information and uncertainty in remote perception research · 2016-12-07 · information and...

Information and Uncertainty inRemote Perception Research

BRENDA J DUNNE AND ROBERT G JAHN

Princeton Engineering Anomalies ResearchPrinceton University

Princeton NJ 08544-5263e-mail pearlabprincetonedu

AbstractmdashThis article has four purposes 1) to present for the first time inarchival form all results of some 25 years of remote perception research at thislaboratory 2) to describe all of the analytical scoring methods developed overthe course of this program to quantify the amount of anomalous informationacquired in the experiments 3) to display a remarkable anti-correlationbetween the objective specificity of those methods and the anomalous yield ofthe experiments and 4) to discuss the phenomenological and pragmaticimplications of this complementarity The formal database comprises 653experimental trials performed over several phases of investigation The scoringmethods involve various arrays of descriptor queries that can be addressed toboth the physical targets and the percipientsrsquo description thereof the responsesto which provide the basis for numerical evaluation and statistical assessment ofthe degree of anomalous information acquired Twenty-four such recipes havebeen employed with queries posed in binary ternary quaternary and ten-leveldistributive formats Thus treated the database yields a composite z-scoreagainst chance of 5418 ( p 5 3 3 10 2 8 one-tailed)

Numerous subsidiary analyses agree that these overall results are notsignificantly affected by any of the secondary protocol parameters tested or byvariations in descriptor effectiveness possible participant response biasestarget distance from the percipient or time interval between perception effortand agent target visitation However over the course of the program there hasbeen a striking diminution of the anomalous yield that appears to be associatedwith the participantsrsquo growing attention to and dependence upon theprogressively more detailed descriptor formats and with the correspondingreduction in the content of the accompanying free-response transcripts Thepossibility that increased emphasis on objective quantification of thephenomenon somehow may have inhibited its inherently subjective expressionis explored in several contexts ranging from contemporary signal processingtechnologies to ancient divination traditions An intrinsic complementarity issuggested between the analytical and intuitive aspects of the remote perceptionprocess that like its more familiar counterpart in quantum science brings withit an inescapable uncertainty that limits the extent to which such anomalouseffects can be simultaneously produced and evaluated

Keywords remote perceptionmdashremote viewingmdashanomalous informationacquisitionmdashconsciousness-related anomaliesmdashuncertaintymdashcomplementaritymdashPEARmdashengineering anomaliesmdashanalyticaljudging

Journal of Scienti c Exploration Vol 17 No 2 pp 207ndash241 2003 0892-331003

207

Man also possesses a power by which he may see his friends and thecircumstances by which they are surrounded although such persons may bea thousand miles away from him at that time

Paracelsus

I Introduction and Background

This concise statement of the remote perception hypothesis was proffered by therenowned 16th-century physician and philosopher Paracelsus in a section of hiswritings devoted to the role of lsquolsquoactive imaginationrsquorsquo in manrsquos representation ofhis universe(1) His observation was certainly not the first recorded allusion tosuch anomalous human capabilities This lsquolsquopowerrsquorsquo has been acknowledged invirtually every culture since the dawn of human civilization and invoked undera multitude of names including among many others divination prophecyoracle scrying clairvoyance and second sight

In the more recent history of Western science a considerable body ofliterature describing scholarly investigations of lsquolsquoextrasensory perceptionrsquorsquoalready had been amassed when in the mid-1970s Puthoff and Targ at StanfordResearch Institute introduced a new scientific protocol for empirical in-vestigation of the phenomenon they termed lsquolsquoremote viewingrsquorsquo(23) Theirprocedure required one individual referred to as the lsquolsquopercipientrsquorsquo to attempt todescribe the geographical ambience surrounding another person the lsquolsquoagentrsquorsquowhose location was inaccessible to the percipient by any known sensory meansTheir striking data included many perceptions that were virtually photographicin accuracy and produced an overall statistical yield well beyond chanceexpectations Over the subsequent quarter century numerous replications of theoriginal SRI studies have been reported(4ndash16) including a number of originallyclassified government-sponsored investigations(17ndash21) most of which display theambiguous mixtures of successes and failures that seem to characterize mostserious anomalies research Notwithstanding the majority of these studiesdemonstrate a sufficient degree of anomalous information acquisition to justifycontinued scholarly exploration of this mystifying process

One of the largest extant databases comprising 653 formal and 126 non-formalexperimental trials was produced between 1976 and 1999 as one of the threemajor components of the Princeton Engineering Anomalies Research (PEAR)program The other two segments concerning anomalies in humanmachineinteractions and theoretical modeling have been reported extensively in thisjournal and elsewhere The purpose of this paper is to describe the procedures andsummarize the full results of our remote perception studies and to explore theirimplications for better comprehension of this currently inexplicable communi-cation capability To achieve this most concisely we shall refer frequently toa number of earlier publicationsand technical reports wherein all the datasets andanalytical methods are presented in greater detail(22ndash26)

The first phase of this PEAR work evolved from a body of prior experimentsconducted between 1976 and 1979 by one of the authors (BJD) at Mundelein

B J Dunne amp R G Jahn208

College in Chicago and subsequently at the University of Chicago(2728) whichutilized human-judge ranking procedures similar to those of the earlier SRIstudies(29) Despite the impressive yield of these experiments concernsregarding evident vagaries and possible subjective biases in the judgesrsquointerpretations or even anomalous inputs on their part predicated a morequantitative approach to data evaluation(30) A primary focus of the subsequentPEAR studies has been on the development of analytical judging procedurescapable of rendering the free-response raw data into forms amenable to morerigorous quantification and analysis Beyond the acquisition and analysis oflarge composite databases a number of secondary experimental variables suchas the effect of multiple percipients alternative target selection procedures andthe dependence of the phenomenon on spatial and temporal separations havealso been explored Inspired by a section of Puthoff and Targrsquos 1976 paper(3)

wherein they alluded to the ability of some of their percipients to describe targetscenes even before the target had been identified much less visited the majorityof the PEAR trials have been acquired in this precognitive mode And sincemany of the percipients maintain that their experiences are not strictly speakingof a simple visual nature the term lsquolsquoprecognitive remote perceptionrsquorsquo or PRPhas been preferred

II Protocol

In its basic form the PEAR protocol requires a percipient to describe anunknown remote geographical target where an agent is was or will be situatedat a prescribed time The target location is selected randomly before each trialfrom a large pool of potential targets prepared previously by an individual nototherwise involved in the experiment The contents of this pool are stored inseparate sealed envelopes randomly numbered and maintained so that no agentor percipient has access to them Prior to a given trial the target is designated bygeneration of a random number that identifies one of the envelopes which thenis delivered still sealed to the agent who opens it and follows instructions tolocate the target This lsquolsquoinstructedrsquorsquo mode of target selection is complemented bya lsquolsquovolitionalrsquorsquo protocol option typically followed when the agent is traveling onan itinerary unknown to the percipient in a region for which no prepared poolexists In these trials the agent simply selects the target from among the variouslocal sites accessible at the time specified for the trial

In either version the percipient is asked to spend 15 to 20 minutes attemptingto visualize or experience the target and to record these impressions in a free-response stream-of-consciousness form either orally into a tape recorder or inwriting optionally including drawings Unlike some of the procedures followedat SRI and elsewhere where percipients are trained to use particular strategiesor where perceptions are generated in a laboratory setting with an experimenterpresent and actively eliciting information PEAR percipients are free to choosetheir own subjective strategies and physical locations and experimenters are not

Remote Perception Research 209

present during the perception process While the majority of data have beenacquired in the precognitive mode wherein the perceptions are generated andrecorded before the target is selected a substantial subset of trials have beenexecuted in a retrocognitive mode wherein perceptions are generated after theagent has visited the target and a smaller number have been performed in lsquolsquorealtimersquorsquo In all cases strict precautions are taken to ensure that perceptions arerecorded and filed before percipients have any sensory access to informationabout the targets and no ordinary means of communication between percipientsand agents is available until after that point

The agents who in almost all cases are known to the percipients are asked tosituate themselves at the target sites at the agreed-upon times and to immersethemselves in the scenes for about 15 minutes At the close of the visitationperiods they record their impressions of the target scenes supplementing themwith hand-drawn sketches if desired and whenever possible by one or morephotographs to corroborate their verbal descriptions Like the percipients agentsare free to employ their own subjective strategies They simply are encouragedto attempt in some way to share their target experiences with the percipients

All of the participants in the PEAR experiments have been uncompensatedvolunteers none of whom has claimed exceptional abilities in this regard Noexplicit tactical instructions are given although an attitude of playfulness isencouraged and emphasis is placed on enjoymentof the experience rather than onachievement per se Transcript styles of individual percipients vary widelyranging from a few cryptic details at one extreme to lengthy impressionistic flowsof imagery on the other No systematic records have been maintained on therelative effectiveness of the various personal strategies deployed by theparticipants or on any of their psychological or physiological characteristicsThey are encouraged however to furnish subjective reports of their experiencesand these anecdotaldescriptionshave providedvaluableglimpses into some of themore qualitative aspects of the underlying process For example severalpercipients have commented that they found it helpful to clear their mindsvisualize a blank screen and wait for an image of the agent to appear Some agentsreport that they imagine that the percipients are with them at the target scene andthat they carry on mental conversationswith them pointingout various aspects ofthe sites On some occasions agents have observed that they found their attentiondrawn to components of the scene that they had overlooked initially only todiscover later that these features had been part of the percipientrsquos descriptionsalmost as if the percipientrsquos consciousness had guided their attention Manyparticipants have indicated that they feel more like they are sharing a commonexperience rather than lsquolsquotransmittingrsquorsquo information from one person to another

III Analytical Judging Methods Development and Initial Applications

As mentioned earlier evaluation of the original Chicago experiments thathad produced highly significant statistical results had been based on rankings


assigned by independent human judges to each of the free-response perceptionswhen compared with photographs of all the targets in its local series(30) Toassess the potential statistical impact of inter-judge variability in those studies27 transcripts comprising the first three experimental series had been subjectedto repeated re-judging by five separate individuals Although approximately halfof these trials demonstrated a strong consistency in the ranks assigned by boththe primary and secondary judges and confirmed the acquisition of significantextra-chance information the others received a wide range of ranks suggestingthat the matches originally assigned to these trials had most likely been arbitraryAlso evident in this review was the inherent inefficiency of an approachwhereby the entire informational content of a given perception was reduced toa single datum ordinal at best in a small experimental series

Beyond the accumulation of new empirical data the first major thrust ofthe embryonic PEAR program was an attempt to alleviate some of theseshortcomings by developing standardized methods of quantifying the in-formation content of the free-response data via a series of computer algorithmsThe first step in this direction was the establishment of a code or alphabet of 30simple binary descriptive queries that could be addressed to all targets andperceptions The questions ranged broadly from factual eg whether the scenewas indoors or outdoors whether water was present etc to more impression-istic eg whether the scene was confined or expansive noisy or quiet etc Theresponses entered into a computerized database manager as strings of 30 bitswere submitted to an assortment of analytical scoring algorithms that couldprovide numerical evaluation of the thus-specified information content of anygiven trial and once scored the statistical merit of the perception results couldbe evaluated by an assortment of computerized analytical ranking procedures(22)

Specifically the algorithms scored each transcript against all the targets in thepool and then ranked them in order of descending score

While still dependent upon a ranking procedure this descriptor-based processhad the advantages that such ranking could proceed on a more standardizedanalytical basis and that many more alternative targets could be ranked by thecomputer than by a human judge As a first test of this approach one series ofeight trials from the earlier Chicago database was encoded ex post facto into thebinary format by five independent encoders Reassuringly most of the responseswere found to be in close agreement with each other ie the computer-assignedranks of the better trials were highly consistent with those of the original humanjudges and those of the weaker trials were comparably equivocal

With these scoring methods so qualified 35 new trials were generatedfollowing the same protocol used in the earlier experiments but now the targetsand perceptions were descriptor-encoded ab initio by the agents at the targetsites and by the percipients after completing their free-response descriptionsAlthough the statistical results of these new trials were not as strong as those ofthe ex post factondashencoded data they were still highly significant Perhaps evenmore importantly the general agreement among the various scoring algorithms


confirmed that the analytical methodology was indeed capable of providingreliable quantification of the intrinsically impressionistic remote perception dataTo obviate the possibility that the particular list of descriptors employed somehowcould process even random inputs to apparently significant scores a lsquolsquocalibrationrsquorsquoexercise was undertaken wherein artificial lsquolsquotargetrsquorsquo and lsquolsquoperceptionrsquorsquo datamatrices of the same size as the actual data matrices were constructed from theoutput of a random event generator The same computational schemes wereapplied to various combinations of these both with each other and with the truedata with results that were all well within chance expectation(25)

