Raw capture
SRI Targeting Requirements Task - May 1982
SRI Targeting Requirements Task - May 1982
- Source: https://www.cia.gov/readingroom/docs/CIA-RDP96-00787R000300060001-4.pdf
- Mirror used for extraction: https://documents2.theblackvault.com/documents/cia/stargate/STARGATE%20%234%20176/Part0001/CIA-RDP96-00788R001300320001-3.pdf
- Authors: Harold E. Puthoff, Russell Targ, Beverly S. Humphrey, Keith Harary
- Prepared for: Defense Intelligence Agency
- Report date: May 1982
- Report identifiers: SRI Project 3279-4; CIA-RDP96-00788R001300320001-3 on the extracted copy
- Sourced: 2026-09-04
- Extraction: Gemini 3.7 Flash through AOP from the 31-page scanned PDF, checked against local OCR at the title, methods, tables, conclusions, and final reference page
- Related analysis: ../../sources/keith-harary
Full report
TARGETING REQUIREMENTS TASK (U)
Final Report
May 1982
By:
Harold E. Puthoff
Russell Targ
Beverly S. Humphrey
Keith Harary
Prepared for:
Defense Intelligence Agency
Washington, D.C. 20301
Attention: DT-1A
Contract: MDA903-81-C-0292
SRI Project: 3279-4
Special Access Program for GRILL FLAME
Restrict Dissemination to Only Individuals with Verified Access
Not Releasable to Foreign Nationals
SRI International
333 Ravenswood Avenue
Menlo Park, California 94025 U.S.A.
(415) 326-6200 • Cable: SRI INTL MPK • TWX: 910-373-2046
The Black Vault
The Black Vault is the largest online Freedom of Information Act (FOIA) document clearinghouse in the world. The research efforts here are responsible for the declassification of hundreds of thousands of pages released by the U.S. Government & Military.
Discover the Truth at: http://www.theblackvault.com
This document is made available through the declassification efforts and research of John Greenewald, Jr., creator of The Black Vault.
TARGETING REQUIREMENTS TASK (U)
Final Report
Covering the Period October 1980 to October 1981
May 1982
By:
Harold E. Puthoff
Russell Targ
Beverly S. Humphrey
Keith Harary
Prepared for:
Defense Intelligence Agency
Washington, D.C. 20301
Attention: DT-1A
Contract No: MDA903-81-C-0292
SRI Project: 3279-4
Approved by:
Robert S. Leonard, Director, Radio Physics Laboratory
David D. Elliott, Vice President, Research and Analysis Division
Classified by: DT-1A
Review on: 31 May 2002
SRI/GF-0030
SPECIAL ACCESS PROGRAM FOR GRILL FLAME.
RESTRICT DISSEMINATION TO ONLY INDIVIDUALS WITH VERIFIED ACCESS.
NOT RELEASABLE TO FOREIGN NATIONALS
SECRET
CONTENTS
- LIST OF TABLES iii
- EXECUTIVE SUMMARY iv
- I OBJECTIVE 1
- II INTRODUCTION 2
- III PROTOCOLS 4
- A. General Protocol 4
- B. Viewer Selection 4
- C. Distribution of Trials Across Session Conditions 5
- D. Transcript Evaluation 6
- IV RESULTS 8
- A. Trial Collection 8
- B. Data Summaries 8
- C. Overall Findings 8
-
- Evidence for Remote Viewing 15
-
- Distribution of Results across Targeting Modes 15
-
- Effects of Mid-Session Feedback 17
-
- Caveats 18
-
- V SUMMARY AND CONCLUSIONS 21
- REFERENCES 24
TABLES
- 1 Distribution of Trials in Targeting Study 5
- 2 0-to-7 Point Evaluation Scale for Target/Transcript Correspondence 7
- 3 Summary of RV Data for Viewer 557 9
- 4 Summary of RV Data for Viewer 753 10
- 5 Summary of RV Data for Viewer 688 11
- 6 Summary of RV Data for Viewer 807 12
- 7 All Data from 48 RV Trials, with Mean Values for Each Viewer and Each Session Category 13
- 8 Summary of Coordinate RV Trials with Mid-Session Feedback 14
EXECUTIVE SUMMARY
In this report we present the results of a four-month investigation, carried out by SRI International, to determine the relative effectiveness of various targeting procedures in use in remote viewing (RV). Three such procedures were investigated:
(1) Beacon targeting, in which the viewer has had some personal contact with, or is given the photograph of, an individual at the target site.
(2) Coordinate targeting, in which the viewer is given the geographical coordinates of the target site.
(3) Abstract targeting, in which the viewer is only told that there is a target site to be described.
