A REMOTE VIEWING EVALUATION PROTOCOL REVISED JULY 1983

Document Type: 
Collection: 
Document Number (FOIA) /ESDN (CREST): 
CIA-RDP96-00788R001800050001-8
Release Decision: 
RIPPUB
Original Classification: 
S
Document Page Count: 
48
Document Creation Date: 
November 4, 2016
Document Release Date: 
April 10, 2000
Sequence Number: 
1
Case Number: 
Publication Date: 
December 1, 1982
Content Type: 
REPORT
File: 
AttachmentSize
PDF icon CIA-RDP96-00788R001800050001-8.pdf1.79 MB
Body: 
For Release 2000/08/08 : CIA-RDP96-00788 R001800050001-8 Final Report December 1982 11 A REMOTE VIEWING EVALUATION PROTOCOL (U) DEFENSE INTELLIGENCE AGENCY WASHINGTON, D.C. 20301 SG1J CONTRACT MDA908-82-C-0034 SPECIAL ACCESS PROGRAM FOR GRILL FLAME. RESTRICT DISSEMINATION TO ONLY INDIVIDUALS WITH VERIFIED ACCESS. CLASSIFIED BY: DT-5A REVIEW ON: 31 December 2002 Copy No. .....f....... This document consists of 48 pages. SRI/G F-024 7 333 Ravenswood Ave. ? Menlo Park, California 94025 NOT RELEASABLE TO (415) 326-6200 ? Q -'o TWX: 910-373-1246 NATIONALS ed For Release 2000/0 - DP96-007888001800 I Approved For Release 20p0Lqjfq0D0788R001800050001-8 LIST OF ILLUSTRATIONS (U) . . . . . . . . . . . . . . . . . . . . iii LIST OF TABLES (U) . . . . . . . . . . . . . . . . . . . . . . . iv I OBJECTIVE (U) . . . . . . . . . . . . . . . . . . . . . . . 1 I I INTRODUCTION (U) . . . . . . . . . . . . . . . . . . . . . . 2 III ANALYSIS PROTOCOL (U) . . . . . . . . . . . . . . . . . . . 4 IV CONCLUSIONS (U) . . . . . . . . . . . . . . . . . . . . . . 11 Appendix A--EVALUATION OF FOUR COORDINATE REMOTE VIEWINGS (U). . 12 Appendix B--SUMMARY OF EVALUATION TECHNIQUE (U). . . . . . . . . 43 REFERENCES (U . . . . . . . . . . . . . . . . . . . . . . . . . . 44 UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20I A& MDXXW0788R001800050001-8 1 Sample RV Assessment Form (U) . . . . . . . . . . . . . . . . 5 A-1 Transcript 1 (U) . . . . . . . . . . . . . . . . . . . . . . 14 A-2 Transcript 2 (U) . . . . . . . . . . . . . . . . . . . . . . 15 A-3 Transcript 3 (U) . . . . . . . . . . . . . . . . . . . . . . 16 A-4 Transcript 4 (U) . . . . . . . . . . . . . . . . . . . . . . 17 A-5 Target Site Hango (U) . . . . . . . . . . . . . . . . . . . . 19 A-6 Target Site Fernando de Noronha (U) . . . . . . . . . . . . . 20 A-7 Target Site Inverness (U) . . . . . . . . . . . . . . . . . . 21 A-8 Target Site Punkaharju (U) . . . . . . . . . . . . . . . . . 22 UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 200W0N0;LA 66EQ788R001800050001-8 1 Task-Defined Relevance Scale for Target Elements (U) . . . 6 2 Quality Assessment Scale (U) . . . . . . . . . . . . . . . . 8 3 Numerical Score Conversion Table (U) . . . . . . . . . . . . 8 A-1 Target Element Relevance Ratings for Hango (U) . . . . . . . 23 A-2 Target Element Relevance Ratings for Inverness (U) . . . . . 23 A-3 Target Element Relevance Ratings for Fernando de Noronha (U) . . . . . . . . . . . . . . . . . . . . . . . . . 24 A-4 Target Element Relevance Ratings for Punkaharju (U). . . . . 24 A-5 RV Assessment Forms for Calibration of Remote Viewing (U). . 25 A-6 A Rank Ordering of Weighted Averages (U) . . . . . . . . . . 41 UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 sw Approved For Release 20UINQAMl O788RO01800050001-8 (U) The objective of this task was to develop an evaluation procedure to assess the relative quality of a set of different remote viewing (RV) responses. 1 UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 2 g 1L61ED0788R001800050001-8 (U) To quantify the analysis procedure, we have divided the task into four separate areas: subject response, target/task definition, quality assessment, and numerical analysis. Figure 1 is a sample RV Assessment Form that has been designed to emphasize the separation of the analysis tasks. Each of the parts of the form are described below. (U) The subject's response should be prepared for analysis without any knowledge of either the target site or the overall task. The aim of this method of response preparation is to reduce a possibly redundant, rambling response to a coherent set of concepts. To meet