With growing confidence in the viability of this analytical methodology anadditional 51 prior trials from Chicago and PEAR were then transcribed into thenew descriptor format increasing the total number of ex post factondashencoded trialsto 59 comprising all the original human-judged trials that met formal protocolcriteria and had adequate target documentation to permit such retrospectiveencoding Here and henceforth formal trials are defined as those that follow thestandard protocol described earlier and also meet all of the following criteria

1 The agent and percipient are specified to one another2 The date and time of the agentrsquos target visitation are specified to the

percipient3 The agent is present at the target within 15 minutes of the specified time and

is consciously committed to his or her experimental role during that period4 Both agent and percipient produce verbal descriptions and complete the

descriptor response forms5 Both agent and percipienthave adequate familiarity with the applicationand

interpretation of the descriptor questions and with the general protocol6 Photographs written descriptions or other substantiating target informa-

tion are available

By 1983 the 59-trial ex post factondashencoded database had been supplementedby 168 new ab initiondashencoded trials plus 73 others that for various reasons didnot meet formal protocol criteria bringing the total to 300 Of the non-formaltrials 21 were categorized as lsquolsquoquestionablersquorsquo where failure to meet the formalcriteria was due to protocol violations such as the lack of adequatesubstantiating target information evidence that one or both of the participantsdid not understand the application or interpretation of the descriptor questionsor the vulnerability of the trial to sensory cueing Another 52 trials weredesignated in advance as lsquolsquoexploratoryrsquorsquo wherein intentional deviations fromformal protocol such as deliberately not informing the percipient of the agentrsquosidentity or not specifying the time of target visitation were undertaken(24)

IV Statistical Evaluations via Empirical Chance Distributions

Beyond its evident success in dispassionate ranking of the trials in any givenexperimental series the descriptor-based scoring method offered a far more


desirable and powerful capability ie the direct calculation of the statisticalmerit of individual trial scores or groups of scores To achieve this an empiricallsquolsquochancersquorsquo distribution was constructed by scoring every perception in the 300-trial database against every possible target except its correct one thuscompounding a large array of deliberately mismatched scores the distributionof which displayed classical Gaussian features and could serve as a statisticalreference Several variationsof this scoring techniquewere explored all of whichconsisted of calculating a score for each trial based on the proportion of matchesand mismatches in the percipient and agent responses to the 30 descriptor queriesusing a set of generalized a priori probabilities derived from the 300 targetscomprising the database as descriptor weighting factors For example since moretargets tended to be outdoors than indoors a correct positive response to the querylsquolsquoIs the scene indoorsrsquorsquo was assigned a greater weight than a correct negativeresponse and its incremental contribution to the total score was proportionatelylarger The sum of the score increments from all 30 descriptors constituted thelsquolsquoabsolute scorersquorsquo for a given trial which was then divided by some normalizingfactor such as the maximum score that would have been achieved had all 30target and perception descriptor responses agreed yielding a lsquolsquonormalized scorersquorsquoThe statisticalmerit of this normalized score was then established by comparing itwith the chance distribution of similarly normalized mismatched scores

The descriptor response check sheets also contained a column labeledlsquolsquounsurersquorsquo in addition to the standard lsquolsquoyesrsquorsquo and lsquolsquonorsquorsquo options which permittedparticipants to indicate any ambiguities they might experience in relating theirsubjective impressions in strictly binary terms These lsquolsquounsurersquorsquo responses weredisregarded in the binary calculations but they provided the basis forinvestigating the potential benefits of ternary-based algorithms(23) Seven suchternary scoring methods were explored all of which showed good internalconsistency but none of which indicated any substantial advantage over thebinary calculations Given their added computational complexity subsequentstudy was limited to only five binary-based methods

deg Method A The number of descriptors answered correctly divided by thetotal number of descriptors (ie a count of the numerical fraction ofcorrect responses ignoring the a priori descriptor probabilities)

deg Method B The sum of all descriptors answered correctly each weighted bythe reciprocal of its a prioriprobabilitydividedby the sum of all descriptorsso weighted (This method weighted the value of correct responses ininverse proportion to their a priori probabilities and normalized the scoreby the highest possible score obtainable by this method for a given target)

deg Method C The same numerator as Method B divided by the total numberof descriptors normalized by the lsquolsquochancersquorsquo score derived from the a prioriprobabilities

deg Method D The sum of all descriptors correctly answered lsquolsquoyesrsquorsquo eachweighted by the reciprocal of its a priori probability plus the unweighted


sum of all descriptors answered lsquolsquonorsquorsquo the total divided by the sum of alldescriptors labeled lsquolsquoyesrsquorsquo in the target each weighted by the reciprocal ofits a priori probability plus the unweighted sum of all descriptors labeledlsquolsquonorsquorsquo in the target with the resultant score weighted by the highestpossible score for that target (This process effectively removed from thecalculation those descriptors on which the percipient responded nega-tively whether correctly or incorrectly and thereby served to counterveneuse of a negative response to imply ignorance of the descriptor rather thanits explicit absence)

deg Method E The same numerator as Method D divided by the total numberof descriptors ie by the lsquolsquochancersquorsquo score

Table 1 summarizes the results of these 300 trials grouped by experimentalcriteria as assessed by each of these five recipes

The most instructive feature of these results is the consistency of anomalousyield across these five diverse scoring schemes Regardless of the algorithmemployed for all but the exploratory trials the composite results indicate highlysignificant increments of anomalous information in the matched scores that arenot present in the mismatched score distributions constructed from the same rawdata Even the null results of the 52 exploratory trials are informative in theirindication that the features violated in these excursions from the standardprotocol ie the percipientsrsquo knowledge of the agent or of the time of targetvisitation may be requisites to generation of the anomalous effect Given theevident insensitivity of the results to the particular scoring strategy deployed itwas agreed that only one method would henceforth be used as the standard forevaluating future binary-encoded trials Method B was selected for this purposesince it treated positive and negative descriptor responses in a symmetrical andintrinsically normalized fashion

These results made it clear that the new analytical methodology was capableof relatively objective quantitative assessment of the inherently subjectiveremote perception phenomenon Unlike the less efficient labor-intensive humanjudging methods it not only could calculate individual trial scores but couldprovide robust indications of the statistical quality of large databases On theother hand the analytical judging process introduced certain imperfections of itsown For example the forced lsquolsquoyesrsquorsquo or lsquolsquonorsquorsquo responses were limited in theirability to capture the overall ambience or context of a scene or nuances ofsubjective or symbolic information that might be detected by human judgesFurthermore while restricting the extracted information to the 30 specifiedbinary descriptors minimized the reporting task for the participants it precludedutilization of other potentially relevant features in the transcripts such asspecific colors textures architectures or any other details not covered by thequestions These shortcomings were partially offset by the continued re-quirement that percipients first generate free-response descriptions from whichthe descriptor responses were then derived a procedure intended to retain the


spontaneity of the PRP experience as well as to preserve the raw data ina suitable format for further study Nonetheless it became evident that afterseveral experiences with the descriptor utilization many participants tended tolimit their attention and descriptions to those features that they now knew werespecific to the questions

These limitations notwithstanding the evident advantages of the analyticaljudging techniques encouraged further exploration beginning with a compre-hensive evaluation of the effectiveness of the individual descriptors inconstructing the trial scores From this it was determined that the entire groupof descriptors originally selected by some combination of anecdotal experienceand intuition actually comprised a reasonably uniform set in terms of theireffectiveness in quantifying informational bits across a broad range of targettypes None was found to be extremely effective none was seriously deficientSub-division of the descriptors into classifications of natural vs man-made

TABLE 1Summary of Binary PRP Data as of 1983

Scoringmethod

Chancemean

ChanceSD

Meanscore

Compositez-score

Probability(one-tailed)

Trialsp 05

Trialsp 05

Formal data (N 5 227)A 05610 1053 06113 7197 3 3 102 13 28 (4) 12 (2)B 05042 1207 05590 6833 4 3 102 12 40 (6) 18 (3)C 10005 2380 11101 6941 2 3 102 12 35 (5) 14 (2)D 06512 0935 06926 6672 1 3 102 11 33 (6) 15 (3)E 10034 1330 10676 7277 2 3 102 13 35 (4) 14 (2)

Formal plus questionable data (N 5 248)A 05610 1053 06071 6894 3 3 102 12 30 (4) 12 (2)B 05042 1207 05536 6442 6 3 102 11 42 (7) 17 (3)C 10005 2380 10998 6574 2 3 102 11 37 (6) 15 (2)D 06512 0935 06887 6321 1 3 102 10 34 (6) 14 (2)E 10034 1330 10619 6924 2 3 102 12 37 (4) 15 (2)

Exploratory data (N 5 52)A 05610 1053 05538 2 0493 (31) 0 (3) 0 (6)B 05042 1207 05023 2 0115 (45) 2 (3) 4 (6)C 10005 2380 10277 0824 20 3 (2) 6 (4)D 06512 0935 06419 2 0719 (24) 1 (2) 2 (4)E 10034 1330 10246 1148 13 5 (1) 10 (2)

All data (N 5 300)A 05610 1053 05979 6070 6 3 102 10 30 (7) 10 (2)B 05042 1207 05447 5809 3 3 102 9 44 (10) 15 (3)C 10005 2380 10873 6320 1 3 102 10 40 (8) 13 (3)D 06512 0935 06806 5447 3 3 102 8 35 (8) 12 (3)E 10034 1330 10554 6773 6 3 102 12 42 (5) 14 (2)

Note The original version of this table published in Technical Report 83003 contained an error thatinadvertently inflated the results from Method A suggesting that this method produced larger effectsthan the others With this corrected the results are reasonably consistent across all five methods Numbers in parentheses indicate number of trials with negative z-scores p 05


objective vs subjective permanent vs transient and indoor vs outdoor alsoindicated no significant differences in effectiveness The interdependenceamong the various descriptors eg that outdoor scenes were less likely to beconfined or that indoor scenes were less likely to involve airplanes or roadvehicles was also explored by a variety of statistical methods all of whichconfirmed that while such correlations might blunt the incisiveness of the fulldescriptor net somewhat they could not compromise the validity of theresults(242531)

Thus by the close of this phase of the program a number of useful generalconclusions had emerged

1 Although the various methods produced differing scores for some of theindividual trials the overall statistical yield was uniformly highlysignificant and relatively insensitive to the particular scoring andnormalizing recipes employed

2 There was general agreement between the results of the various analyticalmethods and those of the impressionistic assessments by human judgesparticularly for the perceptions of higher statistical merit

3 The use of ternary descriptor responses wherein participants were offeredthe option of lsquolsquopassingrsquorsquo on a given descriptor did not yield sufficientlymore consistent or accurate results compared to the binary methods tojustify the added computational complexity

4 Defining a lsquolsquouniversalrsquorsquo target pool in terms of a sufficiently large numberof actual targets made it possible to calculate a set of generalized a prioridescriptor probabilities that could be used for scoring any individualperception efforts in the database regardless of its particular local seriespool

5 Calculation of the statistical merit of individual perception efforts byreference to an empirical chance distribution derived from a large numberof deliberately mismatched targets and perceptions proved to be a farmore powerful strategy than the computerized analytical ranking withinindividual small series

6 The 30 descriptors originally chosen through a combination of empiricismand intuition although clearly non-independent nonetheless displayeda reasonably flat profile of effectiveness in building the scores of thesignificant transcripts

V Secondary Parameters

With the effectiveness of the analytical methodology thus established and thecomputerized ranking procedures superseded by the more powerful statisticalprocedure that compared the scores of individual trials or groups of trials witha lsquolsquouniversalrsquorsquo mismatch distribution a second phase of ab initiondashencoded datageneration was initiated that extended over several years Since the protocols


descriptor questions and scoring algorithms remained identical to thosedeployed in the previous phase these new trials could legitimately be combinedwith the earlier data to provide a larger database for structural segmentations By1988 the total PEAR PRP binary-descriptor database consisted of 411 trialsproduced by a total of 48 participants Of these 336 trials qualified as formal 54as exploratory and 21 as questionable Of the 336 formal trials 125 followedthe instructed protocol wherein the target was selected at random from a pre-existing pool and 211 utilized the volitional protocol wherein the agent was inan area for which no prepared pool existed

Sorting the data by another criterion 291 trials 216 of which qualified asformal were generated under the standard protocol wherein a single percipientattempted to describe the location of a single agent In the remaining 120 trialsall of which met the formal criteria two or more percipients addressed the sametarget The number of percipients addressing a given target ranged from two toseven and each perception was scored as a separate trial against its appropriatetarget In all but two of the multiple-percipient trials the percipients were awarethat others were involved in the experiment although they did not always knowtheir identities The participating percipients always were separated spatiallyfrom each other and in most cases attempted their perception efforts at differenttimes One series of formal trials and a few of the exploratory trials involvedmore than one agent but in each of these cases only one pre-specified set oftarget encodings was included in the scoring process the second set was usedonly for informal comparison