In our experiments with four remote viewers, three of whom performed reliably in the RV task (RV of San Francisco Bay Area sites), we did not find any overall significant differences in the efficacy of three targeting modes, subject to some variation because of individual preferences. Instead, reliable RV functioning with results of comparable accuracy was obtained with all three techniques.
As an additional task, we investigated the usefulness of giving the viewer limited mid-session feedback as to the general nature of the target site. We found that this procedure did not result in increased accuracy of description.
I OBJECTIVE
The objective of the "Targeting Requirements Task" was to determine the relative effectiveness of various targeting procedures for use in remote viewing (RV). If differences in relative effectiveness were found, SRI International was also to determine whether such differences depend on the characteristics of individual remote viewers or are widespread in nature.
II INTRODUCTION
In both SRI and client studies in RV over the past decade, several methods have been used to target the remote viewer on the site. Much of the early work used a person located at the target site as a target for the remote viewer.1–3* We refer to this as Beacon RV, because in some sense the individual at the site can be said to act as a "homing" beacon. A second technique, which has often been used in operational RV, and around which a training program is being developed,4 is Coordinate RV. In this procedure, the target site coordinates (latitude and longitude in degrees, minutes, and seconds) are given (with no further information) to the remote viewer who is to view the site. A third technique, which has been used occasionally with good success both in laboratory work and in operational viewing, we call Abstract RV. In this approach, the remote viewer is simply told that there is a target site to be described; no further information is given.
These three techniques, with variations,† have been used successfully, at SRI, in the client community, and elsewhere. However, no systematic comparison of their relative effectiveness has been made to date. This study compares the results of the use of the targeting techniques as described above under otherwise uniform RV conditions. The results are examined to determine whether significant quantitative differences exist as far as the quality of the RV product is concerned. These three representative techniques were chosen for this study because they span the range, from the concrete to the abstract, of the targeting techniques typically required in operational tasks.
Specifically, the targeting mode is varied over the three techniques. These techniques are designated here as Techniques A, B, and C (for Abstract, Beacon, and Coordinate, respectively). A variation of Technique C, designated C', is also incorporated into the study to examine whether modest feedback given to the viewer at mid-session about the general nature of the site increases accuracy in the remainder of the session.
* References are listed at the end of the report.
† For example, in Beacon RV, the remote viewer may be introduced to the outbound person who is to act as a beacon, or simply be shown his photograph.
III PROTOCOLS
A. General Protocol
The general protocol for the study is to closet a remote viewer with an experimenter at SRI, and, at a prearranged time, have the viewer describe an undisclosed remote site using the required targeting technique. The target site, one of sixty located in the San Francisco Bay Area within a 30-min driving radius of SRI, is selected by random number access to a target pool by a second experimenter in charge of overall protocol. For each viewer, target sites are used without replacement as the series progresses, so that no individual viewer has the same site twice. In all cases, the interviewer is blind to the target so that he is free to question the remote viewer to clarify his descriptions without fear of leading.
During the prearranged viewing period lasting 15-min, the viewer makes drawings of and records on tape his impressions of the target site. At the end of this viewing period, the interviewer collects the data for the file, finds out from the protocol experimenter what the target site was, and then takes the viewer to the site for feedback.
B. Viewer Selection
To evaluate fairly the effects of varying the target conditions, we chose to carry out the study with four relatively inexperienced SRI viewers, as opposed to the more experienced viewers who exhibit strong preferences for certain targeting techniques.
C. Distribution of Trials Across Session Conditions
Each of the four remote viewers chosen was asked to contribute twelve trials apiece, three trials each for each of the four techniques, A, B, C, and C'. This method provides a total of 48 trials, 12 in each of the four categories, distributed as shown in Table 1 below. Each of the viewers used the four techniques in a balanced, random intermixed order (e.g., BACC'ACB...) as is usual in psychological studies with several stimulus categories.
Table 1: DISTRIBUTION OF TRIALS IN TARGETING STUDY
| Viewer | Category A | Category B | Category C | Category C' |
|---|---|---|---|---|
| 557 | 3 | 3 | 3 | 3 |
| 753 | 3 | 3 | 3 | 3 |
| 807 | 3 | 3 | 3 | 3 |
| 688 | 3 | 3 | 3 | 3 |
The protocol experimenter tells the interviewer at the beginning of the session which technique is to be used. For Technique A, the interviewer simply informs the viewer that there is a target site to be described; no further information is given.