this requirement we have developed a set of initial guidelines to the conceptualization procedure. (U) A concept is defined as a paraphrase of a single idea that has been expressed in the RV verbal or drawing response. That coherent idea should not be fragmented into component parts. For example, a response might be of the form, "I see a large, textured, gray building." The single concept that expresses this idea should be "large, textured, gray building," rather than four separate concepts--one for each word in the phrase. Each concept should be entered under the "Transcript Concept" column in the RV Assessment Form. (U) For this initial evaluation technique, a particular concept should be used only once in the analysis. (Some weighting factor propor- tional to concept frequency could be utilized, but, for the initial attempt, only unique concepts are used.) If in the construction of the transcript concept list a concept later in the transcript is a duplicate of an earlier one, it should be so noted by placing the concept number of the original concept in the "D" column. UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20UN A& D0788ROO1800050001-8 b 4) N N O ~D .-+ h M lT -1 O .-~ P. M CT .y O N o O M Pte. O .,y' N ,--a ur CO N ~ ON N ~D O e O O O r-I rl N N N M M M O ,-I N M eM ~ tL1 h OD 01 O ~ N C7 ~M m N ri rl N N 4 O N M d' b t0 GO O O H rl N N 0 0 m ra a d a 44 a+' ri N M t!~ f,D N OD Oi O N r~l UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 2ocJ A$$LFI 788R001800050001-8 (U) To utilize an analysis procedure that is capable of quantitative assessment, it is necessary to define, in advance, what the goals of the assessment are. In the RV Assessment Form, columns "Element of Target" and "Relevance" are provided to clearly define the goal of the analysis. In the ideal situation, an RV target should be completely specified in advance. A target typically consists of a number of target elements, each of which may have varying relevance with regard to the overall RV task. For any given target, an independent list of target elements should be prepared. The selection of what constitutes a target element is left completely to the discretion of the task coordinator. The target element must be selected with little regard to task relevance (target element relevance is accounted for later). Because an RV target consists, in principle, of an essentially infinite number of possible elements, discre- tion needs to be exercised in the selection process. (U) For each target element identified for the site, the task coordinator must define a relevance rating. This rating allows the coordinator to tailor the analysis to the task requirements. Table 1 shows the scale that is used for the target element relevance rating. (U) TASK-DEFINED RELEVANCE SCALE FOR TARGET ELEMENTS (U) Rating Relevance Scale 1 A target element of trivial interest 2 A target element of minor interest 3 A target element of intermediate interest 4 A target element of major interest 5 A target element of key interest UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 200W(WQA6$1b6EQ788RO01800050001-8 (U) For each transcript concept on the RV Assessment Form (Figure 1), the analyst should attempt to find the element on the list of target elements that he/she considers to be the best match. The analyst should be quite liberal in the concept/element matching (i.e., the quality of the match should be considered at this point in the analysis). If he/she is able to identify a target element that might be considered a match to the given concept, a 1 is placed in the "p" (present) column on the assessment form. If no element can be identified, a 0 is placed in the "p" column. After making a target element identification, the selected target element, and its corresponding overall relevance rating should be entered in the appropriate columns on the assessment form. (U) Having identified a corresponding target element for each con- cept, it is now appropriate to assess the quality of the match. The quality assessment is done on the basis of how well the single concept in question matches the selected