Table 2 presents the summary statistics obtained using binary Method B forthis combined PRP database and its various subsets The empirical chancedistribution used as a reference was derived from all the formal trials in thissame database and comprised more than 100000 mismatched scores Inaddition to the subsets addressing planned variations of the protocol eg abinitio vs ex post facto encoding single vs multiple percipients and instructedvs volitional assignment of targets summaries for ad hoc subdivisions of thedatabase by seasonal and regional target groupings are also included For eachindependently calculated subset the table displays the number of trials themean score the effect size (defined as the mean z-score of all the trials in thegiven subset) with associated 99 confidence intervals the standard deviationof the trial z-score distribution (expectation 5 1) and the composite z-score(calculated by multiplying the effect size by the square root of the numberof trials in the subset) with its associated one-tailed probability against chanceThe last three columns list the number of trials in each subset with z 1645( p 05) (numbers in parentheses indicate z 2 1645) the correspondingpercentage of those significant trials and the percentage of scores where p 50(greater than the chance mean score) Each group is scored using the locala priori descriptor probabilities associated with that subset and except for thegroups labeled lsquolsquoAll Trialsrsquorsquo and lsquolsquoNon-Formal Trialsrsquorsquo the various subsetsconsist of formal trials only All are calculated with reference to the universal


TA

BL

E2

Bin

ary

PRP

Dat

aS

umm

arie

s(S

cori

ngM

etho

dB

)

Sub

set

T

rial

sM

ean

scor

eE

ffec

tsi

ze

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Prob

abili

ty(o

ne-t

aile

d)

Tri

als

p

05

T

rial

sp

0

5

Tri

als

p

50

All

tria

ls41

15

364

279

61

351

060

564

78

310

29

47(1

2)11

(3

)

59

For

mal

tria

ls33

65

447

347

61

521

083

635

51

310

210

44(8

)13

(2

)

62

Non

-for

mal

tria

ls75

496

92

046

62

780

910

20

399

655

3(4

)4

(5

)44

Ab

initi

o27

75

345

263

61

611

033

437

86

310

26

31(5

)11

(2

)

59

Ex

post

fact

o59

594

27

546

417

120

35

792

33

102

914

(2)

24

(3

)75

Sin

gle

perc

ipie

nt21

65

489

382

61

941

098

561

31

310

28

34(6

)16

(3

)

60

Mul

tiple

perc

ipie

nt12

05

404

312

62

511

049

341

63

310

24

12(3

)10

(3

)

63

Inst

ruct

edta

rget

s12

55

653

516

62

671

140

577

14

310

29

23(5

)18

(4

)

65

Vol

itio

nal

targ

ets

211

532

22

446

191

106

63

549

23

102

425

(3)

12

(1

)60

Sum

mer

tria

ls24

45

466

363

61

831

099

566

37

310

29

35(5

)14

(2

)

65

Win

ter

tria

ls92

540

73

156

286

104

33

017

13

102

313

(2)

14

(2

)57

Chi

cago

targ

ets

316

189

957

65

871

189

533

05

310

28

10(1

)32

(3

)

81

Pri

ncet

onta

rget

s10

65

504

394

62

861

110

406

02

310

25

14(3

)13

(3

)

62

Tar

gets

else

whe

re19

95

267

199

61

941

051

281

02

310

23

20(3

)10

(2

)

58

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05


chance distribution of mismatched scores (N 5 106602 mean 5 5025 andstandard deviation 5 1216)

The overall results of these analyses leave little doubt by any criterion thatthe PRP perceptions contain considerably more information about the designatedtargets than can be attributed to chance guessing Although the superior results ofthe ex post facto trials relative to the ab initio trials are particularly striking littledifference is found between single- and multiple-percipient performances andthere is no evidence of seasonal dependencies (In assessing these results it isimportant to keep in mind that the statistical z-scores reflect both the averageeffect size and the number of trials in each subset So for example although thesingle-percipient data produce a substantially larger z-score than the smallermultiple-percipient subset their relative effect sizes are very close and the largeconfidence intervals indicate that the two groups are statistically indistinguish-able Similar remarks pertain to the seasonal discriminations)

The substantial difference between the yields of the ex post facto and ab initiodata raise some concern that the former on which the descriptor questions andmethodology initially had been based could have introduced a spurious scoreinflation into the composite database Therefore these analyses were repeatedusing only the formal ab initio data The composite results of these 277 trialspresented in Table 3 continue to display a robust overall effect and confirm thatthe bottom-line yield of the overall PRP database cannot be discounted on thebasis of any such inflation It is interesting to note however that in thissomewhat more restricted dataset the difference between the instructed andvolitional subsets is considerably smaller and only marginally significant andthe geographical distinction between Princeton targets and those elsewhere oncethe ex post facto Chicago trials are excluded becomes statistically non-significant

The difference between the average effect sizes of the instructed and volitionaltrials is worth closer examination since these two subsets might have beenexpected to display disparities in their empirical a priori descriptor probabilityestimates Given the less formal nature of the target selection process in thevolitional trials it was possible that the agentrsquos knowledge of the percipientrsquospersonal preferences or target response patterns could have influenced the targetselection and representation thereby introducing an undue bias into the volitionaltrial scores In the full database summarized in Table 2 there was indeeda statistically significant difference between the results of these two subsets (z 5241) but it was actually the instructed subset that produced the larger effect sizeThe formal ab initio data only (Table 3) still showed a larger effect in theinstructed trials although the difference here was considerably smaller (z 5173) Thus the concern that the target selection process employed in thevolitional trials might have contributed to artificial enhancement of the resultsappeared to be unfounded If anything these comparisons suggested that thevolitional target selection process may actually have had an inhibitory effect onthe phenomenon rather than imposing an advantage


TA

BL

E3

For

mal

Ab

Initi

oD

ata

Sum

mar

ies

(Sco

ring

Met

hod

B)

Subs

et

Tri

als

Mea

nsc

ore

Eff

ect

size

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Pro

babi

lity

(one

-tai

led)

T

rial

sp

0

5

Tri

als

p

05

T

rial

sp

5

0

All

tria

ls27

75

345

263

61

611

034

437

86

310

26

31(5

)11

(2

)

59

Sing

lepe

rcip

ient

194

537

02

846

197

106

33

949

43

102

524

(6)

12

(3

)56

M

ultip

lepe

rcip

ient

835

321

243

62

750

974

221

50

135

(1)

6(1

)

64

Inst

ruct

edta

rget

s94

541

63

226

296

111

53

122

93

102

411

(5)

12

(5

)61

V

olit

iona

lta

rget

s18

35

308

233

61

941

020

314

88

310

24

21(1

)11

(

05

)60

Sum

mer

tria

ls19

55

374

287

61

951

058

401

33

310

25

24(4

)12

(2

)

62

Win

ter

tria

ls82

530

82

336

285

100

22

107

018

7(2

)9

(2

)56

Prin

ceto

nta

rget

s10

65

504

394

62

811

125

406

02

310

25

14(4

)13

(4

)

62

Tar

gets

else

whe

re17

15

243

180

61

971

000

234

89

310

23

16(1

)9

(05

)

59

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05


The magnitude and consistency of the anomalous yield in these data arepresented graphically in Figure 1 where the results of all 336 formal trials aredisplayed in the form of a cumulative deviation of the actual scores from chanceHere the stronger yield of the early ex post facto trials is strikingly evidentNonetheless the remainder of the trace while less steep also shows a clear andsystematic deviation from chance expectation

Further details on the analytical judging methodology and individual trialresults as well as examples of target photos and transcripts from some specifictrials may be found in Refs 24ndash26 32 and 33 and a process that verifies thatthe scores are not inflated by shared percipientagent coding biases is describedin Appendix A of this paper

VI Distance and Time Dependencies

Beyond the secondary parameters discussed in the previous section a numberof other variables were explored in the course of these experiments that provedhelpful in illuminating some of the fundamental characteristics of the anomalouscommunication process Two features of particular importance are thedependence of the results on the physical distance separating the percipientand the target and on the time interval between the perception effort and theagentrsquos visitation of the target The spatial distances in this database ranged fromless than one mile to several thousand miles and the temporal separations fromseveral days before to several days after target visitation Figures 2 and 3 displaythe results of regression analyses of the dependence of the trial scores on thesetwo parameters In each the horizontal dashed line denotes the empirical meanz-scores the central dotted line indicates the linear regression fits to the dataand the outer dotted lines are the 95 confidence intervals thereof Since theregressions are statistically indistinguishable from the lines of constant meanshift we conclude that within the ranges of this database there are nosignificant correlations of effect size with either distance or time In particularwhen a regression of the data is plotted as a function of the reciprocal square of

Fig 1 Cumulative deviation of 336 binary-encoded formal trials


the distance the results specifically refute any 1r2 dependence of the anomalouslsquolsquosignalrsquorsquo Furthermore if the data are segregated into subsets of the moreextreme spatially and temporally displaced trials and those more proximate theaverage effect sizes of the former remain statistically indistinguishable fromthose of the latter(2425)

The lack of evidence for attenuation of the remote perception yield withincreased distance or time severely limits the possibilities for theoreticalexplication in terms of any known physical process However these findings didprompt the testable hypothesis that other anomalies being explored by PEARmight display similar non-local characteristics and led to an extensive study ofremote humanmachine interactions Here again significant intention-correlatedmean shifts have been observed that are statistically indistinguishable from thosein the local experiments Not only are the scales of these anomalous effectsinsensitive to intervening distance and time but they display the same structuralpatterns as those of the corresponding local experiments(34) Indeed thesimilarities between the humanmachine and remote perception results providedthe first indications that these two forms of anomaly previously regarded asdistinct phenomena actually might derive from the same mechanism ofinformation exchange

VII FIDO Scoring

By 1985 the PEAR program had amassed a substantial body of experimentaldata that both confirmed the reality and robustness of the remote perceptionphenomenon and demonstrated the efficacy of the analytical scoring techniquesAlthough the ab initiondashencoded trials had produced a smaller average effect size

Fig 2 336 binary-encoded formal trial scores as a function of distance


than that of the ex post facto subset this was attributed primarily to an inherentadvantage for the earlier data of having the descriptor questions and analyticaltechniques based on those trials The results of the ab initio experiments werestill highly significant statistically and the sacrifice of some of theimpressionistic yield of the earlier efforts was deemed a reasonable price topay for the capacity for more incisive quantitative measurement of theinformation content of the data Notwithstanding the diminished effect sizeprompted a new phase of investigation with the goal of achieving a betterunderstanding of the cause of this attenuation and recovering the stronger yieldsobtained in the original experiments

In the course of generating the ab initio data several participants hadcomplained that the forced binary responses seemed somewhat inhibitory andincapable of capturing many aspects of their experiences suggesting that thismight have contributed to the deterioration of the results It was clearly evidentthat many of the target scenes and most of the perceptions contained ambiguousfeatures that could not be answered easily with simple lsquolsquoyesrsquorsquo or lsquolsquonorsquorsquo responsesFor example an agent might be indoors but looking out a window at an outdoorscene and thus unsure whether to characterize the scene as indoors or outdoorsA feature might have captured the agentrsquos attention during the target visitationbut not have been an integral component of the scene itself such as a briefconversational exchange with a passerby in an otherwise unpopulated areacomplicating the response to the question lsquolsquoAre people presentrsquorsquo This problemwas particularly evident in percipientsrsquo efforts to identify specific details froma perception that often emerged as a less than coherent stream of consciousnessmuch as in the difficulty of recalling features from fragments of dream imagery

In an effort to make the analytical judging process more lsquolsquouser friendlyrsquorsquoa quaternary descriptor response alternative was devised playfully termed

Fig 3 336 binary-encoded formal trial scores as a function of time


FIDO an acronym for lsquolsquoFeature Importance Discrimination Optionrsquorsquo This newformat provided participants with four response options for each descriptora rating of lsquolsquo4rsquorsquo identified a feature as a clearly dominant component of thescene lsquolsquo3rsquorsquo meant the feature was present but not particularly important lsquolsquo2rsquorsquoindicated uncertainty as to the presence or absence of the feature and lsquolsquo1rsquorsquo wasa statement of the definite absence of the feature Since implementation of theFIDO program required rewording of the descriptors combination of the FIDOtrials with the earlier databases was not feasible but it did provide anopportunity to clarify or redefine some of the existing questions that had posedoccasional interpretational difficulties After an extensive assessment whichincluded having several people encode a variety of test scenes with the newquaternary descriptors and comparing their responses for consistency a revisedset of 32 descriptors was created and a new body of experiments undertaken Inall other respects the same protocol was followed as in the earlier studiesalthough data were now generated on a trial-by-trial basis rather than in seriesof arbitrary length The FIDO program ran for four years beginning in 1985 andproduced a total of 167 trials