For Technique B, the viewer is either introduced in person to the outwardbound experimenter who will act as a beacon (Beacon Trial One), or is simply shown a photograph of an otherwise unknown outwardbound experimenter (Beacon Trials Two and Three). The reason for this intertrial variation is to obtain additional information about the amount of viewer/beacon contact required.
For Technique C and C', the viewer is read the coordinates (in degrees, minutes, and seconds) for the site. For Technique C', the interviewer obtains from the protocol experimenter before session start an envelope containing general information about the site (e.g., "target site is a building exterior," "target site is an open outdoor area," and so forth.) In mid-session, after the viewer has described the site to the best of his ability, the interviewer opens the feedback envelope and gives this additional information to determine whether it stimulates increased accuracy and detail in the viewer's subsequent images of the site.
D. Transcript Evaluation
In early programs, transcript analysis was carried out exclusively on the basis of blind judging (matching) of transcripts to target sites.1, 2 This technique, although excellent with regard to demonstrating the presence or absence of a viable RV function, did not provide a uniform measure from transcript to transcript of the quality of RV functioning. In the previous program, SRI, in cooperation with the client, developed a 0-to-7 point rating scale to be applied "nonblind", post hoc to the evaluation of transcripts.3 For no correspondence between transcript and target site, a 0 is assigned; for excellent correspondence a 7; and for intermediate correspondence an intermediate rating. The precise criteria for each rating is shown below in Table 2.
A comparison (in the previous program) of the ratings produced with this approach against the ratings produced by the blind-judging approach for a 36-trial series showed statistically significant positive correlation between the two techniques. Furthermore, application of the 0-to-7 point scale to randomly matched transcripts and targets from that study yielded chance results. These two findings taken together establish that application of the 0-to-7 point scale provides a reliable, objective measure of RV quality. This method was, therefore, chosen for evaluation of the transcripts for this targeting study.
Table 2: 0-TO-7 POINT EVALUATION SCALE FOR TARGET/TRANSCRIPT CORRESPONDENCE
| Point | Value Assigned to the Point |
|---|---|
| 7 | Excellent correspondence, including good analytical detail (e.g., naming the site by name), and with essentially no incorrect information. |
| 6 | Good correspondence with good analytical information (e.g., naming the function) and relatively little incorrect information. |
| 5 | Good correspondence with unambiguous unique matchable elements, but some incorrect information. |
| 4 | Good correspondence with several matchable elements intermixed with incorrect information. |
| 3 | Mixture of correct and incorrect elements, but enough of the former to indicate viewer has made contact with the site. |
| 2 | Some correct elements, but not sufficient to suggest results beyond chance expectation. |
| 1 | Little correspondence. |
| 0 | No correspondence. |
IV RESULTS
A. Trial Collection
In accord with the protocols outlined in Section III, a total of 48 trials were carried out, 12 with each of four remote viewers. As summarized in Table 1, each viewer contributed three trials for each of the four techniques.
B. Data Summaries
Data summaries for each of the four remote viewers are tabulated in Tables 3 through 6, and a collective summary is provided in Table 7. Listed in the individual viewer Tables 3 through 6 are the trial numbers (1 through 12) and associated sites, targeting techniques and 0-to-7 point-scale accuracy ratings. (Two columns appear in the accuracy ratings for Category C'. Ratings in the first column were assigned on the basis of material produced before feedback only, while those in the second column apply to the transcript as a whole, including material generated following feedback. The effects of mid-session feedback are treated in detail in Table 8, in which we present a detailed session-by-session summary.)
Table 3: SUMMARY OF RV DATA FOR VIEWER 557
(Overall mean = 2.3—Does not show evidence for RV)
| Trial Number | Target Site | Targeting Technique | Accuracy Rating (Before / Entire)* | Mode | Mean Rating / Mode |
|---|---|---|---|---|---|
| 2 | Allied Arts | A | 5 | Abstract | 3.3 |
| 7 | Cabana Hotel | A | 2 | Abstract | 3.3 |
| 12 | Bubble Building | A | 3 | Abstract | 3.3 |
| 1 | Padre Statue | B | 1.5 | Beacon | 1.5 |
| 6 | Baylands | B | 3 | Beacon | 1.5 |
| 10 | Depot Tunnel | B | 0 | Beacon | 1.5 |
| 3 | Mausoleum | C | 3 | Coordinates | 1.8 |
| 5 | Railroad Trestle | C | 1 | Coordinates | 1.8 |
| 9 | Boathouse | C | 1.5 | Coordinates | 1.8 |
| 4 | Pool Complex | C' | 4 4 | Coordinates with Mid-Session Feedback* | 2.5; 3.0 |
| 8 | Grocery Store | C' | 2 3 | Coordinates with Mid-Session Feedback* | 2.5; 3.0 |
| 11 | Underground Garage | C' | 1.5 2 | Coordinates with Mid-Session Feedback* | 2.5; 3.0 |
* First set of evaluations are made on the basis of material generated before feedback; second set, on the basis of the entire transcript.
