for
remote viewing was observed in this experiment. In a third experiment
designed to explore the role of feedback upon remote viewing quality, two
of four viewers produced independently significant evidence for remote
viewing. A number of speculations are offered as to possible mechanisms
including real-time data acquisition and global precognition with noise
reduction.
Ionian
Approved For Release 2000/08/08 : - - 789R002200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
(U) TABLE OF CONTENTS
ABSTRACT ................................................................ ii
LIST OF TABLES ........................................................... iv
LIST OF FIGURES .......................................................... iv
I INTRODUCTION ................................................. 1
A. Real-Time vs. Precognition Experiment .......................... 4
B. The Tachistoscope Experiment .................................. 8
III RESULTS AND DISCUSSION ...................................... 10
A. Real-Time vs. Precognition Experiment .......................... 10
B. The Tachistoscope Experiment .................................. 12
REFERENCES .............................................................. 16
Approved For Release 2000/0U i E D2200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
(U) LIST OF TABLES
1.
Results for Viewer 372 ................................................... 10
2.
Real-Time vs. Precognition Results ......................................... 11
3.
Tachistoscope Feedback Experiment ....................................... 12
1.
(U) LIST OF FIGURES
Timing for the Beacon RV Experiment ..................................... 5
2.
Idealized Curves of Possible Relationships Between RV Quality and
Intensity of Feedback ................................................... 9
3.
RV Quality vs. Feedback Intensity: Viewer 009 .............................. 14
4.
RV Quality vs. Feedback Intensity: Viewer 177 .............................. 15
Approved For Release 2000/0UNCFFCDP5>M2200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
I INTRODUCTION (U)
(S/NF) Since 1973, remote viewing (RV) has been observed under a wide
variety of different conditions. A few of many possible examples are
coordinate RV* (targeting by geographical coordinates),' beacon RV (known
person at the remote site) ,2 abstract targeting (targeting by the word
"target" or by a random number or binary number) ,3 and targeting by remote
tasking, in which the task is sealed in an envelope which is
geographically isolated from the viewer. To first order, all of these
(and more) have been demonstrated successfully in the laboratory and under
operational conditions.
(S/NF) The main difficulty in trying to understand the various
successes of RV from a fundamental point of view is that RV appears to
require a large number of basic theories to explain the variety of
observables. How is it possible to describe access to remote information
with a single unifying concept when the target has been specified by a
complex series of random events, separated in time and space, and these
events are completely unknown to the viewer? This problem has been one of
the main sources of criticism about the existence of RV, in that nothing
else in nature appears to have such properties. (It is beyond the scope
of this report to argue this point. It suffices to say that most of the
great advances in science contributed to the organization and
understanding of seemingly unrelated data. The ultraviolet catastrophe
and early atomic spectra are but two examples of the confusion prior to
the understanding provided by early quantum theory.)
(U) SRI has been developing a heuristic model of psychoenergetic functioning4 that has
the potential of providing some understanding of the RV confusion described above. It is based
upon a concept called precognition. Since the 1930s, the parapsychology literature has been
reporting experiments that claim to demonstrate the existence of precognition--remote viewing
of target material that had not been specified at the time of the viewing. As yet, there is not a
meta-analysis of this literature, but there is a review of the experimental support for
* (U) References may be found at the end of this report.
Approved For Release 2000/08/08: CIAMOMPOWM002200200001-5
 Approved For Release 2000/08U tot[ P9 S I F I E D200200001-5
(U)
precognition.b Morris found a number of compelling studies that support the concept, but in his
opinion all have possible real-time--albeit, at times, somewhat strained--explanations. In 1976,
SRI reported four successful RV trials conducted in a precognition mode.2 Similarly, the
Princeton Engineering Anomalies Research (PEAR) group have conducted -500 RV
precognitive trials with results beyond chance expectation.e. 7 Working with a selected viewer, M.
Besent, Krippner et al. found highly significant evidence in 1972 to support precognition in a
dream environment.8 Finally, working with Besent in 1987, Honorton continued to find stable
and strong evidence for precognition in a computer driven forced-choice experiment.9
(U) A typical precognition protocol for a single RV trial is as follows:
(1) At 10:00, a viewer and monitor are sequestered.