target element. The judgement is to be made without regard to any other issues, such as importance of the concept to the transcript, or importance (relevance) of the target element to the target. Table 2 shows the quality assessment scale that is used for this part of the analysis. The appropriate quality score from Table 2 is entered in the "Quality" column on the RV Assessment Form for each concept for which a matching target element has been identified. from the relevance and quality (Tables 1 and 2) evaluation as follows: S' = P X R X Q , where P is the value in the "p" column (0 or 1); R is the relevance evaluation; and 0 is the quality assessment. S' can assume values ranging between 0 and 25. Table 3 demonstrates how to determine the final score, S, for a given value of S' for each concept. The conversion table is UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20c 0JN LA5-3 &788R001800050001-8 II INTRODUCTION (U) (U) In addressing the remote viewing (RV) evaluation question, we consider two separate requirements: Absolute evaluation of a single response for a single unknown target. ? Relative evaluation of a set of responses for a series of known targets. The first of these is of the most interest in an operational setting. As part of the RV enhancement task, we have considered this problem in two ways. First, by conducting an operational RV session between two calibra- tion RV sessions, a tentative a priori assessment of operational efficacy can be determined. The evaluation is made on the basis of performance during the calibration sessions, and on the basis of adherence to a pre- determined session structure. (U) A second technique for an a priori evaluation was explored as part of the Fiscal Year 1982 program in an audio-linguistic task. This task provided indications that careful linguistic analysis, when coupled with technical audio analysis, could yield an assessment in the absence of knowledge about the target. (U) Various techniques have been used in the pasts in an attempt to solve the relative evaluation problem. The most common of these was the simple rank ordering of all responses, as assessed against all possible targets used in an experimental series. In this procedure, a judge is presented with n RV transcripts and n target sites. His task is to arrange (U) References are listed at the end of this report. UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 2000/ 96-00788 R001800050001-8 (U) the transcripts in order of the best to least match for each of the n targets. A simple numerical counting procedure is then used to estimate the likelihood that the judge's transcript/target matches are by chance alone.2 This early technique contained little systematic structure for determining the final order of matches. (S) The first step toward systematizing the rank order judging procedure was to preprocess the raw data in the transcript by "concep- tualizing" both the verbal and the pictorial responses. Conceptualizing a transcript requires an analyst to paraphrase the transcript into a list of coherent statements. This concept list is then compared and scored concept-by-concept to each of the targets in the experiment. The resulting scores are averaged for each response, and all responses are rank-ordered on the basis of these scores.3 This improved analysis procedure was applied to a number of experiments within the Technology Transfer Task for INSCOM (U) The problem with the above technique is that there are no guide- lines as to how the analyst should paraphrase the transcript; furthermore, the method in which the concepts are to be assessed against the targets remains undefined. The purpose of the Evaluation Task in FY 1982 was to identify a procedure that corrected these deficiencies. Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20U1GLAM00788R001800050001-8 Table 2 (U) QUALITY ASSESSMENT SCALE (U) Rating Discrimination Scale 1 Poor description; only one or two aspects of the material match. 2 Fair description; a few aspects of the material match, but a large ambiguity exists. 3 Reasonable description; many aspects of the material match, but there remains some ambiguity. 