Table 4: SUMMARY OF RV DATA FOR VIEWER 753*
(Overall mean = 3.3—Indicates contact with target sites)
| Trial Number | Target Site | Targeting Technique | Accuracy Rating (Before / Entire)† | Mode | Mean Rating / Mode |
|---|---|---|---|---|---|
| 2 | Dome House | A | 0 | Abstract | 1 |
| 8 | Bridge in Park | A | 1 | Abstract | 1 |
| 9 | Salt Pile | A | 2 | Abstract | 1 |
| 1 | Children's Playground | B | 6.5 | Beacon | 6 |
| 5 | Swimming Pool | B | 5.5 | Beacon | 6 |
| 10 | Church on Hill | B | 6 | Beacon | 6 |
| 3 | Varsity Theatre | C | 5 | Coordinates | 2.7 |
| 7 | Banana Records Building | C | 0 | Coordinates | 2.7 |
| 11 | Tennis Courts | C | 3 | Coordinates | 2.7 |
| 4 | Cemetery | C' | 1 2 | Coordinates with Mid-session Feedback† | 3.5; 3.5 |
| 6 | Miniature Golf | C' | 6 5.5 | Coordinates with Mid-session Feedback† | 3.5; 3.5 |
| 12 | Victorian House | C' | 3.5 3 | Coordinates with Mid-session Feedback† | 3.5; 3.5 |
* The viewer shows significant differences between conditions A, B, C (one-way analysis of variance; $ ext{df}_1 = 2$, $ ext{df}_2 = 6$: $F = 7.69$—$F = 5.14$ required for $p < 0.05$).
† First set of evaluations are made on the basis of material generated before feedback; second set, on the basis of the entire transcript.
Table 5: SUMMARY OF RV DATA FOR VIEWER 688
(Overall mean = 4.0—Indicates contact with target sites with good correspondences)
| Trial Number | Target Site | Targeting Technique | Accuracy Rating (Before / Entire)* | Mode | Mean Rating / Mode |
|---|---|---|---|---|---|
| 2 | Church on Hill | A | 6 | Abstract | 5.2 |
| 5 | Locomotive Playground | A | 6 | Abstract | 5.2 |
| 8 | Ely Chevrolet | A | 3.5 | Abstract | 5.2 |
| 1 | Mills Florist | B | 5 | Beacon | 3.8 |
| 7 | Salt Pile | B | 3.5 | Beacon | 3.8 |
| 12 | SRI Bike Shed | B | 3 | Beacon | 3.8 |
| 3 | Stanford Shopping Center | C | 3 | Coordinates | 2.7 |
| 6 | Boy Scout Fire Circle | C | 3 | Coordinates | 2.7 |
| 9 | Palo Alto City Hall | C | 2 | Coordinates | 2.7 |
| 4 | Underground Garage | C' | 5 3.5 | Coordinates with Mid-Session Feedback* | 4.2; 3.5 |
| 10 | Methodist Church | C' | 5 4.5 | Coordinates with Mid-Session Feedback* | 4.2; 3.5 |
| 11 | Art Museum | C' | 2.5 2.5 | Coordinates with Mid-Session Feedback* | 4.2; 3.5 |
* First set of evaluations are made on the basis of material generated before feedback; second set, on the basis of the entire transcript.
Table 6: SUMMARY OF RV DATA FOR VIEWER 807
(Overall mean = 4.1—Indicates contact with target sites with good correspondences)
| Trial Number | Target Site | Targeting Technique | Accuracy Rating (Before / Entire)* | Mode | Mean Rating / Mode |
|---|---|---|---|---|---|
| 2 | Wallbanger's | A | 6.5 | Abstract | 4.2 |
| 7 | Shielded Room | A | 2.5 | Abstract | 4.2 |
| 10 | Baylands | A | 3.5 | Abstract | 4.2 |
| 1 | Mills Florist | B | 3.5 | Beacon | 3.5 |
| 9 | Bubble Building | B | 2 | Beacon | 3.5 |
| 12 | Miniature Golf | B | 5 | Beacon | 3.5 |
| 3 | Hoover Tower | C | 3.5 | Coordinates | 4.7 |
| 4 | Tennis Courts | C | 5.5 | Coordinates | 4.7 |
| 5 | Mausoleum | C | 5 | Coordinates | 4.7 |
| 6 | Glass Slipper Motel | C' | 4.5 3.5 | Coordinates with Mid-Session Feedback* | 4; 3.3 |
| 8 | Victorian House | C' | 3.5 3 | Coordinates with Mid-Session Feedback* | 4; 3.3 |
| 11 | Varsity Theatre | C' | 4 3.5 | Coordinates with Mid-Session Feedback* | 4; 3.3 |
* First set of evaluations are made on the basis of material generated before feedback; second set, on the basis of the entire transcript.