The standard FIDO scoring matrix illustrated below assigned a score of 5 toeach correctly matched response to options lsquolsquoabsentrsquorsquo and lsquolsquodominantrsquorsquo wherethere was agreement on the clear presence or absence of a given feature A scoreof 4 was assigned to correct matches of lsquolsquopresentrsquorsquo or lsquolsquounsurersquorsquo Mismatches oflsquolsquoabsentrsquorsquo vs lsquolsquounsurersquorsquo or lsquolsquopresentrsquorsquo vs lsquolsquodominantrsquorsquo where percipient and agentagreed on the presence or absence of a feature but assigned it different degreesof importance received a score of 3 if the percipient was less confident than theagent but only 2 if the percipient was more confident An lsquolsquounsurersquorsquo vslsquolsquopresentrsquorsquo mismatch received a score of 2 mismatches of lsquolsquoabsentrsquorsquo vslsquolsquopresentrsquorsquo lsquolsquoor unsurersquorsquo vs lsquolsquodominantrsquorsquo were assigned a score of 1 and a totalmismatch of lsquolsquodominantrsquorsquo vs lsquolsquoabsentrsquorsquo was scored as 0

The scores derived from the 32 descriptor comparisons were added to producea total score for each individual trial as in the previous binary analyses Amatrix was then constructed that scored all the targets against all the perceptionsand the scores of the correct matches compared with the distribution ofmismatched scores Rather than attempting to establish a priori probabilities forthese more complex descriptor options the FIDO calculations were carried outusing a method similar to binary Method A which simply divided the sum of

Absent Unsure Present Dominant

Absent 5 3 1 0 Unsure 2 4 2 1

TargetPresent 1 2 4 2

Dominant 0 1 3 5 Aacute Perception


the descriptor scores by the total number of descriptors ignoring any a prioridescriptor probabilities The composite z-score thus calculated for the 167 FIDOtrials was 1735 indicating a marginally significant overall achievement but onethat was reduced even further from the high yield of the previous data

Five alternative algorithms subsequently were applied ex post facto to theseFIDO data in an effort to understand the cause of the lower yield and to devisemore effective scoring strategies Two of these methods simply returned the datato the original binary and ternary formats to ascertain whether the lower yieldwas attributable to an analytical insensitivity of the new technique or to poorerpercipient performance The binary reduction treated all responses of 4 or 3 asa lsquolsquoyesrsquorsquo and all 2 or 1 responses as a lsquolsquonorsquorsquo while the ternary reduction treateda response of 4 as a lsquolsquoyesrsquorsquo a response of 1 as a lsquolsquonorsquorsquo and a response of 2 or 3 asan lsquolsquounsurersquorsquo A fourth method ignored everything but exact matches assigninga score of 1 for each descriptor response in the perception that matched that inthe target Two additional methods allowed partial credit for close matchessimilar to that of the standard FIDO algorithm One assigned a score of 2 for anexact match and a score of 1 for an ambiguous match the other assigneda weight of 4 to an exact match and a score of only 1 for an ambiguous match Asummary of the results produced by these six methods is presented in Table 4

Other than the binary-reduction version which produced nearly as manyextra-chance lsquolsquomissesrsquorsquo as lsquolsquohitsrsquorsquo the results from the other five methods alldisplayed relatively close concurrence marginally significant compositez-scores and effect sizes only about half that of the ab initio trials and onlyabout a fifth as large as that of the ex post facto subset Although the proportionsof trials with positive scores were above 50 in all the calculations neitherthese nor the numbers of significant trials exceeded chance expectation ClearlyFIDO had not achieved its goal of enhancing the PRP yield despite its potentialsensitivity to subtle or ambiguous informational nuances in the data Despitesome variability among the z-scores calculated for individual trials by thedifferent scoring methods the general consistency across most of the scoringmethods for the composite database suggested that the decreased yield was notdirectly due to inadequacies in the FIDO scoring algorithms per se but to a moregeneric suppression of the anomalous information channel

This suspicion was reinforced by a supplemental exercise in which anindependent human judge was asked to rank the fits between the agentsrsquo free-response transcripts and their coded descriptors This ranking effort wasadmittedly subjective and arbitrary and complicated by the varied lengths oftranscripts and the presence or absence of drawings photos or other illustrativematerial However of the 167 targets the judge determined that 162 (97)showed reasonably good correspondences between the agentsrsquo verbal descrip-tions and their descriptor responses A similar exercise was performed on thepercipientsrsquo encodings of their transcripts with comparable results Thus theFIDO descriptors themselves seemed adequate for capturing both the target


information and the percipientsrsquo imagery The diminishment of the yieldevidently had its source elsewhere

VIII Distributive Scoring

Shortly after completion of the FIDO analyses an REG-based humanmachine study had indicated that operator pairs of opposite sex workingtogether with a shared intention produced substantially stronger effects thansame-sex pairs or individual operators(35) This in turn had led toa comprehensive examination of nine of PEARrsquos humanmachine databaseswhich were found to display significant gender-related differences in individualoperator achievement(36) Although hints of possible gender-related trends hadalso been noted in the PRP data the previous pool of contributing percipientsand agents had been too small and disproportionately balanced to determinewhether such gender-pairing might be a significant factor in these experimentsas well To explore this hypothesis a new body of remote perceptionexperiments was performed using a balanced pool of same- and opposite-sexparticipant pairs each contributing an equal number of trials

This new protocol required each percipientagent pair to generate a seriesconsisting of five trials Ideally the same pair would produce another five-trialseries with their roles reversed Since a concern had been raised that providingfeedback to participants at the conclusion of each trial could introduce a possiblebias in subsequent trials feedback to participants was withheld until all fivetrials of a series were completed and each target selected from the pool ininstructed experiments was returned before the next trial To preclude anypossibility of shared response bias all analyses were based solely on local subsetcomparisons within a given series

As an added attempt to improve the scoring methodology a new descriptorcheck sheet was designed that permitted participants to respond to each questionon a distributive scale of 0 to 9 to indicate the relative prominence of each of 30descriptor features Similar to the prior methods the results were evaluated by

TABLE 4Summary of FIDO Data by Six Scoring Methods (N 5 167)

Scoringmethod

Effectsize

Compositez-score Probability

Trialsp 05

Trialsp 05

Trialsp 50

FIDO 01343 1735 041 10 (8) 6 (5) 54Binary 00761 0984 163 13 (12) 8 (7) 53Ternary 01598 2065 019 5 (6) 3 (4) 56Exact 01495 1932 027 17 (6) 10 (4) 54Distributive 01453 1878 030 12 (6) 7 (4) 57Weighted distributive 01467 1896 029 15 (6) 9 (4) 55

Numbers in parentheses indicate number of trials with negative z-scores p 05


constructing a 5 3 5 matrix for each series by scoring every target against everyperception These individual scores in turn were drawn from various 10 3 10matrices that cross-indexed and assigned values to every possible pair of 0ndash9descriptor rankings Again several different recipes were applied

deg A direct-match matrix that awarded a score of 1 for any exact descriptormatch and 0 for any mismatch

deg A binary matrix that treated any response of 0ndash4 as a lsquolsquonorsquorsquo and anyresponse of 5ndash9 as a lsquolsquoyesrsquorsquo with a correct match assigned a score of 1 andan incorrect match a score of 0

deg A ternary matrix that treated 0ndash2 as a lsquolsquonorsquorsquo 3ndash6 as an lsquolsquounsurersquorsquo and 7ndash9as a lsquolsquoyesrsquorsquo and assigned a score of 2 to any correct lsquolsquoyesrsquorsquo or lsquolsquonorsquorsquo match1 to a correct lsquolsquounsurersquorsquo match and 0 to any other response

deg A distributive matrix that assigned a score of 2 for a direct match 1 fora mismatch by one or two levels in the descriptor rankings and 0 for anyother mismatches

deg An extended distributive matrix that assigned a score of 10 to a directmatch 5 to an adjacent match 2 to a response two points removed fromthe correct rank 1 to a response three points removed and 0 to any otherresponse

deg A weighted distributive matrix that assigned scores of 9 for direct matchesat the extremes of the range (0 or 9) with decreasing credit as the matchapproached the middle of the range ie correct matches of 1 or 8received a score of 8 matches of 2 or 7 received a 7 etc Scoring foradjacent matches followed a similar pattern of reduced credit as the rankapproached the middle of the range

As before the sum of the individual descriptor scores constituted the total scorefor a given trial and the scores of the five matched trials were compared withthose of the 20 mismatched scores to determine the statistical merit of eachseries

Thirty experimental series comprising 150 trials were generated using thisdistributive protocol by 12 participant pairs 8 of whom produced at least twoseries together with the percipientagent roles reversed The results aresummarized in Table 5

Once again there was reasonably good agreement among the six scoringrecipes but the overall results were now completely indistinguishable fromchance No more than the expected number of significant trials emerged in theanalyses and the low statistical resolution in defining the local empiricalchance backgrounds a consequence of the small size of the scoring matricesmade calculation of individual trial z-scores virtually meaningless In a certainsense this was reminiscent of one of the problems that had stimulateddevelopment of the analytical judging methodologies 18 years earlier namelythe statistical inefficiency of assessing the informational content of individualtrials in small experimental series But now the phenomenon itself seemed to


have disappeared And given the lack of any statistical yield in these data itwas not possible to ascertain whether there was any evidence of co-operatoror gender differences the question that had originally prompted thisexploration

In pondering this paradox we became cognizant of a number of subtler lessquantifiable factors that also might have had an inhibitory effect on theexperiments such as the laboratory ambience in which the experiments werebeing conducted For example during the period in which the FIDO data werebeing generated we were distracted by the need to invest a major effort inpreparing a systematic refutation to an article critical of PEARrsquos earlier PRPprogram(3738) Although most of the issues raised in that article were irrelevantincorrect or already had been dealt with comprehensivelyelsewhere and shown tobe inadequate to account for the observed effects(23) this enterprise deflecteda disproportionateamountof attentionfrom and dampened the enthusiasmfor theexperiments being carried out during that time Beyond this in order to forestallfurther such speciouschallenges it led to the impositionof additionalunnecessaryconstraints in the design of the subsequent distributiveprotocol Although it is notpossible to quantify the influence of such intangible factors in the study ofconsciousness-related anomalies where unknown psychological factors appear tobe at the heart of the phenomena under study they cannot be dismissed casually

IX Review and Discussion

The evidence acquired in the early remote perception trials had raisedprofound questions in the minds of the PEAR researchers similar no doubt tothose of the countless others who over the course of history had experiencedfirst-hand the validity of Paracelsusrsquo remarkable claim The possibility thatordinary individuals can acquire information about distant events by theseinexplicable means even before they take place challenges some of the mostfundamental premises of the prevailing scientific worldview PEARrsquos effortsto devise strategies capable of representing the information acquired in the

TABLE 5Summary of Distributive Data by Six Scoring Methods (30 Series 150 Trials)

Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50

Direct match 2 00088 2 0108 543 2 (0) 6 (6) 4 (4) 46Binary 2 00684 2 0838 799 0 (1) 8 (3) 5 (2) 47Ternary 2 00342 2 0419 662 0 (0) 5 (5) 3 (3) 55Distributive 2 00501 2 0613 730 1 (0) 5 (5) 3 (3) 51Extended

distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52

Weighteddistributive

2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



remote perception process in a manner amenable to quantitative analysis hadfollowed the traditional scientific method ie to design experiments capableof reproducing the phenomenon under carefully controlled conditions tosystematically eliminate sources of extraneous noise in order to bring thephenomenon in question into sharper focus and to pose theoretical models todialogue with these empirical results

The early phases of the program provided encouraging indications that thiscould be accomplished via a set of standardized descriptor queries addressed toboth the agentrsquos description of the physical target and to the percipientrsquos stream-of-consciousness narrative that would serve as an lsquolsquoinformation netrsquorsquo to capturethe essence of the anomalous communication Ex post facto application of thistechnique to existing data seemed to confirm the efficacy of this approachproducing results that were consistent with previous human judge assessmentsand encouraging continued explorations In the second phase of the program abinitio utilization of this method in a new body of experiments also producedhighly significant results While the average effect size of these was somewhatsmaller than that of the original ex post facto subset this was attributed primarilyto the fact that these were the data on which the descriptor questions andanalytical techniques had been based Nevertheless the statistical yield of the abinitio data still was sufficiently robust to indicate that the new method couldserve its intended purpose adequately

Yet like so much of the research in consciousness-related anomaliesreplication enhancement and interpretation of these results proved elusive Asthe program advanced and the analytical techniques became more sophisticatedthe empirical results became weaker It appeared as if each subsequentrefinement of the analytical process intended to improve the quality andreliability of the lsquolsquoinformation netrsquorsquo had resulted in a reduction of the amount ofraw information being captured This diminution of the experimental yieldprompted extensive examination of numerous factors that could have contributedto it After exploring and precluding various possible sources of statistical orprocedural artifact however we were forced to conclude that the cause of theproblem most likely lay somewhere in the subjective sphere of the experience