(2) From 10:10 to 10:25, the viewing is conducted. The viewer is asked to provide
information about a target that will be generated at the completion of the
session.
(3) At 10:30, an assistant, who is blind to the session, randomly selects a target
from a large pool of target material.
(4) At 10:35, the viewer and monitor debrief the session using the selected target
as feedback material.
Even though substantial numbers of precognition experiments have been done elsewhere, SRI
has conducted all of the RV sessions (with the exception of the four trials in 1976) using
variations of a real-time protocol--the target material has been randomly selected prior to the
RV session.
(U) If precognition is a fact of nature, then it represents a possible mechanism by which
the viewer has access to the target material regardless of when that material was generated.
Conceptually, if there is an answer to an experimental question then, in principle, a viewer can
gain access to the "answer book" to obtain the necessary information. What occurs to generate
the "answer book" is completely superfluous. Such an "answer book" would provide a
convenient explanation of how RV occurs in spite of all the different and elaborate targeting
procedures. The "answer book" in these cases is the result of the viewing.
(U) An obvious difficulty arises if all this is true. What constitutes an "answer book?"
The most direct "answer book" might be the reporting of the target material to the viewer after
the session (feedback). Unfortunately, the situation is more complex. There are no examples in
nature that are fundamentally anthropomorphic. If information from a future time is available,
then most certainly it is available to anyone. The implication, then is that viewers obtain
Approved For Release 2000/08708 * GTi-R lAba-d 78 QQ2200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
(U)
information regardless of the feedback parameters--including no feedback to the viewer at all.
Quickly, we arrive at a position that the precognition model of information acquisition is
fundamentally unfalsifiable--therefore a nonproductive concept. Yet as indicated above, many
experiments suggest that precognition may be possible.
(U) One approach to the problem is to examine precognition empirically. During FY
1987, therefore, SRI conducted two different experiments that addressed separate aspects of the
concept. The first of these,10 initiated during FY 1986, involves using a tachistoscopic display of
the feedback material to attempt to manipulate the information available to the viewer from the
future. The other was a standard RV experiment, using a protocol similar to the one described
above, in an attempt to replicate the earlier precognition results. This report describes these
experiments in detail.
* (U) This report constitutes the deliverables for Objective F, Tasks la and 1b.
Approved For Release 2000/08/f18' CIA=RDP9F-0~89RE 00200001-5
 Approved For Release 2000/08/08;`IA-RDP96-00789ROO2200200001-5
II METHOD OF APPROACH (U)
A. (U) Real-Time vs. Precognition Experiment
(U) During FY 1987, SRI conducted two experiments to examine the effects of target
generation time in RV data acquisition. The first of these was with a selected viewer using natural
locations within 30 minutes driving time from SRI as target material. In the other experiment,
four experienced viewers used photographs from the National Geographic Magazine as target
material. In both experiments the time of target generation (before or after the RV session) was
unknown to either viewers or monitors.
1. (U) The Beacon RV Series
(U) To examine the role of target generation upon out-bound RV experiments, SRI
asked an experienced viewer (Viewer 372) to participate in a 20-trial series. Viewer 372 has
been calibrated in this particular task in that he/she has demonstrated significant RV
performance in all (2) of the beacon experiments conducted at SRI. 11, 12 Furthermore, Viewer
372 has expressed strong preference for this type of experiment rather than those that use
photographs as targets.
(S/NF) The target material consisted of 66 natural outdoor
locations within a half-hour's drive of SRI. The sites were selected on
the basis of the past performance of Viewer 372. Thus, the target
selection criteria allowed sites that would be more difficult for novice
viewers. The intent was to produce a target pool with a variety of
different material. For Viewer 372, the variety could be architectural
(and other details) as well as general gestalt features.
(U) The viewer and the monitor were blind to both the target pool and the
individual target selections. At the beginning of each trial, the viewer and monitor were
sequestered in the RV laboratory. The assistant then selected the target generation time and, if
appropriate, the target site. The target selection time for each trial was determined according to
lmj~
Approved For Release 2000/08/08 : OiM~llDr99~6789R002200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
(U)
a balanced random protocol. In the real-time condition, the target site was selected prior to the
remote viewing, and the assistant (beacon) had to be at the target site. In the precognition
condition, both the target selection and the beacon activity occurred after the remote viewing was
concluded.