Table 7: ALL DATA FROM 48 RV TRIALS, WITH MEAN VALUES FOR EACH VIEWER AND EACH SESSION CATEGORY
| Viewer | A (Abstract) | B (Beacon) | C (Coordinates) | C' Before Feedback | C' Including Feedback | Viewer Mean* |
|---|---|---|---|---|---|---|
| 557 | 5 2 3 $ar{x} = 3.3$ | 1.5 3 0 $ar{x} = 1.5$ | 3 1 1.5 $ar{x} = 1.8$ | 4 2 1.5 $ar{x} = 2.5$ | 4 3 2 $ar{x} = 3.0$ | 2.3 (non-RV) |
| 753† | 0 1 2 $ar{x} = 1.0$ | 6.5 5.5 6 $ar{x} = 6.0$ | 5 0 3 $ar{x} = 2.7$ | 1 6 3.5 $ar{x} = 3.5$ | 2 5.5 3 $ar{x} = 3.5$ | 3.3 |
| 688 | 6 6 3.5 $ar{x} = 5.2$ | 5 3.5 3 $ar{x} = 3.8$ | 3 3 2 $ar{x} = 2.7$ | 5 5 2.5 $ar{x} = 4.2$ | 3.5 4.5 2.5 $ar{x} = 3.5$ | 4.0 |
| 807 | 6.5 2.5 3.5 $ar{x} = 4.2$ | 3.5 2 5 $ar{x} = 3.5$ | 3.5 5.5 5 $ar{x} = 4.7$ | 4.5 3.5 4 $ar{x} = 4.0$ | 3.5 3 3.5 $ar{x} = 3.3$ | 4.1 |
| Category Mean | 3.4 | 3.7 | 3.0 | 3.5 | 3.3 | — |
| Category Mean of 3 Showing RV Function | 3.4 | 4.4 | 3.3 | 3.9 | 3.4 | — |
* Mode C' entry before feedback only used in calculating viewer mean.
† The viewer shows significant differences between conditions A, B, C (one-way analysis of variance; $ ext{df}_1 = 2$, $ ext{df}_2 = 6$; $F = 7.69$—$F = 5.14$ required for $p < 0.05$).
Table 8: SUMMARY OF COORDINATE RV TRIALS WITH MID-SESSION FEEDBACK (Type C' Trials)
| Viewer | Site | Rating Before Feedback | Rating Including Feedback | Postfeedback |
|---|---|---|---|---|
| 557 | Pool Complex | 4 | 4 | Some additional AOL,* but no significant further detail. |
| Grocery Store | 2 | 3 | Some additional correct detail: sign above entrance overhang, planters, benches, wall. | |
| Underground Garage | 1.5 | 2 | Some additional AOL, plus additional correct detail about columns and a description of entrance ramp inside garage. | |
| 753 | Cemetery | 1 | 2 | Some additional AOL, but also good feeling tones: formal, garden sense with someone working on it, as additional detail. |
| Miniature Golf | 6 | 5.5 | A few sparse further correct details (large sphere, concrete) and some additional AOL. | |
| Victorian House | 3.5 | 3 | Brought in AOL from another target site. | |
| 807 | Glass Slipper Motel | 4.5 | 3.5 | Detail of tree in wooden planter is good additional detail, but feedback called up a lot of additional AOL. |
| Victorial House | 3.5 | 3 | AOL descriptions of Macy's and other places were elicited by feedback. No further correct detail was provided. | |
| Varsity Theatre | 4 | 3.5 | Confirmed original perceptions, and added AOL about office buildings and grass. | |
| 688 | Underground | 5 | 3.5 | Feedback led to further mention of items already named, called up AOL and created confusion. No new correct details elicited by feedback. |
| Methodist Church | 5 | 4.5 | Further mention of details previously provided, along with some additional bits of AOL. | |
| Art Museum | 2.5 | 2.5 | After feedback, AOL details of specific building details provided--these were not correct. |
* AOL pertains to Analytical Overlay of memory and imagination.
C. Overall Findings
Most of the findings of this study are obtained by examination of Table 7. We, therefore, turn our attention for a moment to a detailed examination of this table.