4 Good description; a large number of aspects of the material matches, but it is possible to conceive of material that would be a better match. 5 Excellent description; all or nearly all aspects of the material match. (U) NUMERICAL SCORE CONVERSION TABLE (U) S P x R x Q Score Normalized Score 0 0 0.00 1 1 0.35 2 2 0.71 3 3 1.07 4 4 1.43 5 5 1.79 6 6 2.14 8 7 2.50 9 8 2.86 10 9 3.21 12 10 3.57 15 11 3.93 16 12 4.29 20 13 4.64 25 14 5.00 UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20UbCLA& D0788R001800050001-8 (U) used to eliminate the nonuniformly-distributed gaps in scoring numbers that occur if one simply uses the product S'. Thus, the final score for each concept ranges from 0 to a maximum of 14. This conversion table is provided as part of the RV Assessment Form. (U) If an assessment of an individual concept is required, the final score for each concept/target-element match can be related to the quality assessment scale by using the conversions shown in the third column of Table 3 and on the assessment sheet. It should be noted, however, that the integer scores are used to simplify the remaining calculations. (U) To determine a final evaluation of the complete transcript assessed against a given target, a weighted average of concept scores is computed. To assist in the calculation of the weighted average, a tally box score is provided at the bottom of the RV Assessment Form. For each of the possible scores, 0 through 14, the number of concepts that attained that particular score are counted. For example, if 3 concepts were evaluated with a score of 12, a 3 is entered in the box below the 12 score. If the frequency of occurrence of score Si is fi, then the final weighted average is computed by Ak = E fJ 'If j X Si /Z fJ Ak = 0.357 Ak (U) The normalized, weighted average score, Ak, is entered in the weighted average box on the assessment sheet. The weighted average score has been normalized to be within the range 0 C U UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 o .~ N M r' q M ti 00 Ol .~-4 N C4 m - M Lh ? - t? vi Approved For Release 20'0708105" 04DP9R00788R0018000500 w w yg a UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20YQ%(T8TOr8": CTi~ NAP-00788 ROO 1800050001-8 $ ~' n $0 n a0 0~ O ' ' o - N M ~ o . o c o m o O o 0 h ~' v ri d' N ra M & L os AY eM F1 Ln LO er to QQ O E a o o '+ o ~+ o ~ o - o ~ - ~+ FT rl 7 N O ~ 0 M $4 4) ' &1 00 -H H w F cd .+ + a + + a U) a . a~ a 4I a q- 0 P4 0 o w ,~ N N M N tD N 00 Of 04 ~-1 r4 N rl M d' .-1 M N UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20001 crJ `:'+CIR-KQP9b=00788R001800050001-8 O rl N M N t0 P OD 0! O 14 ~4 .~-~ m O '+ N M ep M O OD T O N N a 89 UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 2O 6GN14A SJl F1 0788R001800050001-8 O ,-I N M w M (D N OD Q! O C4 m V O N M N m GD m O N N fa N N 0 fA Ui U) N tD d' rl M N rl rl ?~ O ri N N N It sY M 1 O U] U M M Ui V7 41 Q q v ~ a F ro q 14 r-4 14 H -H 14 14 a 1-1 - , 0 U 41 ri N UI ? d m H qo~ 0 U P U p .,- a q q TI 44 a 0 ,~ rl N M tM U! t0 N 00 Q~ .~-~ "4 M N UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 200W0V081A 96.788RO01800050001-8 1 O .r N M w n tD n a O! 0 ?+ N O rt N M a h 0a 00 M 0 N N s 00 O O O M O O O old v O M r-1 N OY t) N Q ?i o o .-~ ~ 0 0 0 a rfi ~ N to O F M w n to n 00 a H N 1-1 N r-4 r-1 UNCLASSIFIED d) rb P 4J o CL U 4) V) U r. 0 U Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20c 6l04 4AP 1F4 D 788R001800050001-8 I o .r N ri .r n m n ao o, o w - - w O .~ N M t0 00 Of m r~i N N u N N N 0 C) N 0 at O' ri rl ri N 0 ea % a 112 a~ `+ N N N o sr er ep W O W Pq ~ fA ~ G~0 ~ ~ pg qa O E F E 93 0 m rl ++ N 0 pp F N at -H $4 4J U a . a a m a s ! ,~ o. ~ ,~ fc .-~ N l7 h m N OD O) -4 cn to UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 200 L "F%FI . 788R001800050001-8 o ~+ N N .M n tD n oo w o ?+ w .i .i 4 4 N M V~ H W W q7 N 0 N N 0 O N O O M N O O O O N - r~i.. O eN O M N O O rl tcaD h0 q bD ~ .N } .N UU 7 r+ y } 4 x ?+ as p .7 O U a .0 cd H a 41 b ti y -+ a a N N er 0 01 M 01 Y +? N r. w G C4 b (# al 01 'pr U U a U A U a Cl a O W 0 . U 6 R ~ r1 N M W 1f! ~D N W O -1 ,N N -I V to o Ir gamg.