(U) In order that both experimenter and viewer would be blind to the target
generation time, the timing of events was synchronized according to the diagram shown in Figure
1.
Clock
synch
Clock
synch
Condition
and target
selected,
assistant
leaves for
site
Condition
selected
Assistant
arrives at
site, RV
begins
RV
begins
Assistant
leaves site,
RV ends
RV ends,
site selected,
assistant
leaves for
site
Assistant
arrives at
SRI
Assistant
arrives at
site
Assistant
leaves site
Team goes
to site for
feedback
Assistant
returns to
SRI, team
goes to site
for feedback
0 5 35 50 80 95 125
Time (minutes into session--not to scale)
UNCLASSIFIED
FIGURE 1 (U) TIMING FOR THE BEACON RV EXPERIMENT
(U) After selecting the site, the assistant remained there for 15 minutes.
During that period, the assistant photographed the target with a Polaroid camera and recorded
impressions orally on audio tape. The pictures and recordings were used as a part of the
feedback when the viewer was taken to the site.
(U) The RV session lasted about 15 minutes (corresponding to the time the
beacon was at the site) during which the viewer recorded his/her impressions of the target by
drawing and writing descriptions. Sessions were also video taped as part of a separate
experiment.13 When the session was finished, all responses were copied and the originals were
given to an experiment coordinator. The viewer and monitor then proceeded (with the copies of
the session material) to the site for feedback. After the initial meeting prior to the experiment,
there was no contact between the beacon person and the viewer or monitor, in order to maintain
a strict double blind for the target generation.
C% I El Approved For Release 2000/08UNl~kRDP'969R5~1200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
b. (U) Analysis
(U) After the completion of the 20 trials (10 in each target generation time
condition), the monitor prepared the material for the analyst. The responses were examined to
determine whether any temporal clues existed (none did), and were randomly labeled. Since the
protocol specified that each used target was returned to the pool, it was possible that a target
could be used more than once. The monitor prepared a numbered list of 20 targets (not in
.session order) which included all targets used during the experiment and, if necessary, randomly
chosen decoy targets to complete a set of 20.
(U) The 20 targets and responses were given to two experienced analysts for
judging. The analysts' task was to rate on a continuum scale each response as to the degree of
correspondence with all 20 targets. A rank ordering was then prepared from the continuum
assessments. From these data, two 10x10 matrices were constructed, one each for the two
conditions. Each matrix was arranged so that the correct matches lie on the diagonal. Let To be
the trace of this matrix. By permuting rows (or columns) we calculate how many matrices would
have a trace less than or equal to To. The fraction, To/nl represents the p-value for the
condition. 14
2. The Target Photograph Series
(U) This series was conceptually similar to the one just described. The primary
differences were the target material (200 National Geographic Magazine photographs instead of
natural sites) and the associated timing.
(U) The protocol differed from that shown in Figure 1 in that the various
activities were driven by a series of signals rather than timed to the clock. Once the viewer and
monitor were sequestered in the RV laboratory, the monitor signaled an assistant by telephone,
letting the phone ring once. During a one-minute interval, the assistant determined which target
generation time condition to use (from a balanced random protocol) and either generated a
target or not. If a target was generated for the real-time condition, it was placed in a
predetermined location (on a different floor from the RV laboratory); a blank folder was placed
if the condition was precognition. When the minute had elapsed, the assistant used the one-ring
signal to the monitor that it was time for the session to begin. When the RV session was
completed, the monitor signaled the assistant with one ring. If the condition was real-time, the
assistant waited one minute and returned the one-ring signal to the monitor. If the condition was
Approved For Release 2000/081AK91[6MW18JROA 00200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
(U)
precognition, the assistant generated a target, and replaced the blank folder with the actual
target. At the end of the minute, the assistant returned the one ring signal to the monitor.
(U) The monitor then copied the response material, left the copies with the
viewer, presented the originals to an experiment coordinator, obtained the target, and returned
to the RV laboratory to provide feedback to the viewer. After an extensive debrief period, the
session was concluded.