The transcript ratings for each of the remote viewers, for each of the session categories, are shown in the individual boxes in the table. The techniques, listed across the top, are Abstract (A), Beacon (B), Coordinates (C) and Coordinates with Feedback (C'). The latter (C') technique has two columns of transcript evaluation numbers; those made on the basis of material up to the point of mid-session feedback (first column), and those made for the entire transcript, including material generated after feedback (second column).
The bottom row shows the mean transcript numbers for each targeting technique averaged both for all four viewers and for the three viewers who showed evidence for reliable RV (discussed below). The right-hand column shows each viewer's twelve-trial mean. For Technique C', the numbers before feedback only are used in the calculation of these means so that they are not contaminated by the effects of feedback.
1. Evidence for Remote Viewing
The first overall result of the study is obtained by noting each viewer's twelve-trial mean (Table 7, right-hand column). The twelve-trial means for the four viewers are 2.3, 3.3, 4.0, and 4.1, respectively. Reference to the rating-scale definitions in Table 2 indicates that the last three of the four viewers in Table 7 produced means high enough to constitute evidence for relatively reliable remote viewing, while Viewer 557, the first viewer, did not do so. (For this viewer evidence for RV was not totally lacking because five of the twelve trials rated a 3 or higher; rather, trial-to-trial reliability was lacking.) We conclude, therefore, that robust RV was obtained with three of the four remote viewers.
2. Distribution of Results across Targeting Modes
To evaluate the results of using the alternative targeting techniques, A, B, and C, we examine the A, B, and C columns of Table 7. Examination of the means in the bottom row shows little difference between alternative targeting strategies. This is confirmed in detail by statistical analysis of all the transcript rating numbers, both in the three-category X four-viewer matrix, and in the three-category X three-viewer matrix confined to the three remote viewers showing reliable RV functioning.* Therefore, the results obtained for Target Techniques A, B, and C were essentially the same.
As we examine the fine structure of individual viewer performance profiles, we find that the above conclusion for the group as a whole is especially reflected in the individual responses of the two stronger remote viewers, 688 and 807, who essentially did equally well with each of the three targeting techniques, as did the unreliable viewer, 557.†
Only in the case of the remaining successful remote viewer (753) do we find significant differences in the alternative targeting conditions; the Beacon (B) ratings are elevated, and the Abstract (A) ratings depressed, as compared with mean performance.† In this case, the viewer expressed from the beginning a strong preference for targeting on a beacon person, which seemed "natural," as compared with the increasing abstraction of the Coordinate (C) and Abstract (A) targeting technique. This preference for a particular targeting technique, correlated with better performance for that technique, can be contrasted with the lack of expressed preference on the part of the other viewers plus their relatively stable performance using the alternative techniques.
These results, taken together, lead us to conclude that there are no inherent differences in the use of Abstract (A), Beacon (B) or Coordinate (C) targeting techniques, but personal bias or preference on the part of a viewer can skew the relative effectiveness of these alternative targeting techniques in practice.
Finally, no differences of note were observed in Technique B (Beacon) between the first trial, in which the remote viewer is introduced in person to the individual who is to act as a beacon, and the second and third trials, in which the remote viewer is simply shown the photograph of an otherwise unknown beacon person.‡
* One-way analysis of variance: 3 X 4; $ ext{df}_1 = 2$, $ ext{df}_2 = 33$; $F = 0.47$ ($F = 3.29$ required for $p < 0.05$). 3 X 3; $ ext{df}_1 = 2$, $ ext{df}_2 = 24$; $F = 0.95$ ($F = 3.40$ required for $p < 0.05$).
† One-way analysis of variance: $ ext{df}_1 = 2$, $ ext{df}_2 = 6$ ($F = 5.14$ required for $p < 0.05$). $F(688) = 4.02$; $F(807) = 0.40$; $F(557) = 1.51$; $F(753) = 7.69$.
‡ One-way analysis of variance: $ ext{df}_1 = 1$, $ ext{df}_2 = 6$; $F = 0.25$ ($F = 4.96$ required for $p < 0.05$).
3. Effects of Mid-Session Feedback
In a series of twelve Coordinate Trials (labeled C'), three each contributed by each of the four remote viewers, viewers were given rudimentary mid-session feedback after providing initial descriptions on the basis of coordinate targeting (as in a C Trial). The interviewer then encouraged further response from the remote viewer.
The feedback material used was prepared in advance by the experimenter in charge of overall protocol, and was unknown to the interviewer until that moment in the RV session when he opened an envelope containing feedback information and disclosed its contents to the viewer. The type of feedback given was designed to be as "nonleading" as possible, meant only to give the viewer some verification if he were already on the right track. The feedback was in the form of a single phrase, such as "an expansive interior location" for an underground garage, or "an outdoor open area with structures" for a cemetery.