Throughout the course of the program when participants had been queriedabout their personal reactions to the encoding process their most commoncomplaint was a feeling of being lsquolsquoconstrainedrsquorsquo by the required forced-choicebinary queries In response the FIDO phase was implemented to permitparticipants more freedom in formulating their responses Although the FIDOdatabase appeared to contain a considerable number of impressionisticallysuccessful trials the composite quantitative results now were only marginallysignificant

The failure of FIDO to reinvigorate the PRP program plus the desire toexamine variations in individual performance led to yet another encodingstrategy with even more response flexibility ie the distributive methodologyAlthough this method was intended to alleviateparticipantsrsquo feelings of subjective


constraint concerns about the possibility of participant response biases imposedadditional procedural restrictions It was evident from the null results of the 150distributive trials that all efforts to enhance the effect by progressively moreelaborate analysis techniques not only had failed but even had proven counter-productive Although the judging methodology had been proven to serve itsintended analytical purpose the progressive attenuation of the yield suggestedthat there was some kind of interference taking place between the analyticalmeasures and the generation of the effects they were attempting to measure

The trend is clearly evident on re-examination of the cumulative deviationgraph of Figure 1 which plots chronologically the cumulative results of all 336formal binary-encoded trials and displays a potentially instructive clue to theinexorable decrease in effect size Following the initial sharp slope representingthe strong yield of the original 59 ex post facto trials the slope of the subsequent277 ab initio trials can be seen to consist of two distinct segments The first ofthese comprising the initial 168 ab initio trials (60 through 227 on the x-axis)has a consistent positive slope albeit shallower than that of the earlier ex postfacto data The slope of the second segment (trials 228 through 336) whichconsists of the 109 trials from the second phase of the ab initio experiments isnoticeably flatter The beginning of this second segment would therefore appearto be the point at which the experimental yield began to deteriorate Figure 4plots the comparative effect sizes of the data from these various experimentalperiods reconfirming the systematic decrease of the yield beginning with thesecond phase of the ab initio binary experiments The numerical results of thesesegments are presented in Table 6 (Again the effect sizes displayed in the graphand table were calculated by dividing the z-scores for each database by thesquare root of the number of trials in that subset and thus indicate the averagez-score per trial)

While the composite yield of the total database remains highly significant itis evident that this result is driven primarily by the much stronger yields of theearlier trials bolstered by the substantial size of the overall database itself Thesuccess of the analytical judging technique in the early phases of the programand its apparent insensitivity to the particular scoring matrices invokedconfirms that such an approach can indeed be deployed successfully as a strategyfor quantifying this inherently subjective process Nonetheless somethingclearly changed in the second phase of the ab initio experiments that resulted ina substantial weakening of the effect being quantified Since both phases of theab initio portion of the program utilized identical descriptor questions andscoring algorithms their analytical effectiveness therefore can be ruled out asthe source of the lower yield in the later phases of the program

Another pattern became evident when we returned to the raw free-response datawith this in mind The free-response descriptions in the later trials wereconsiderably shorter than those generated in the earlier ones some of which hadrun to several pages of narrated perceptions Indeed in many of these later trialspercipientsrsquo verbal descriptions consisted of only a few cursory phrases intended


simply to clarify nuances of their descriptor responses and provided little in theway of the stream-of-consciousness imagery they had been asked to generate Itappeared that as the percipients became more familiar with the descriptorquestions their subjective impressions were increasingly guided and circum-scribed by them as though the questions were establishing the informationalframework for their responses The original free-response remote perceptionexperiment thus had taken on the characteristics of a multiple-choice task and thelocus of the experience had shifted from the realm of intuition to that of intellect

X From Analysis to Analogy

Having exhausted the search for the source of the remote perception signaldeterioration in the analytical techniques themselves we are driven to look

Fig 4 Effect sizes of various data subsets

TABLE 6PRP Summaries by Database

Participants

Database

Trials

Series

Agents

Percipients TotalComposite

z-scoreEffectsize Probability

Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6

Later trials 109 13 7 13 15 1291 124 098FIDO 167 9 19 22 25 1735 134 041Distributive 150 30 15 15 16 2 0108 2 009 543

TOTAL 653 88 39 59 69 5418 212 3 3 102 8

Some individuals contributed to more than one database in both percipient and agent capacity


further afield for a satisfactory explanation If we step back to review theprogram from a broader perspective we note that all of the methodologicallsquolsquoimprovementsrsquorsquo introduced to refine the scoring techniques had been directedtoward more efficient extraction of the anomalous information and eliminationof possible sources of artifact or bias Some were efforts to achieve lsquolsquosharperdefinitionrsquorsquo of the remote perception lsquolsquosignalrsquorsquo others were attempts to lsquolsquotightenrsquorsquothe experimental lsquolsquocontrolsrsquorsquo and a few were designed to lsquolsquoclarifyrsquorsquo certaincharacteristics of the communication lsquolsquochannelrsquorsquo All these terms reflect anemphasis on achieving increasingly precise specification and reducing the noiseor uncertainty in the process Yet each increment of analytical refinementappears to have resulted in a systematic reduction not of the lsquolsquonoisersquorsquo but of thelsquolsquosignalrsquorsquo itself This raises the somewhat radical possibility that manifestation ofthe anomaly may actually require a certain degree of the very noise oruncertainty that we had invested so much effort to reduce It is a possibilityhowever for which precedent can be found in other domains of scholarlyinquiry and is therefore worth consideration in the present context

The most immediate technical examples of this complementarity of signaland noise are the humanmachine experiments carried out in our laboratory andelsewhere(39) All of these studies employ some form of random processorand the anomalous effects appear as departures of their random outputs fromchance expectation It is as if the lsquolsquonoisersquorsquo of the random process provides theessential raw material out of which the mind of the operator is able to constructa small amount of ordered lsquolsquosignalrsquorsquo

Such effects are by no means restricted to explicit anomalies research Similardepartures from canonical expectations can be found in contemporaryengineering applications of lsquolsquostochastic resonancersquorsquo wherein a deliberateincrease in the overall level of noise in certain kinds of lasers or sensitiveelectronic circuits can actually enhance the detection of weak fluctuatingsignals(4041) Other studies have demonstrated that the introduction of anelement of chaos into certain types of nonlinear processes such as the interactionof two otherwise independent random oscillators can stimulate synchronousbehavior between the transmitter and the receiver(4243) In each of theseinstances information or order has been introduced into a sensitive nonlinearphysical system not by reducing the ambient noise but by increasing it

Of particular interest for our purpose is the researchersrsquo unanticipatedobservation that in such synchronizationprocesses the receiver actually recordedchanges in the signal before the transmitter recorded the transmission of thosechanges In other words the system seemed capable of anticipating thesynchronization The engineers who carried out the studies remarked that lsquolsquoWewould thus expect that any of those analogous systems which exhibit chaosshould also be liable to anticipating synchronization We thus hope that ourwork will act as a stimulus to explore the opportunities for observinganticipating synchronization in physical chemical biological and socio-economic systemsrsquorsquo(41) Following this suggestion we might note that in


a certain sense the remote perception process qualifies as an example ofa lsquolsquosensitive nonlinear system with a weak fluctuating signalrsquorsquo that exhibitsa certain degree of chaos and that the participants in these experiments functionas lsquolsquotwo otherwise independent random oscillatorsrsquorsquo Hence it well may be thatour signal is also dependent upon a background of random noise for itsmanifestation If so it would appear that it was our attempts to enhance the remoteperception signal by sharpening the specificity of the information channel thatcould in fact have been responsible for the attenuation of the signal

Reaching farther afield for relevant analogies the accepted model ofbiological evolution incorporates the importance of uncertainty in enhancinginformation Darwinian theory postulates that living species adapt to theirenvironment by selecting for specific traits that emerge in the process of randomgenetic mutation This process is itself strongly dependent on the generation oflsquolsquonoisersquorsquo emerging from the massive redundancy of continuously recombinedgenetic information When the randomness of this process is limited as inrepeated interbreeding the short-term advantage of increased predictability ofinherited traits is offset by longer-term weakening of the genetic strain of thespecies

Insights can also be derived from a quite different realm of human experiencenamely the practice of certain mystical divinatory traditions where anomalousrelationships between signal and noise are also evident In most of thesea clearly defined question is submitted to some kind of random process for thepurpose of accessing information unavailable to the conscious mind Typicallythe response comes in imprecise or symbolic form that requires translation intomeaningful or pertinent terms One such example is the renowned Oracle ofApollo at Delphi in ancient Greece a highly respected source of wisdom thatlong played a central role in Greek culture and politics Consultation of theoracle involved a priestess called the Pythia who crowned in laurel and in analtered state of consciousness stimulated by vapors arising from a cleft in theearth over which she sat on a tripod produced a lsquolsquofree responsersquorsquo utterancewhich was then interpreted by the attending priest in response to the seekerrsquosquery Two points of potential relevance here are the non-analytical receptivestate of mind of the lsquolsquopercipientrsquorsquo and the deferment of interpretation by thelsquolsquojudgersquorsquo until after the experience has been completed

Another ancient oracle still widely used is the Chinese lsquolsquoBook of Changesrsquorsquoor I Ching a divination process that involves generation of a sequence ofrandom binary events the results of which are represented as two lsquolsquotrigramsrsquorsquoThese are referred to a table or matrix that identifies each of the 64 possiblecombinations or lsquolsquohexagramsrsquorsquo with a specific text that is then consulted toobtain a response to the original query Notwithstanding the subjective nature ofthe interpretation of the texts a vast body of evidence accumulated over manymillennia testifies to the efficacy of the I Ching in producing accurate andconsequential results Despite the claim of many rationalists that such oraclesare nothing more than bizarre combinations of wishful thinking and lsquolsquomere


chancersquorsquo this is the same lsquolsquoirrationalrsquorsquo formula that seems to underlie the remoteperception phenomena that have now been demonstrated by rigorous analyticalquantification to convey more meaningful information than can be attributed tolsquolsquomere chancersquorsquo Hence the principles invoked by the ancient sages in developingthe I Ching may shed some light on these more contemporary anomalies

Psychologist Carl Jung who devoted more than 30 years to the study of the IChing pointed out in his Foreword to the classic Richard Wilhelm translation(44)

that lsquolsquowe know now that what we term natural laws are merely statistical truthsand thus must necessarily allow for exceptions If we leave things to naturewe see a very different picture every process is partially or totally interferedwith by chance so much so that under natural circumstances a course of eventsabsolutely conforming to specific laws is almost an exceptionrsquorsquo He relates theemphasis placed by the ancient Chinese mind on chance and the subjectiveinterpretation of events to the modern world of quantum mechanics where thereality of inherently random microscopic physical events includes the observeras well as the observed In both domains what Jung refers to as the lsquolsquohiddenindividual quality of things and menrsquorsquo draws on the unconscious and intangiblequalities that undergird the experiences of the conscious mind and the tangiblephysical world respectively in similar fashion to the conceptual frameworkdescribed in our paper lsquolsquoA Modular Model of MindMatter Manifestation(M5)rsquorsquo(45) Both Jungrsquos representation and our own emphasize that the causal andsynchronistic perspectives of reality are complementary rather than mutuallyexclusive Jung maintains that the lsquolsquocoincidencersquorsquo of a synchronistic event occurslsquolsquobecause the physical events are of the same quality as the psychic events andbecause all are the exponents of one and the same momentary situationrsquorsquo(44) Ourrepresentation of this concept speaks of the emergence of both cognitiveexperience and physical events from a common underlying substrate of theunconscious mind and the undifferentiated world of physical potentialitywherein the distinction between mind and matter blurs into uncertainty Giventheir common origin it should not be surprising to observe correlations betweentheir manifested expressions in the worlds of mental and physical lsquolsquorealityrsquorsquo Justas the concept of complementarity in quantum mechanics brings with it a certaindegree of uncertainty that makes it impossible to achieve absolute precision intwo frames of reference simultaneously the complementarity of an lsquolsquoobjectiversquorsquocausal picture of reality and a lsquolsquosubjectiversquorsquo synchronistic one also maynecessitate tolerance of a degree of uncertainty in both dimensions

In many respects the empirical evidence from remote perception as well asfrom other domains of anomalies research is more compatible with an acausalor synchronistic model than with a causal one Although we have recognizedthis in principle our experimental approach and the language we have deployedin describing the effects has betrayed certain causal assumptions For exampledespite repeated comments from participants that the PRP experience felt morelike lsquolsquosharingrsquorsquo than lsquolsquosending and receivingrsquorsquo we persisted in speaking ofinformation lsquolsquotransmissionrsquorsquo Similarly our enduring efforts to extract the


lsquolsquosignalrsquorsquo from the lsquolsquonoisersquorsquo also reflected a more deterministic orientation YetJungrsquos model the ancient divinatory traditions evolutionary theory contempo-rary signal processing research and humanmachine anomalies all suggest thatnoise may be a requisite component of the process of signal generation and thatobjective linear causality may not prevail under these circumstances