~ 01 y q U UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20OUNCLASfAF;E-D788R001800050001-8 o O - N M 'I U9 t0 N 00 Of 0 - N M d' rl r4 -4 O ~ N M V' U1 tD 00 O ~ ~ ~ N N N O W M 0 CO N N - tD 0 N er tD w i d a N eM N N V m U'7 N eM u N ~a v Cl U) Cl Cl eM N Cl M U! M to 11 93 ,SA O p ?d gC O N F Y 0 a 0 00 a ~+ 14 'r o rl ti 0 m b 4 14 v a m 0 41 a) FOi 44 m 61 N f0-, Myy F ep N q NO 0 a 600 z ,4 V .) M 0 a 7 4J 144 O 4~ -44 10 tOr 14 1-1 ;H 4-4 c a P. a w op. ,-1 N M el' U! t0 N 00 0) O F N M tM H UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 200"`CfRPIEQ788ROO18OOO5OOO1-8 g m o ~ N M .r as to n ao w o ~ ~ ~ ~ O r1 N M ep b t0 fb O) ~ ~ N N N N 4 ro N M M N 0 00 w 0 M O 0 M q V M M M N M 4 ff! ? yb + ?t fi P . It .r a M M a M 4J ttl w w 84 40 ca p to A m x ?? x x a m x x x 4J v 'i N M eN O P L7 F U W m b 14 b 14 yy Fi U V F 4J U yy F 4J f) P N 0 {~ q -4 Ol H 44 ++ a a U P+ a P4 F W W aI. ,$ rl N M V to m t- 00 Ot 0 ' V UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788ROO1800050001-8 Approved For Release 2044NQ:AS 0788ROO1800050001-8 0 N N M ~N N t0 t'? 00 Ol 0 r4 Cl 1 $ .4 m d' M 0 OD Q1 0 N In w N N O N N M M M N ti Cl .d Cl rl 0 0 Cl - M rd 0 W V N m m 0 94 -P 4.3 m 0m -1 as m U m M r m H Q -1 m N an H QQ N o F A' a .-4 .-1 -1 r rl 0 d 0 r+ i-1 0 u at mq 0) r4 b q 0 k ) ma N M V 44 0 b d k 4-) cV NO 7 H 1 4) 4A d 44 w F4 91 O b r-i 1H 14 u a a V a . + 8 v a C x u a a w 0 w .,1 ? ,$ r1 N M N fa t. 00 Ql ~d-1 UNCLASSIFIED o A. U 4) UD U 0 0 0 Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20UN A F{EDO788ROO18OOO5OOO1-8 8 f0 O - N M w w t0 A ao Ol O - ." 1 N rl c N a .-1 O N N 8'1 a fn t0 OD M N O N N r -1 0 O N O N M N O O N N N C1 M N M i o a - , ep M .-~ r4 a .-4 u0 u7 N r N ~!' rl N N V N N N A yym q rl O c d N P a m cd !.. a~ N F4 P ~ yC~l a ~ yN 01 ~y ? a m d a a N 0 44 13. Gi p N N M 41 V 44 01 w +, H t7 vv W ~ A d :3 4J v 91 gp 0 ym (A c~ F -H r4 ?~ F1 F7 F ail 3 N N E H 00 Oi O -I N M V N UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788ROO1800050001-8 Approved For Release 20(W/WQJM$R 1D0788R001800050001-8 x o N M r' n ro n ao o O r+ N .i rMi O rl N M V In w w Qb 0 CQ N N N a O M O O M O O M M N M Sp O N N N ep "4 V I t eM N M ~ y ~ , M M M M M M ifi M N N ~O U ~p ~" 2 F Ol 14 0 a ri ., N r, O r~ O O .~ H N H H 44 b N M ~p q d is 4J 4J 44 W -H W CF7 W Fi A q F $ h .~ r? N q H t0 A OD Of O N rl N rl M rl rl N rl UNCLASSIFIED y 4-) u G 0 0 w 0 Z* Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20 UNGLA1S 00788ROO1800050001-8 8 tp O rl N M ~M h m N OD fi O - . N rl M N -d O .-~ N M N t9 CD N N N O O N O N ca V eM O aw M 'd' 6l t. 6) -'~i d M V~ M r-1 .-1 ~-1 eP rl I!7 N IC) a 112 , eM er V~ N 'd~ eN V' tf) V~ N eM N 4H 0 4J W r4 :OH d ld 4 H 4 14 x "q H 02 H m+ CA 4 4 E : N 93 4J A cd V. N CI bD h0 N \ rl rl M 44 M A 4~ ~ 4J W (D 3 q d ~ 0 V ~ ~ Qdy p U~ F q A v W A m 0)F E ti -a i + A i o" R N M a v, w ti 00 Of 1.4 UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788ROO1800050001-8 Approved For Release 20 *L-A1SC6&O788ROO1800050001-8 O O N M ~N M t0 h 00 01 0 r4 cq ~ - CO O ~M N t0 00 Of 0 M O O .1 N C ) r -i . i N N ~p O O 0 m 0 1f) N O N 0 ti 0) N M Cl eM rl r-1 rl ~-1 V~ rl ~ N r-1 a .12 to M LO eN N M N 44 0 0 0 0 41 q i 44 ~ 4+ 4~ ~ 44 4J ?C A 07 ~ i0 O l 14 cd -1 cd 0 ' 4) ~'rpQ , Q 4) 4) 'N~ LO) . F ey N 1-1 a o -I ,-q .4 o r, N .r .i 1-4 H 1-1 1-1 N 4J co 4D ri N r4 C ~p Q ty d $4 k Y +~ d d F 1 { Y a N a b 0 a"'i F v~' ~ ?yp W 3 C7 G1 y Gi ~i q r [~ 3~ e ,~ .