(U) Two experienced viewers (Viewers 009 and 454) contributed 30 sessions
(15 in each condition) each using the above protocol. Experienced Viewer 105 contributed 13
sessions in each condition, and experienced Viewer 177 contributed 9 sessions in the real-time
condition and 12 in the precognition condition. Viewers 105 and 177 exercised their option to
stop this experiment in accordance with the human-use guidelines.
b. (U) Analysis
(U) The analysis has been described elsewhere,15 so only an overview will be
presented here. Using cluster analysis, all 200 targets had previously been assigned to orthogonal
clusters of similar targets (i.e., each cluster of similar targets differed from every other cluster).
The assistant prepared packages (one for each viewer) consisting of all the responses randomly
ordered. Next, he/she generated an ordered list (on target ID) of seven targets for each
response consisting of the actual target and six decoys (a different set of seven for each
response). The decoys were chosen from clusters different from each other and different from
the target cluster. The decoy clusters were chosen randomly from a set of 18, weighted by the
number of targets in each cluster. Once a cluster was selected, the decoy was randomly selected
from within the cluster. This procedure assured that all targets were equally likely to be chosen
as a decoy.
(U) The response material, and the target lists were presented to two analysts
for judging. The analysts rank ordered each set of seven targets for each response in accordance
with the best to the worst response/target match. For each viewer, a sum-of-ranks statistic was
computed for the sessions regardless of condition, and separately for the two target generation
conditions.
Approved For Release 2000/08W:C.- 96-~JAIA112200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
B. (U) The Tachistoscope Experiment
(U) During FY 1986, SRI developed the protocol, hardware, and software that were
necessary to conduct a feedback-dependency experiment.10 The basic idea was to try to identify
the time frame within which the viewer is accessing information. The target material is available
in real time and feedback is provided immediately after each session.
(U) One important assumption was necessary in order to provide meaning to the
experiment. Namely, the feedback experience was assumed to be proportional to the cognitive
awareness of the feedback material. Under this assumption, the amount of information available
at feedback time constituted the independent variable. Using techniques derived from vision
research,15 40 targets (selected randomly from the pool of 200 National Geographic Magazine
photographs) were prepared into 8 intensity groups of 5 targets each. Each intensity group
represented the cognitive awareness that each viewer would experience (on the average) at
feedback time. Of the eight intensities, one was zero (i.e., no feedback at all), one was below
subliminal threshold (SL), one was low SL threshold (-.25% recognition), one was mid SL
threshold (.50% recognition), one was high SL threshold (x?75% recognition), and three were
of increasing intensity above 100% recognition. Experientially, the top two intensities were
sufficient to experience nearly complete cognitive awareness of the feedback material. By
definition, those below SL could not be cognitively sensed.
(U) Because of the difficulties outlined in Section I, great care was exercised to limit the
information in the "answer book." At no time in the future would a response be cognitively
compared to its intended target. Three pieces of information are needed to provide complete
knowledge of a session; (1) the target, (2) the response, and (3) the comparison between them.
The target system was prepared by individuals who would not have access to the responses. The
monitor, assistant and viewer did not have access to the targets. Last, the analysts were never
informed which were the correct results on a trial-by-trial basis.
(U) Technically, the tachistoscope (the device to display the feedback material) was
controlled by a computer in such a way that everyone was blind to target selection during a trial.
For example, the slide tray always began and ended in the zero position, and a positive feedback
loop assured that the intended target was displayed at the correct time.
(U) Three experienced viewers (Viewers 009, 105, and 177) contributed 40
trials (5 at each intensity level) each. A novice (Viewer 137) also contributed 40 trials. A
random order of intensities of feedback was determined (by computer) once (and differently) for
Approved For Release 2000/0UNC{DPStA200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
(U)
each viewer. Once the order had been set, the trials cycled through the list of intensities until the
40 trials were complete. For a given intensity, a random selection (with replacement) was made
in real-time from the five possible targets. Once the target selection was complete, the computer
stepped the slide tray to the appropriate position and signaled the monitor (on a remote terminal)
to begin the session.