If one defines lsquolsquonoisersquorsquo in the remote perception context as the percipientrsquosuncertainty or lack of conscious knowledge about the target and lsquolsquosignalrsquorsquo asthe content of valid information acquired in the process these diverse analogiescan be quite instructive For example the early experiments wherein percipientswere asked simply to generate an unfocused free-response stream ofconsciousness were in this sense more lsquolsquonoisyrsquorsquo than the later efforts wherepercipientsrsquo imagery was guided by a more structured information lsquolsquogridrsquorsquo orlsquolsquofilterrsquorsquo of descriptor queries In those trials that were only encoded ex postfacto the participants had no knowledge of the information filter that would beimposed only well after the data were generated and they seemed more easilyable to access information about the targets In the first generation of ab initiobinary-encoded trials when descriptor check-sheets were something of a noveltyand percipients were still urged to generate their free-response descriptionsbefore attempting descriptor encoding the transcripts tended to be somewhatshorter but most of them still comprised a free-association type of narrativeThese trials also produced highly successful results albeit of a somewhatsmaller average effect size By the time of the later ab initio experimentshowever when we had acquired greater confidence in the efficacy of theanalytical judging approach less importance was placed on the raw free-response data and this shift of emphasis was reflected in the abbreviated evencursory percipient responses In retrospect it is apparent from the content ofthese shorter transcripts that the percipients were anticipating the descriptorquestions and inadvertently focusing their attention on those particular aspectsof their experience Although the intent of the quaternary and then distributivedescriptor questions was to relieve the participantsrsquo sense of lsquolsquoconstraintrsquorsquo thesemore complex forms of questions appear to have had the opposite effect forcingpercipients to pay even more attention to the nuances of the information grid andthus filtering out any signal that was not perceived to be lsquolsquorelevantrsquorsquo In this waythe background lsquolsquonoisersquorsquo was reduced even further and more structuredcognitive processes associated with achieving internal consistency in whathad essentially become a forced-choice task effectively restricted the flow ofunconscious imagery

It is also telling that until recently this trend had not even been perceived asa problem by the researchers Typing 30 numbers into a computer was mucheasier than the task of evaluating lengthy verbal transcripts and the ability toacquire a quantitative indication of the merit of an individual trial increasinglyreplaced the spontaneous excitement of finding apparent correspondences in theraw data The shift in experimental perspective from predominantly subjectiveto almost totally analytical was so gradual that little consideration was given to


the possible costs of such a transition For example combination of the datafrom the first and second phases of the ab initio experiments was justified solelyon technical grounds with no serious consideration given to the implications ofa change from ranking the quality of a trial to measuring its specific informationcontent other than the relative efficiency and statistical power of the twoapproaches The subsequent effort expended on refining the technical andanalytical components of the program rather than on trying to understand whatthe participants were really trying to tell us when they complained of feelinglsquolsquoconstrainedrsquorsquo by the descriptor questions further exacerbated the overemphasison quantitative precision that ultimately may have suffocated the subtle butessential subjective signal

The larger effect size of the lsquolsquoinstructedrsquorsquo vs the lsquolsquovolitionalrsquorsquo trials alsosupports the importance of retaining an adequate component of noise oruncertainty in the system When percipients attempted to describe scenes chosenby a random process that precluded utilization of any prior knowledge about theagentrsquos habits or personal preferences their perceptions contained a largercomponent of anomalous information In the volitional protocol where onemight imagine a certain a priori advantage percipientsrsquo rational expectationsmay have imposed yet another kind of information filter that inhibited the subtlelsquolsquosignal detectionrsquorsquo process In other words the strongest lsquolsquosignalsrsquorsquo appear tohave been generated under the lsquolsquonoisiestrsquorsquo conditions ie in the absence orminimization of any orderly or rational form of structural information (It maybe interesting to note in this regard that approximately 66 of the ab initiobinary trials 98 of the FIDO trials and 77 of the distributive trials followedthe volitional protocol whereas 53 of the ex post facto trials were instructed)

One might even speculate that the overall success of these experimentsderives in considerable measure from the lsquolsquoirrationalrsquorsquo nature of the remoteperception task itself When requested to describe a spatially and temporallyremote scene without access to any known sensory channel percipients areforced to abandon any rational strategy for fulfilling such an assignment Withcognitive functioning thus confounded by uncertainty leaving the consciousmind less able to mask the subtle signal with rational associations theunconscious mind of the percipient may better be able to access the lsquolsquohiddenindividual quality of things and menrsquorsquo

Although a degree of uncertainty may indeed be necessary for the generationof remote perception effects the complementary relationship between signal andnoise we are proposing nevertheless requires retention of a comparabledimension of structure in the process Recall for example that the earlyexploratory trials where percipients did not know the identity of the agent or thetime of target visitation produced completely null results (Table 3) As in the IChing or other divinatory arts where it is essential that the querant pose a clearlydefined question the remote perception process also seems to require thepercipient to establish some minimal lsquolsquoboundary conditionsrsquorsquo when addressingthe unknown target If indeed such a process involves an excursion into the


unconscious realm of undifferentiated potential in order to acquire specificinformation some corresponding specific question would appear to bea prerequisite To complement this facilitative function some form ofquantitative assessment of the amount of anomalous information is indispens-able if the study of remote perception is to qualify as a scientific enterprise

To this end we have proposed in several previous publications that a moreastute balance between the analytical and the aesthetic dimensions of suchphenomena needs to guide any future explorations of consciousness-relatedanomalies(3245ndash49) In the article entitled lsquolsquoScience of the Subjectiversquorsquo(49) weobserved how lsquolsquoin the interplay of objective intellect and subjective spirit we aredealing with the primordial conjugate perspectives whereby consciousnesstriangulates its experiencersquorsquo This complementary relationship has now beenconfirmed in the record of our remote perception research That is thesubjective spirit of these experiences appear to be more effectively attainedwhen unencumbered by analytical or cognitive overlays and its inherentuncertainties are both acknowledged and utilized However the equallyimportant role of objective intellect must serve to enhance rather than toinhibit the process and our eventual understanding of it

Acknowledgments

The authors are indebted to a great many people without whom the pro-gram described in this paper could not have been accomplished In particularwe express our sincerest thanks to the 72 participants who gave so generouslyof their time to produce these data and to our many friends and colleagueswho assisted in various stages of judging encoding and re-encoding themSpecial thanks are extended to our PEAR colleagues Roger Nelson and YorkDobyns who were instrumental in virtually every phase of the PRP programincluding the development of protocols descriptor questions and analysesand the interpretations of the data and to Elissa Hoeger for her invaluableassistance in preparing the many tables and references contained herein

This research has been an integral part of the PEAR program since 1979during which time it has been supported by the generosity of many philan-thropic individuals and organizations including Mr Laurance RockefellerMr Richard Adams the Institut fuumlr Grenzgebiete der Psychologie und Psy-chohygiene and other donors who prefer to remain anonymous

Appendix A

Local Descriptor Probabilities and Individual Performance

The scores presented in the summaries of Table 2 had been calculated usingthe local a priori probabilities associated with each subset following the sameprocedure that had been deployed for all of the major analyses in the first phase


of the analytical judging program[25(AppxC)] Those early explorations hadestablished that when the local a priori probabilities were used to scorea particular subset using a given scoring method the empirical chancedistributions resulting for different subsets appeared to be statisticallyindistinguishable It thus had been concluded that a single empirical chancedistribution namely the one resulting from the largest assembly of formal datacould be used as a reliable reference standard for any subset provided that thesubsetrsquos trial scores were computed using its own local a priori probabilities

Unfortunately this uniformity of chance distributions is only approximatelycorrect A re-evaluation of this technique illustrated a mechanism wherebyinternal variations in the a priori probabilities among different subsets of thedatabase could potentially produce artificially inflated or deflated scores in thematched-trial distributions relative to the off-diagonal population of mis-matches For example a given percipientagent pair might happen to sharea similar encoding style such as a tendency to respond affirmatively toambiguous features or particular preferences for certain descriptors whichcould result in their trials having responses that were more closely correlatedthan those of the mismatched scores constituting the reference distributionSimilar biases also might arise from geographical or seasonal variations or otherpossible causes

Since the apparent indistinguishability of the chance distribution for a numberof large data subsets cannot be guaranteed theoretically it is necessary to verifyempirically that the overall results are not in fact spuriously inflated by suchbiasing mechanisms The possible influence of idiosyncratic individual patternsof a priori response probabilities in agent and percipient encoding styles wasexamined using the data produced by the 29 agentpercipient pairs who hadcontributed five or more trials to the composite database (Collectively these 29pairs were responsible for 274 of the 336 formal trials) The results of this testfor local biasing are shown in Figure A which displays an array of traces forthese 274 trials after the style of Figure 1 The individual plotted points are thecumulative z-scores achieved by each of the 29 agentpercipient pairs based onthree distinct calculation methods The lsquolsquonon-localrsquorsquo method calculates each trialscore using the a priori probabilities for the full formal database and computesits z-score against the standard empirical chance distribution for the overalldatabase In other words this trace is simply the composite z-score assigned tothe subset of trials contributed by given agentpercipient pairs extracted fromthe results of the overall database of 336 formal trials In comparison the lsquolsquolocalalpharsquorsquo score is derived by scoring each percipientagent pairrsquos contributions onthe basis of its own internal a priori probabilities but still referring these scoresto the overall empirical chance distribution The lsquolsquolocal distributionrsquorsquo calculationremoves all reference to global distributions and along with it any possibility oflocal-biasing effects by scoring each agentpercipient pairrsquos data not only withits own local a priori probabilities but against its own local mismatchdistribution


With few exceptions all of which are associated with very small datasets thethree scoring strategies produce a reassuring degree of agreement especially inthe composite yields It is evident from Figure A that these three methods are notstatistically distinguishable and that any inflation or deflation of the overalleffect due to local biasing is less than the inherent statistical uncertainty of thescoring procedure It therefore may be concluded that within the limits of thestatistical resolution encoding artifact is not a significant contributor to theseexperimental results

The rank-ordered effect sizes obtained by each of the 28 percipients and 15agents who contributed more than one trial to the database were also examinedSome 25 of the percipients 40 of the agents and 21 of the percipientagent pairs produced statistically significant overall results whereas only 5 ofeach group would be expected to do so by chance All but two percipients andtwo agents generated net positive effects compared to the 50 chanceexpectation and of these four individuals three produced positive results whenfunctioning in the alternate role A separate data subset consisting of only thefirst trials from each of the 38 percipients contributing to the formal databasewas also calculated to examine the possibility that the composite yield mighthave been distorted by large databases produced by any given percipientDespite the small size of this group of trials the results display the same linearconsistency as the full database achieving a highly significant composite z-scoreof 3890 Thus it is also clear that the success of the overall results is notattributable to exceptional performance by only a few participants25

References

1 Hartmann F (1973) Paracelsus Life and Prophecies Blauvelt NY Rudolf Steiner (p 105)2 Targ R amp Puthoff H E (1974) Information transmission under conditions of sensory

shielding Nature 252 602ndash607

Fig A Cumulative z-score progress for three alternative scoring techniques


3 Puthoff H E amp Targ R (1976) A perceptual channel for information transfer over kilometerdistances Historical perspective and recent research Proceedings of the IEEE 64 329ndash354

4 Allen S Green P Rucker K Cohen R Goolsby C amp Morris R L (1975) A remoteviewing study using modified version of the SRI procedure In Morris J D Roll W G ampMorris R L (Eds) Research in Parapsychology Metuchen NJ Scarecrow Press (pp 46ndash48)

5 Hastings A C amp Hurt D B (1976) A confirmatory remote viewing experiment in a groupsetting Proceedings of the IEEE 64 1544ndash1545

6 Rauscher E A Weissmann G Sarfatti J amp Sirag S P (1976) Remote perception of naturalscenes shielding against ordinary perception In Morris J D Roll W G amp Morris R L(Eds) Research in Parapsychology Metuchen NJ Scarecrow Press (pp 41ndash45)

7 Vallee J Hastings A C amp Askevold G (1976) Remote viewing experiments throughcomputer conferencing Proceedings of the IEEE 64 1551ndash1552

8 Whitson T W Bogart D N Palmer J amp Tart C T (1976) Preliminary experiments ingroup lsquolsquoremote viewingrsquorsquo Proceedings of the IEEE 64 1550ndash1551

9 Puthoff H E amp Targ R (1977) Mind Reach New York Delacorte10 Chotas H (1978) Remote viewing in the Durham area Journal of Parapsychology 42 61ndash6211 Tart C T Puthoff H E amp Targ R (Eds) (1979) Mind at Large IEEE Symposia on the