The data from the twelve C' trials with mid-session feedback are summarized in two columns of Table 7 and in Table 8. Comparison of the means in the bottom row for the results with feedback (second C' column) against the results, either of the same session before feedback (first C' column), or the Coordinates targeting without feedback (C column), shows no significant differences, either enhancement or degradation. This holds considering all the viewers, or just the three with reliable functioning.*
Specific session-by-session detail is presented in Table 8. It is clear from these data that feedback, presented in the form described, was not generally helpful in increasing the accuracy of postfeedback elaboration. Instead, in the majority of trials, the feedback appeared to trigger Analytical Overlay (AOL) of images from memory and imagination, resulting in some (though not significant) degradation of the description provided before feedback, at least in those cases where the initial description was good. In the few cases where the rating improved after feedback, the improvement can be attributed to leading from the feedback, because the results in those cases still showed little evidence for RV functioning.
Overall, then, there was no evidence that mid-session feedback led to improved accuracy. Instead, there was a trend (though statistically insignificant) toward degradation of the result by AOL.
* All viewers, one-way analysis of variance: $ ext{df}_1 = 1$, $ ext{df}_2 = 22$, $F = 4.3$ required for $p < 0.05$: $F( ext{C' before and after feedback}) = 0.16$, $F( ext{C, C' after feedback}) = 0.44$.
Three reliable viewers: $ ext{df}_1 = 1$, $ ext{df}_2 = 16$, $F = 4.49$ required for $p < 0.05$: $F( ext{C' before and after feedback}) = 0.53$, $F( ext{C, C' after feedback}) = 0.03$.
4. Caveats
In regard to the effects of mid-session feedback just described, care must be taken not to generalize that intrasession feedback in any form is necessarily unproductive; only that there was no evidence that feedback in the form given was useful. Evidence is emerging in another study,4 for example, that simple statements of "correct," given in immediate response to correct viewer statements can be helpful, parallelling similar evidence in computer "guessing game" studies in which immediate feedback appears to lead to increasingly elevated performance profiles.5 We have shown, however, that descriptive statements of fact about a site, given after a lengthy narrative by a viewer, may not be helpful.
With regard to the effects of a different kind of feedback, post-session access to information about the site, the targeting study was designed to parallel as closely as possible protocols that hold under operational conditions. As such, because feedback to the remote viewer is often made available at some future time, in our study we also provided feedback. In this case we took the viewer to the site following each session.
Such post-experiment feedback provides, however, a confounding factor, both in our study and in operational tasking in general: namely, the possibility of obtaining information via a precognition channel. At this point we have no data on whether a significant portion of the information is transferred via this channel in a typical RV session. It is only known, primarily from RV data generated in other laboratories,6 that a precognitive channel can provide significant amounts of information in studies designed to focus on this aspect.
To determine as best we could whether there was any evidence in this study for precognitive effects, we examined the transcripts and flagged references to future site visitation that might in principle trigger use of a precognitive channel. An average of approximately one reference per transcript met this criterion (49 references in 48 transcripts).
To determine first whether any potential effects of feedback precognition might be distributed unevenly across the session categories, and thereby possibly compromise the effort to compare targeting techniques, a statistical analysis of the distribution of future feedback references in the transcripts was done. (The number of references totalled 16, 11, and 14 for Targeting Techniques A, B, and C, respectively.) No evidence was found for an uneven distribution across session conditions, indicating no evidence for compromise caused by an uneven distribution of future-feedback references.*
To check the matter further, we investigated whether there was any evidence that references to future feedback resulted in higher individual transcript ratings, because a positive correlation between references and ratings might indicate that triggered precognition played a major role. Altogether, with 49 such references distributed across 48 transcripts, we found by statistical test that the correlation coefficient between number of references per transcript and transcript ratings was not significant ($r = 0.08$, $p = 0.70$).
Thus, we find no evidence that statements that might in principle encourage use of a precognitive channel had any effect, either for individual transcript ratings or for the differential comparisons between targeting conditions. The possibility of precognitive influence is, therefore, limited to the global possibility that a significant amount of information comes via the precognitive mode when it is available, simply because it is available. A separate study with feedback withheld on a random basis is required to resolve this global question.
* One-way analysis of variance: $ ext{df}_1 = 2$, $ ext{df}_2 = 33$, $F = 0.52$ ($F = 3.29$ required for $p < 0.05$).