(U) When the viewing was complete, the viewer opened a window cover to
observe the feedback. (The monitor was prevented from seeing the feedback.) When the
viewer was ready, he/she initiated the feedback by pressing a button. One, and only one, display
appeared on the translucent window screen. (Electronics prevented the viewer from receiving
more feedback after the first button press.) The monitor was instructed not to discuss the
experience with the viewers in any way.
(U) Upon the completion of the session the monitor signaled the computer
that the session was over and the computer in turn cycled the slide tray back to zero and stored
the target information in a file. At that time, the viewer was free to go.
b. (U) Analysis
(U) The analysis proceeded exactly as described in Section II A.2.b. In
addition, however, the data were plotted as RV quality vs. feedback intensity. A schematic
representation of this plot and some possible interpretations are shown in Figure 2.
High
RV Quality
Significant RV
Real-time
Null Hypothesis
No RV
Subliminal J Intensity
Threshold
FIGURE 2 (U) IDEALIZED CURVES OF POSSIBLE RELATIONSHIPS BETWEEN
RV QUALITY AND INTENSITY OF FEEDBACK
' e9
Approved For Release 2000/08R~PUFAOi200200001-5
Significant RV Significant RV
Real-time and Precognition
Precognition /
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
r% r-11"
III RESULTS AND DISCUSSION (U)
A. (U) Real-Time vs. Precognition Experiments
1. (U) The Beacon RV Series
(U) Table 1 shows the results of the 20-session beacon RV series.
Table 1
Condition
Trace
p-value
Real-time
45.5
0.154
Precognition
51.5
0.638
Overall, no condition met the criteria for statistical significance. For Viewer 372, this represents
the first time at SRI that a series has not met statistical significance out of four attempts (counting
each condition as a separate attempt). Given Viewer 372's track record, we allow for some
speculation as to possible reasons for the results of this series.
(S/NF) Viewer 372 first participated in a six-trial RV experiment
in FY 1980. That study produced four first place matches and two second
place matches for a combined p-value of 0.003.11 His/her second
participation was in FY 1986. when twelve beacon RV trials were conducted
with an overall p-value of 0.007.12 Combined with the two efforts in FY
1987 (see Table 1) the average p-value is 0.201. Using an exact
calculation,16 the probability of observing an average p-value of 0.201 in
4 experiments is 0.017. This is consistent with a minimum p-value (0.003)
technique17 which yields 0.012.
(S/NF) There are at least two possible hypotheses for this
experiment not reaching significance. The first (and most likely one) is
given by Utts.18 If one is willing to estimate a "hit" rate given that RV
Approved For Release 2000/0 - - 002200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
GOP-
(S/NF)
is real, then it is possible to calculate the probability of observing a
significant study. While it is difficult to ascertain the actual "hit"
rate for RV, Utts provides an estimate for a similar process--Ganzfeld.
For a 10-trial study the probability of observing a significant result is
only 15%. (MCE is 5%, of course.)
(S/NF) Secondly, a new variable was introduced by the nature of
the protocol. The time between the remote viewing and the feedback was
greater than two hours. This represents an order of magnitude increase
over our other experiments. The influence of this increase is currently
unknown.
2. The Target Photograph Series
(S/NF) Table 2 shows the results of the four-viewer real-time vs.
precognition experiment. Based on the sum of ranks and their associated
p-values, there was no significant evidence of RV in this series.
Table 2
(U) REAL-TIME VS. PRECOGNITION RESULTS
Cond.
Viewer
Real-time
Precognition
All
Trials
rt/pc
009
57
(0.375)
62
(0.625)
119
(0.482)
15/15
105
61
(0.905)
51
(0.473)
112
(0.797)
13/13
177
32
(0.283)
46
(0.415)
78
(0.275)
9/12
454
70
(0.912)
68
(0.862)
138
(0.954)
15/15
Totals
220
(0.203)
227
(0.472)
447
(0.179)
52/55
(U) Sum-of-ranks (p-value)
(S/NF) Based on the past performance (in real-time RV) of these
particular viewers, the results are disappointing, Yet, because of their
record, we speculate upon possible reasons why this experiment did not
reach significance.
Approved For Release 2000/08/0 789R002200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
(S/NF) As described above, an estimate (provided by Utts18) of the
probability of a significant 10-15 trail RV series is approximately 15%.