Nature of Extrasensory Perception New York Praeger Special Studies12 Karnes E W Sussman E P Klusman P amp Turcotte L (1980) Failures to replicate remote

viewing using psychic subjects Zetetic Scholar 6 66ndash7613 Schlitz M amp Gruber E (1980) Transcontinental remote viewing Journal of Parapsychology

44 305ndash31714 Puthoff H E Targ R amp May E C (1981) Experimental psi research Implications for

physics In Jahn R G (Ed) The Role of Consciousness in the Physical World Boulder COWestview Press (pp 37ndash86)

15 May E C Utts J M Humphrey B S Luke W L W Frivoid T J amp Trask V V (1990)Advances in remote-viewing analysis Journal of Parapsychology 54 193ndash228

16 Nelson R D Dunne B J Dobyns Y H amp Jahn R G (1996) Precognitive remoteperception Replication of remote viewing Journal of Scientic Exploration 10 109ndash110

17 May E C (1996) The American Institutes for Research review of the Department of DefensersquosSTAR GATE program A commentary Journal of Scienti c Exploration 10 89ndash108

18 Puthoff H E (1996) CIA-initiated remote viewing program at Stanford Research InstituteJournal of Scienti c Exploration 10 63ndash76

19 Targ R (1996) Remote viewing at Stanford Research Institute in the 1970s A memoir Journalof Scientic Exploration 10 77ndash88

20 Bremseth Commander L R (2001) Unconventional Human Intelligence Support Transcendentand Asymmetric Warfare Implications of Remote Viewing Graduate thesis Marine Corps WarCollege Marine Corps University Marine Corps Combat Development Command QuanticoVA

21 Puthoff H E (2001) CIA-initiated remote viewing at Stanford Research Institute TheIntelligencer Journal of US Intelligence Studies 12 60ndash67

22 Jahn R G Dunne B J amp Jahn E G (1980) Analytical judging procedure for remoteperception experiments Journal of Parapsychology 44 207ndash231

23 Jahn R G Dunne B J Nelson R D Jahn E G Curtis T A amp Cook I A (1982)Analytical judging procedure for remote perception experimentsmdashII Ternary coding andgeneralized descriptors Technical Note PEAR 82002 Princeton Engineering AnomaliesResearch School of EngineeringApplied Science Princeton University Princeton NJ

24 Dunne B J Jahn R G amp Nelson R D (1983) Precognitive remote perception TechnicalNote PEAR 83003 Princeton Engineering Anomalies Research School of EngineeringAppliedScience Princeton University Princeton NJ

25 Dunne B J Dobyns Y H amp Intner S M (1989) Precognitive remote perception IIIComplete binary data base with analytical refinements Technical Note PEAR 89002 PrincetonEngineering Anomalies Research School of EngineeringApplied Science Princeton UniversityPrinceton NJ

26 Dunne B J amp Jahn R G (2002) Information and uncertainty 25 years of remote perceptionresearch Technical Note PEAR 200201 Princeton Engineering Anomalies Research School ofEngineeringApplied Science Princeton University Princeton NJ

27 Dunne B J amp Bisaha J P (1979) Precognitive remote viewing in the Chicago area Areplication of the Stanford experiment Journal of Parapsychology 43 17ndash30


28 Bisaha J P amp Dunne B J (1979) Multiple subject and long-distance precognitive remoteviewing of geographical locations In Tart C T Puthoff H E amp Targ R (Eds) Mind atLarge IEEE Symposia on the Nature of Extrasensory Perception New York Praeger SpecialStudies (pp 109ndash124)

29 Solfvin G Kelly E amp Burdick D (1978) Some new methods for analysis for preferentialranking data Journal of the American Society for Psychical Research 72 93ndash109

30 Dunne B J (1979) Precognitive Remote Perception A Critical Overview of the ExperimentalProgram Masterrsquos thesis Committee on Human Development University of Chicago ChicagoIL

31 Andresen J (1984) Statistical Tests of Experimental Remote Perception Trials Undergraduatethesis Department of Civil Engineering School of EngineeringApplied Science PrincetonUniversity Princeton NJ

32 Jahn R G amp Dunne B J (1987) Margins of Reality The Role of Consciousness in thePhysical World (Section III) San Diego CA Harcourt Brace Jovanovich

33 Jahn R G Dunne B J amp Nelson R D (1987) Engineering anomalies research Journal ofScienti c Exploration 1 21ndash50

34 Dunne B J amp Jahn R G (1992) Experiments in remote humanmachine interaction Journalof Scientic Exploration 6 311ndash332

35 Dunne B J (1993) Co-operator experiments with an REG device In Rao K R (Ed)Cultivating Consciousness for Enhancing Human Potential Wellness and Healing WestportCT Praeger (pp 111ndash121) (Also available as Technical Note PEAR 91005 PrincetonEngineering Anomalies Research School of EngineeringApplied Science Princeton UniversityPrinceton NJ)

36 Dunne B J (1998) Gender differences in humanmachine anomalies Journal of Scienti cExploration 12 3ndash55

37 Hansen G P Utts J amp Markwick B (1992) Critique of the PEAR remote-viewingexperiments Journal of Parapsychology 56 97ndash113

38 Dobyns Y H Dunne B J Jahn R G amp Nelson R D (1992) Response to Hansen Utts andMarkwick Statistical and methodological problems of the PEAR remote viewing [sic]experiments Journal of Parapsychology 56 115ndash146

39 Jahn R G Dunne B J Nelson R D Dobyns Y H amp Bradish G J (1997) Correlations ofrandom binary sequences with pre-stated operator intention A review of a 12-year programJournal of Scienti c Exploration 11 345ndash367

40 McNamara B Wiesenfeld K amp Roy R (1988) Observation of stochastic resonance in a ringlaser Physical Review Letters 60 2626ndash2629

41 McNamara B amp Wiesenfeld K (1989) Theory of stochastic resonance Physical Review A 394854ndash4869

42 Jones R J Rees P Spencer P S amp Shore K A (2001) Chaos and synchronisation of self-pulsating laser diodes Journal of the Optical Society of America B 18 166ndash172

43 Sivaprakasam S Shahverdiev E M Spencer P S amp Shore K A (2001) Experimentaldemonstration of anticipating synchronization in chaotic semiconductor lasers with opticalfeedback Physical Review Letters 87 154101

44 Jung C G (1950) Foreword to The I Ching (R Wilhelm trans) Princeton NJ PrincetonUniversity Press pp xxxindashxxxix

45 Jahn R G amp Dunne B J (2001) A modular model of mindmatter manifestation (M5)Journal of Scienti c Exploration 15 299ndash329

46 Jahn R G (1989) Anomalies Analysis and aesthetics Journal of Scienti c Exploration 3 15ndash26

47 Jahn R G (1993) The complementarity of consciousness In Rao K R (Ed) CultivatingConsciousness for Enhancing Human Potential Wellness and Healing Westport CT Praeger(pp 149ndash163) (Also available as Technical Note PEAR 91006 Princeton EngineeringAnomalies Research School of EngineeringApplied Science Princeton University PrincetonNJ)

48 Dunne B J (1997) Subjectivity and intuition in the scientific method In Davis-Floyd R andArvidson S (Eds) Intuition The Inside Story New York Routledge (pp 121ndash128)

49 Jahn R G amp Dunne B J (1997) Science of the subjective Journal of Scientic Exploration11 201ndash224


Man also possesses a power by which he may see his friends and thecircumstances by which they are surrounded although such persons may bea thousand miles away from him at that time

Paracelsus

I Introduction and Background

This concise statement of the remote perception hypothesis was proffered by therenowned 16th-century physician and philosopher Paracelsus in a section of hiswritings devoted to the role of lsquolsquoactive imaginationrsquorsquo in manrsquos representation ofhis universe(1) His observation was certainly not the first recorded allusion tosuch anomalous human capabilities This lsquolsquopowerrsquorsquo has been acknowledged invirtually every culture since the dawn of human civilization and invoked undera multitude of names including among many others divination prophecyoracle scrying clairvoyance and second sight

In the more recent history of Western science a considerable body ofliterature describing scholarly investigations of lsquolsquoextrasensory perceptionrsquorsquoalready had been amassed when in the mid-1970s Puthoff and Targ at StanfordResearch Institute introduced a new scientific protocol for empirical in-vestigation of the phenomenon they termed lsquolsquoremote viewingrsquorsquo(23) Theirprocedure required one individual referred to as the lsquolsquopercipientrsquorsquo to attempt todescribe the geographical ambience surrounding another person the lsquolsquoagentrsquorsquowhose location was inaccessible to the percipient by any known sensory meansTheir striking data included many perceptions that were virtually photographicin accuracy and produced an overall statistical yield well beyond chanceexpectations Over the subsequent quarter century numerous replications of theoriginal SRI studies have been reported(4ndash16) including a number of originallyclassified government-sponsored investigations(17ndash21) most of which display theambiguous mixtures of successes and failures that seem to characterize mostserious anomalies research Notwithstanding the majority of these studiesdemonstrate a sufficient degree of anomalous information acquisition to justifycontinued scholarly exploration of this mystifying process

One of the largest extant databases comprising 653 formal and 126 non-formalexperimental trials was produced between 1976 and 1999 as one of the threemajor components of the Princeton Engineering Anomalies Research (PEAR)program The other two segments concerning anomalies in humanmachineinteractions and theoretical modeling have been reported extensively in thisjournal and elsewhere The purpose of this paper is to describe the procedures andsummarize the full results of our remote perception studies and to explore theirimplications for better comprehension of this currently inexplicable communi-cation capability To achieve this most concisely we shall refer frequently toa number of earlier publicationsand technical reports wherein all the datasets andanalytical methods are presented in greater detail(22ndash26)

The first phase of this PEAR work evolved from a body of prior experimentsconducted between 1976 and 1979 by one of the authors (BJD) at Mundelein




II Protocol























tion are available





















Scoringmethod

Chancemean

ChanceSD

Meanscore

Compositez-score


Trialsp 05

Trialsp 05




All data (N 5 300)A 05610 1053 05979 6070 6 3 102 10 30 (7) 10 (2)B 05042 1207 05447 5809 3 3 102 9 44 (10) 15 (3)C 10005 2380 10873 6320 1 3 102 10 40 (8) 13 (3)D 06512 0935 06806 5447 3 3 102 8 35 (8) 12 (3)E 10034 1330 10554 6773 6 3 102 12 42 (5) 14 (2)


















TA

BL

E2

Bin

ary

PRP

Dat

aS

umm

arie

s(S

cori

ngM

etho

dB

)

Sub

set

T

rial

sM

ean

scor

eE

ffec

tsi

ze

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Prob

abili

ty(o

ne-t

aile

d)

Tri

als

p

05

T

rial

sp

0

5

Tri

als

p

50

All

tria

ls41

15

364

279

61

351

060

564

78

310

29

47(1

2)11

(3

)

59

For

mal

tria

ls33

65

447

347

61

521

083

635

51

310

210

44(8

)13

(2

)

62

Non

-for

mal

tria

ls75

496

92

046

62

780

910

20

399

655

3(4

)4

(5

)44

Ab

initi

o27

75

345

263

61

611

033

437

86

310

26

31(5

)11

(2

)

59

Ex

post

fact

o59

594

27

546

417

120

35

792

33

102

914

(2)

24

(3

)75

Sin

gle

perc

ipie

nt21

65

489

382

61

941

098

561

31

310

28

34(6

)16

(3

)

60

Mul

tiple

perc

ipie

nt12

05

404

312

62

511

049

341

63

310

24

12(3

)10

(3

)

63

Inst

ruct

edta

rget

s12

55

653

516

62

671

140

577

14

310

29

23(5

)18

(4

)

65

Vol

itio

nal

targ

ets

211

532

22

446

191

106

63

549

23

102

425

(3)

12

(1

)60

Sum

mer

tria

ls24

45

466

363

61

831

099

566

37

310

29

35(5

)14

(2

)

65

Win

ter

tria

ls92

540

73

156

286

104

33

017

13

102

313

(2)

14

(2

)57

Chi

cago

targ

ets

316

189

957

65

871

189

533

05

310

28

10(1

)32

(3

)

81

Pri

ncet

onta

rget

s10

65

504

394

62

861

110

406

02

310

25

14(3

)13

(3

)

62

Tar

gets

else

whe

re19

95

267

199

61

941

051

281

02

310

23

20(3

)10

(2

)

58

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05







TA

BL

E3

For

mal

Ab

Initi

oD

ata

Sum

mar

ies

(Sco

ring

Met

hod

B)

Subs

et

Tri

als

Mea

nsc

ore

Eff

ect

size

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Pro

babi

lity

(one

-tai

led)