V SUMMARY AND CONCLUSIONS
In this study, "Targeting Requirements Task", we investigated the relative effectiveness of three alternative RV targeting techniques in use at the present time. The techniques are:
(1) Beacon targeting, in which the remote viewer has had personal contact with, or is given a photograph of an individual located at the target site at the time of viewing.
(2) Coordinate targeting, in which the remote viewer is given the geographical coordinates (latitude and longitude, in degrees, minutes and seconds) of the remote site to be described.
(3) Abstract targeting, in which the remote viewer is told only that there is a site to be described.
In addition, as a secondary task we also investigated the efficacy of giving the remote viewer limited mid-session feedback as to the general nature of the target site whose more detailed description we were seeking.
To address these issues, we collected a total of 48 RV trials over a four-month period, using San Francisco Bay Area locations as the target sites. These 48 trials, twelve from each of four remote viewers, were divided into two groups: thirty-six trials evenly distributed across the three targeting techniques (Beacon, Coordinate and Abstract), and an additional twelve coordinate trials in which mid-session feedback was given, to be compared with those coordinate trials without mid-session feedback. Relatively inexperienced viewers were used to minimize a priori bias with regard to the efficacy of one targeting technique over another.
Before discussing the specific results of the study we note that the findings were obtained under conditions approximating those which hold in typical operational RV sessions, which include the possibility of eventual future feedback to the viewer as to "ground truth." The results obtained in this study, as in many operational tasks, are, therefore, subject to the caveat that a global precognitive channel could be operative, and it is recommended that this issue be examined separately in future work.
The results of this study are summarized as follows:
- Three of the four viewers exhibited reliable RV functioning.
- For the viewers as a group (and for the successful viewers as a subgroup), no significant differences as to the efficacy of one targeting technique over another emerged; all three techniques provided useful data of comparable accuracy, indicating that there is little, if any, intrinsic difference between the modes.
- For one of the successful viewers, who quickly developed an order of preference for targeting techniques, significant differences were noted, aligned with the expressed preferences; we take this to indicate that the apparent intrinsic equality of the technique evidenced in the overall results of the study can be modulated by personal preference or bias, and so the choice of targeting must be tempered by this factor.
- In the case of Beacon Targeting, no significant difference between personal contact and the use of a photograph was evident.
- Mid-session feedback in the form given (limited feedback as to the general nature of the site, following the development of a coherent 15- or 20-min narration by the viewer) yielded no significant improvement in accuracy, and some (though statistically nonsignificant) evidence for degradation of accuracy, at least in the better transcripts.
We, therefore, conclude that remote viewers can describe remote sites of interest with equal accuracy, using Beacon, Coordinate, or Abstract Targeting Techniques, subject only to their individual preferences. Attempts to increase the accuracy of such results by providing mid-session descriptive feedback as to the general nature of the site, are, however, not likely to be successful.
REFERENCES
- H. E. Puthoff and R. Targ, "A Perceptual Channel for Information Transfer over Kilometer Distances: Historical Perspective and Recent Research," Proc. IEEE, Vol. 64, pp. 329-354 (March 1976), UNCLASSIFIED.
- H. E. Puthoff, R. Targ, and E. C. May, "Experimental Psi Research: Implications for Physics," in The Role of Consciousness in the Physical World, Ed. R. Jahn, AAAS Selected Symposium 57 (Westview Press, Inc., Boulder, CO, 1981), UNCLASSIFIED.
- R. Targ, H. E. Puthoff, B. S. Humphrey, and E. C. May, "Special Orientation Techniques (U)," Final Report, SRI Project 8465, SRI International, Menlo Park, CA (June 1980), SECRET/NOFORN.
- H. E. Puthoff, "RV Reliability, Enhancement, and Evaluation (U)," Final Report, SRI Project 3279-1, SRI International, Menlo Park, CA (February 1982), SECRET.
- C. T. Tart, Learning to Use Extrasensory Perception (University of Chicago Press, Chicago, IL, 1976), UNCLASSIFIED.
- J. P. Bisaha and B. J. Dunne, "Multiple Subject and Long-Distance Precognitive Remote Viewing of Geographical Location," in Mind At Large: Institute of Electrical and Electronic Engineers Symposia on the Nature of Extrasensory Perception, Ed. C. T. Tart, H. E. Puthoff, R. Targ (Praeger Press, New York, NY, 1979), UNCLASSIFIED.
SPECIAL ACCESS PROGRAM FOR GRILL FLAME
RESTRICT DISSEMINATION TO ONLY INDIVIDUALS WITH VERIFIED ACCESS
NOT RELEASABLE TO FOREIGN NATIONALS
SECRET