Yet it remains surprising that no significant series was observed in eight
attempts.
(S/NF) A possible problem is that this particular experiment was
conducted after the successful tachistoscope experiment (described below).
That experiment required 40 trials from each viewer. Since this
experiment required 30 trails from each viewer, a given viewer had to
produce 70 remote viewings in approximately 80 days.
(S/NF) In summary, then, we were unable to demonstrate a
significant RV phenomenon in the real-time vs. precognition experiments.
Considering the vast amount of data in the literature that claim the
existence of precognition, we recommend that the study should be continued
at a later date.
(U) Table 3 shows the sum of ranks and associated p-values for the tachistoscope
feedback experiment.
Table 3
(U) TACHISTOSCOPE FEEDBACK EXPERIMENT
Viewer
Result
009
131
(0.012)
105
182
(0.962)
137
159
(0.484)
177
104
(3.5x10 8)
*
(U) Sum-of-ranks (p-value)
(S/NF) Viewers 009 and 177 produced independently significant
results. There are a number of ways in which we could combine these data,
but the most conservative is a binomial calculation assuming an event
probability of 0.05. Two successes in four trials corresponds to an exact
Approved For Release 2000/08/08000" 1"
89R002200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
(S/NF)
p-value of 0.014. A more realistic estimate is provided by a minimum
p-value (3.5 x 10-6) technique which yields 1.4 x 10-4.16 The important
point, however, is that this experiment produced strong evidence for an
informational anomaly.
(S/NF) Figures 3 and 4 show RV quality (1 is low, 7 is high) plotted
against intensity of the feedback for Viewers 009 and 177, respectively.
Shown also is the regression line for each viewer. These figures are to
be compared to Figure 2, the idealized expectations. The result that is
easiest to understand in Figure 2 is the positive correlation showing
increased RV performance with increased feedback intensity. We did not
observe any such correlation with either of the significant viewers. In
fact, the linear correlation coefficients were not significant.
(S/NF) The lack of positive correlation in the light of significant
evidence of RV complicates the interpretation considerably. The most
obvious conclusion is that the viewers obtained their data in real time
and not from the later feedback. But, if the argument posited in Section
I is correct (that precognition is unfalsifiable), then the experiment was
doomed to failure from the start. Another equally likely hypothesis is
that the underlying assumption that cognitive awareness constitutes
feedback information is incorrect. If this were true, we would expect to
see no correlation with intensity even if the precognition model were
correct.
(S/NF) Viewer 177's average sum of ranks was significantly (p < 0.02)
greater than his/her sum of ranks in the real-time vs. precognition
experiment. Viewer 009 produced a strong and similar trend that obtained
a probability against chance of 0.08. Assuming these differences are
meaningful, we can speculate that something in the tachistoscope
experiment resulted in a significant noise reduction. Possibly, short
exposures to feedback material allow the viewer to focus only upon the
major items and thus reduce the noise--the precognition model is assumed
here. In any event, continuing this experiment would shed light on the
difficult feedback interpretation problem.
0% r- a ET
Approved For Release 2000/08/08: - 02200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
(S/NF) In summary, the direct real-time vs. precognition experiment
failed to demonstrate significant evidence of RV. The tachistoscope
feedback experiment, however, provided considerable evidence of RV.
I r = 0.230 df = 38
SECRET/NOFORN
FIGURE 3 (U) RV QUALITY VS. FEEDBACK INTENSITY: VIEWER 009
Approved For Release 2000/08/08 : C - - 89R002200200001-5
OR '000
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
um- w
r=0102 df=38
'n
w ww w w w w w w w w
w w w w w w
p
w + rt w
a.
w w w w
n
w
w w w
0 1000 2000 3000 4000
Intensity
SECRET/NOFORN
FIGURE 4 (U) RV QUALITY VS. FEEDBACK INTENSITY: VIEWER 177
Approved For Release 2000/08/08 : C - - R002200200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
IV REFERENCES (U)
1. Puthoff, H. E., "RV Reliability, Enhancement, and Evaluation (U)," Final Report, Project
3279, SRI International, Menlo Park, California, February 1982, SECRET.