~ N M V' M t0 h 00 01 ~ ~ ~ N N to UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 20UM6LA& D0788ROO18OOO5OOO1-8 (U) Two types of overall assessment were chosen to emphasize the versatility of the evaluation procedure, (1) a simple rank ordering based on weighted average scores, and (2) a concept-by-concept, non-parametric, statistical technique. Table A-6 shows the results of the first method, the rank ordering. For convenience, the correct matches are underlined. (U) A RANK ORDERING OF WEIGHTED AVERAGES (U) Session/Target Inverness Hango Fernando de Noronha Punkaharju 2 4.13* 2.72 0.22 2.11 3 1.12 2.65 2.49 2.42 4 1.21 1.38 2.22 2.21 1 1.53 2.52 2.36 0.90 Scores computed with non-uniform target relevance factors. (U) From Table A-6, we see that there were 3 first-place matches and 1 fourth-place match. The probability of obtaining 3 of 4 possible first-place matches from chance fluctuations alone are less than 0.051. The point spread between the best match (Inverness) and the worst match (Punkaharju) are in qualitative agreement with a subjective "first look" at the quality of the transcripts as well. (U) The second analysis determines the significance of the difference between the correct concept/target matches and a control set of matches. All concept/target matches that are not the correct matches act as an internal control set. To avoid any invalid assumptions as to the correct parent distribution, a non-parametric statistical test, the Mann-Whitney U-Test, was chosen for the analysis.4 UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788ROO1800050001-8 Approved For Release 20UNQA&59J O788ROO1800050001-8 (U) It is beyond the scope of this report to review the details of the Mann-Witney U-Test; thus, only the results are quoted here. The probability that the set of correct concept/target matches is statistically indistinguishable from the control concept/target matches is less than 0.071. (U) There are a number of additional statistical procedures that could be used to analyze the results of this evaluation technique. The two cited above, however, represent a spread in complexity that demonstrates the internal consistency of the basic evaluation procedure. With only four similar RV sessions, the evaluation technique nearly reached the 0.05 level of statistical significance with each of the two statistical procedures, a result indicating a successful outcome with regard to the overall assessment procedure. UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 200'Oi08/IIS" C1A'FM9'6=00788 R001800050001-8 Appendix B (U) SUMMARY OF EVALUATION TECHNIQUE (U) Step Action 1 Task coordinator defines the evaluation goal. He/she identifies target elements and assigns target element relevance factors as appropriate. 2 Analyst conceptualizes responses and prepares an RV assessment sheet for each response. 3 Repeated concepts are noted in the "D" column. 4 Copies of the sheets from Item 2 are made; one for each possible target used in the analysis. FOR EACH POSSIBLE RESPONSE/TARGET COMBINATION: 5 Identify a target element for each concept not marked in the "D" column; mark a 1 in the "p" column and write the target element and its relevance factor from Step 1 in the appropriate columns. (Write 0 and blanks if no element can be found.) 6 Using Table 3, assign a quality rating for all present (p = 1) concept/element combinations. 7 Compute the score as follows: a. Calculate relevance (R) X quality (Q) b. Convert R X Q to an integer between 0 and 14 using the conversion table provided. 8 Enter the number of concepts that obtained each possible score in the space provided. 9 Calculate the weighted average using: l Ak = 0.357 [Z fj /f i Sj/E fj ff where: S. is the score and f. is the number of concepts tl.at obtained score j = 0, 1, 2, ..., 14 10 For each response, rank order the weighted averages. UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8 Approved For Release 200 GLASSWbE-D788R001800050001-8 1. H. E. Puthoff and R. Targ, "A Perceptual Channel for Information Transfer over Kilometer Distances: Historical Perspective and Recent Research," Proc. IEEE, Vol. 64, No. 3, pp. 329-354 (March 1976). 2. C. Scott, "On the Evaluation of Verbal Material in Parapsychology: A Discussion of Dr. Pratt's Monograph," J. Soc. Psych. Res., Vol. 46, No. 752, pp. 79-90 (June 1972). 3. R. Targ, H. E. Puthoff, and E. C. May, "State of the Art in Remote Viewing Studies at SRI," Technical Session on Research in Psycho- energetics, Proc. 1977 IEEE International Conf. on Cybernetics and Society, Washington, D.C. (September 20, 1977). 4. S. Siegel, Nonparametric Statistics, pp. 116-120 (McGraww-Hill, 1956). UNCLASSIFIED Approved For Release 2000/08/08 : CIA-RDP96-00788R001800050001-8