2. Puthoff, H. E. and Targ, R., "A Perceptual Channel for Information Transfer over
Kilometer Distances: Historical Perspective and Recent Research," Proceedings of the
IEEE, Vol. 64, No. 3, March, 1976, UNCLASSIFIED.
3. Puthoff, H. E., Targ, R., Humphrey, B. S., and Harary, K., "Targeting Requirements
Task (U)," Final Report, Project 3279, SRI International, Menlo Park, California, May,
1982, SECRET.
4. May, E. C., "Intuitive Data Sorting: An Informational Model of Psychoenergetic
Functioning," Final Report--Objective E, Tasks 3 and 4, Project 1291, SRI International,
Menlo Park, California, December, 1986, UNCLASSIFIED.
5. Morris, R. L., "Assessing Experimental Support for True Precognition," Journal of
Parapsychology, Vol. 46, pp. 321-336, December, 1982, UNCLASSIFIED.
6. Jahn, R. G., "The Persistent Paradox of Psychic Phenomena: An Engineering
Perspective," Proceedings of the IEEE, Vol. 70, No. 2, pp. 136-170, 1982,
UNCLASSIFIED.
7. Dunne, B. J., Jahn, R. G, and Nelson, R. D, "Precognitive Remote Perception,"
Technical Note PEAR 83003, Princeton Engineering Anomalies Research Laboratory,
Princeton University School of Engineering/Applied Science, Princeton, New Jersey,
1983, UNCLASSIFIED.
8. Krippner, S., Honorton, C., and Ullman, M., "A Second Precognitive Dream Study with
Malcom Bessent," Journal of the American Society for Psychical Research, Vol. 66, pp.
269-279, 1972, UNCLASSIFIED.
9. M. Bessent produced another highly significant result in a computerized forced-choice
real-time vs. precognition experiment at the Psychophysical Research Laboratories.
Private Communication, C. Honorton, 1987, UNCLASSIFIED.
10. Piantanida, T., Lantz, N. D., and May, E. C., "A Remote Viewing Feedback Dependency
Experiment," Interim Report--Objective E, Task 1, Project 1291, SRI International,
Menlo Park, California, December, 1986, UNCLASSIFIED.
11. Targ, R., Puthoff, H. E., Humphrey, B. S., and May, E. C., "Special Orientation
Techniques (U)," Final Report, Project 8465, SRI International, Menlo Park, California,
June, 1980, SECRET.
12. May, E. C., "Enhance Human Performance Investigation (U)," Final Report, Project
1291, SRI International, Menlo Park, California, December, 1986, SECRET.
13 May, E. C., "Enhance Human Performance Investigation (U)," Final Report, Project
1291, SRI International, Menlo Park, California, December, 1987, SECRET.
Approved For Release 2000/08M rUck P'9"If AW9WCA00200001-5
 Approved For Release 2000/08/08 : CIA-RDP96-00789R002200200001-5
UNCLASSIFIED
14. Scott, C., "On the Evaluation of Verbal Material in Parapsychology: A Discussion of Dr.
Pratt's Monograph," Journal of the Society for Psychical Research, Vol. 46, No 752, pp.
79-90, June, 1972, UNCLASSIFIED.
15. Humphrey, B. S., May, E. C., Utts, J. M., Frivold, T. J.,Luke, W. W., and Trask, V. V.,
"Fuzzy Set Applications in Remote Viewing Analysis (U)," Final Report--Objective A,
Task 3, SRI Menlo Park, California, December, 1987, SECRET.
16. Mood, A. M., Graybill, F. A., and Boes, D. C., Introduction to the Theory of Statistics,
Third Edition, p. 238, McGraw-Hill Series in Probability and Statistics, McGraw-Hill,
New York, New York, 1963, UNCLASSIFIED.
17. Hedges, L. V., and Olkin I., Statistical Methods for Meta-Analysis, p. 34, Academic
Press, Inc., New York, New York, 1985, UNCLASSIFIED.
18. Utts, J. M., "The Ganzfeld Debate: A Statistician's Perspective," Journal of
Parapsychology, Vol. 50, pp. 393-402, December, 1986, UNCLASSIFIED.
Approved For Release 2000/0&JM4Aw%$1F8tW200200001-5