T

rial

sp

0

5

Tri

als

p

05

T

rial

sp

5

0

All

tria

ls27

75

345

263

61

611

034

437

86

310

26

31(5

)11

(2

)

59

Sing

lepe

rcip

ient

194

537

02

846

197

106

33

949

43

102

524

(6)

12

(3

)56

M

ultip

lepe

rcip

ient

835

321

243

62

750

974

221

50

135

(1)

6(1

)

64

Inst

ruct

edta

rget

s94

541

63

226

296

111

53

122

93

102

411

(5)

12

(5

)61

V

olit

iona

lta

rget

s18

35

308

233

61

941

020

314

88

310

24

21(1

)11

(

05

)60

Sum

mer

tria

ls19

55

374

287

61

951

058

401

33

310

25

24(4

)12

(2

)

62

Win

ter

tria

ls82

530

82

336

285

100

22

107

018

7(2

)9

(2

)56

Prin

ceto

nta

rget

s10

65

504

394

62

811

125

406

02

310

25

14(4

)13

(4

)

62

Tar

gets

else

whe

re17

15

243

180

61

971

000

234

89

310

23

16(1

)9

(05

)

59

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05










VII FIDO Scoring




























Scoringmethod

Effectsize


Trialsp 05

Trialsp 05

Trialsp 50




















Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50


distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52


2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



















Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References























































II Protocol























tion are available





















Scoringmethod

Chancemean

ChanceSD

Meanscore

Compositez-score


Trialsp 05

Trialsp 05




All data (N 5 300)A 05610 1053 05979 6070 6 3 102 10 30 (7) 10 (2)B 05042 1207 05447 5809 3 3 102 9 44 (10) 15 (3)C 10005 2380 10873 6320 1 3 102 10 40 (8) 13 (3)D 06512 0935 06806 5447 3 3 102 8 35 (8) 12 (3)E 10034 1330 10554 6773 6 3 102 12 42 (5) 14 (2)


















TA

BL

E2

Bin

ary

PRP

Dat

aS

umm

arie

s(S

cori

ngM

etho

dB

)

Sub

set

T

rial

sM

ean

scor

eE

ffec

tsi

ze

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Prob

abili

ty(o

ne-t

aile

d)

Tri

als

p

05

T

rial

sp

0

5

Tri

als

p

50

All

tria

ls41

15

364

279

61

351

060

564

78

310

29

47(1

2)11

(3

)

59

For

mal

tria

ls33

65

447

347

61

521

083

635

51

310

210

44(8

)13

(2

)

62

Non

-for

mal

tria

ls75

496

92

046

62

780

910

20

399

655

3(4

)4

(5

)44

Ab

initi

o27

75

345

263

61

611

033

437

86

310

26

31(5

)11

(2

)

59

Ex

post

fact

o59

594

27

546

417

120

35

792

33

102

914

(2)

24

(3

)75

Sin

gle

perc

ipie

nt21

65

489

382

61

941

098

561

31

310

28

34(6

)16

(3

)

60

Mul

tiple

perc

ipie

nt12

05

404

312

62

511

049

341

63

310

24

12(3

)10

(3

)

63

Inst

ruct

edta

rget

s12

55

653

516

62

671

140

577

14

310

29

23(5

)18

(4

)

65

Vol

itio

nal

targ

ets

211

532

22

446

191

106

63

549

23

102

425

(3)

12

(1

)60

Sum

mer

tria

ls24

45

466

363

61

831

099

566

37

310

29

35(5

)14

(2

)

65

Win

ter

tria

ls92

540

73

156

286

104

33

017

13

102

313

(2)

14

(2

)57

Chi

cago

targ

ets

316

189

957

65

871

189

533

05

310

28

10(1

)32

(3

)

81

Pri

ncet

onta

rget

s10

65

504

394

62

861

110

406

02

310

25

14(3

)13

(3

)

62

Tar

gets

else

whe

re19

95

267

199

61

941

051

281

02

310

23

20(3

)10

(2

)

58

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05







TA

BL

E3

For

mal

Ab

Initi

oD

ata

Sum

mar

ies

(Sco

ring

Met

hod

B)

Subs

et

Tri

als

Mea

nsc

ore

Eff

ect

size

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Pro

babi

lity

(one

-tai

led)

T

rial

sp

0

5

Tri

als

p

05

T

rial

sp

5

0

All

tria

ls27

75

345

263

61

611

034

437

86

310

26

31(5

)11

(2

)

59

Sing

lepe

rcip

ient

194

537

02

846

197

106

33

949

43

102

524

(6)

12

(3

)56

M

ultip

lepe

rcip

ient

835

321

243

62

750

974

221

50

135

(1)

6(1

)

64

Inst

ruct

edta

rget

s94

541

63

226

296

111

53

122

93

102

411

(5)

12

(5

)61

V

olit

iona

lta

rget

s18

35

308

233

61

941

020

314

88

310

24

21(1

)11

(

05

)60

Sum

mer

tria

ls19

55

374

287

61

951

058

401

33

310

25

24(4

)12

(2

)

62

Win

ter

tria

ls82

530

82

336

285

100

22

107

018

7(2

)9

(2

)56

Prin

ceto

nta

rget

s10

65

504

394

62

811

125

406

02

310

25

14(4

)13

(4

)

62

Tar

gets

else

whe

re17

15

243

180

61

971

000

234

89

310

23

16(1

)9

(05

)

59

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05










VII FIDO Scoring




























Scoringmethod

Effectsize


Trialsp 05

Trialsp 05

Trialsp 50




















Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50


distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52


2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



















Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References








































































tion are available





















Scoringmethod

Chancemean

ChanceSD

Meanscore

Compositez-score


Trialsp 05

Trialsp 05




All data (N 5 300)A 05610 1053 05979 6070 6 3 102 10 30 (7) 10 (2)B 05042 1207 05447 5809 3 3 102 9 44 (10) 15 (3)C 10005 2380 10873 6320 1 3 102 10 40 (8) 13 (3)D 06512 0935 06806 5447 3 3 102 8 35 (8) 12 (3)E 10034 1330 10554 6773 6 3 102 12 42 (5) 14 (2)


















TA

BL

E2

Bin

ary

PRP

Dat

aS

umm

arie

s(S

cori

ngM

etho

dB

)

Sub

set

T

rial

sM

ean

scor

eE

ffec

tsi

ze

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Prob

abili

ty(o

ne-t

aile

d)

Tri

als

p

05

T

rial

sp

0

5

Tri

als

p

50

All

tria

ls41

15

364

279

61

351

060

564

78

310

29

47(1

2)11

(3

)

59

For

mal

tria

ls33

65

447

347

61

521

083

635

51

310

210

44(8

)13

(2

)

62

Non

-for

mal

tria

ls75

496

92

046

62

780

910

20

399

655

3(4

)4

(5

)44

Ab

initi

o27

75

345

263

61

611

033

437

86

310

26

31(5

)11

(2

)

59

Ex

post

fact

o59

594

27

546

417

120

35

792

33

102

914

(2)

24

(3

)75

Sin

gle

perc

ipie

nt21

65

489

382

61

941

098

561

31

310

28

34(6

)16

(3

)

60

Mul

tiple

perc

ipie

nt12

05

404

312

62

511

049

341

63

310

24

12(3

)10

(3

)

63

Inst

ruct

edta

rget

s12

55

653

516

62

671

140

577

14

310

29

23(5

)18

(4

)

65

Vol

itio

nal

targ

ets

211

532

22

446

191

106

63

549

23

102

425

(3)

12

(1

)60

Sum

mer

tria

ls24

45

466

363

61

831

099

566

37

310

29

35(5

)14

(2

)

65

Win

ter

tria

ls92

540

73

156

286

104

33

017

13

102

313

(2)

14

(2

)57

Chi

cago

targ

ets

316

189

957

65

871

189

533

05

310

28

10(1

)32

(3

)

81

Pri

ncet

onta

rget

s10

65

504

394

62

861

110

406

02

310

25

14(3

)13

(3

)

62

Tar

gets

else

whe

re19

95

267

199

61

941

051

281

02

310

23

20(3

)10

(2

)

58

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05







TA

BL

E3

For

mal

Ab

Initi

oD

ata

Sum

mar

ies

(Sco

ring

Met

hod

B)

Subs

et

Tri

als

Mea

nsc

ore

Eff

ect

size

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Pro

babi

lity

(one

-tai

led)

T

rial

sp

0

5

Tri

als

p

05

T

rial

sp

5

0

All

tria

ls27

75

345

263

61

611

034

437

86

310

26

31(5

)11

(2

)

59

Sing

lepe

rcip

ient

194

537

02

846

197

106

33

949

43

102

524

(6)

12

(3

)56

M

ultip

lepe

rcip

ient

835

321

243

62

750

974

221

50

135

(1)

6(1

)

64

Inst

ruct

edta

rget

s94

541

63

226

296

111

53

122

93

102

411

(5)

12

(5

)61

V

olit

iona

lta

rget

s18

35

308

233

61

941

020

314

88

310

24

21(1

)11

(

05

)60

Sum

mer

tria

ls19

55

374

287

61

951

058

401

33

310

25

24(4

)12

(2

)

62

Win

ter

tria

ls82

530

82

336

285

100

22

107

018

7(2

)9

(2

)56

Prin

ceto

nta

rget

s10

65

504

394

62

811

125

406

02

310

25

14(4

)13

(4

)

62

Tar

gets

else

whe

re17

15

243

180

61

971

000

234

89

310

23

16(1

)9

(05

)

59

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05










VII FIDO Scoring




























Scoringmethod

Effectsize


Trialsp 05

Trialsp 05

Trialsp 50




















Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50


distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52


2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



















Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References





































































Scoringmethod

Chancemean

ChanceSD

Meanscore

Compositez-score


Trialsp 05

Trialsp 05




All data (N 5 300)A 05610 1053 05979 6070 6 3 102 10 30 (7) 10 (2)B 05042 1207 05447 5809 3 3 102 9 44 (10) 15 (3)C 10005 2380 10873 6320 1 3 102 10 40 (8) 13 (3)D 06512 0935 06806 5447 3 3 102 8 35 (8) 12 (3)E 10034 1330 10554 6773 6 3 102 12 42 (5) 14 (2)


















TA

BL

E2

Bin

ary

PRP

Dat

aS

umm

arie

s(S

cori

ngM

etho

dB

)

Sub

set

T

rial

sM

ean

scor

eE

ffec

tsi

ze

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Prob

abili

ty(o

ne-t

aile

d)

Tri

als

p

05

T

rial

sp

0

5

Tri

als

p

50

All

tria

ls41

15

364

279

61

351

060

564

78

310

29

47(1

2)11

(3

)

59

For

mal

tria

ls33

65

447

347

61

521

083

635

51

310

210

44(8

)13

(2

)

62

Non

-for

mal

tria

ls75

496

92

046

62

780

910

20

399

655

3(4

)4

(5

)44

Ab

initi

o27

75

345

263

61

611

033

437

86

310

26

31(5

)11

(2

)

59

Ex

post

fact

o59

594

27

546

417

120

35

792

33

102

914

(2)

24

(3

)75

Sin

gle

perc

ipie

nt21

65

489

382

61

941

098

561

31

310

28

34(6

)16

(3

)

60

Mul

tiple

perc

ipie

nt12

05

404

312

62

511

049

341

63

310

24

12(3

)10

(3

)

63

Inst

ruct

edta

rget

s12

55

653

516

62

671

140

577

14

310

29

23(5

)18

(4

)

65

Vol

itio

nal

targ

ets

211

532

22

446

191

106

63

549

23

102

425

(3)

12

(1

)60

Sum

mer

tria

ls24

45

466

363

61

831

099

566

37

310

29

35(5

)14

(2

)

65

Win

ter

tria

ls92

540

73

156

286

104

33

017

13

102

313

(2)

14

(2

)57

Chi

cago

targ

ets

316

189

957

65

871

189

533

05

310

28

10(1

)32

(3

)

81

Pri

ncet

onta

rget

s10

65

504

394

62

861

110

406

02

310

25

14(3

)13

(3

)

62

Tar

gets

else

whe

re19

95

267

199

61

941

051

281

02

310

23

20(3

)10

(2

)

58

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05







TA

BL

E3

For

mal

Ab

Initi

oD

ata

Sum

mar

ies

(Sco

ring

Met

hod

B)

Subs

et

Tri

als

Mea

nsc

ore

Eff

ect

size

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Pro

babi

lity

(one

-tai

led)

T

rial

sp

0

5

Tri

als

p

05

T

rial

sp

5

0

All

tria

ls27

75

345

263

61

611

034

437

86

310

26

31(5

)11

(2

)

59

Sing

lepe

rcip

ient

194

537

02

846

197

106

33

949

43

102

524

(6)

12

(3

)56

M

ultip

lepe

rcip

ient

835

321

243

62

750

974

221

50

135

(1)

6(1

)

64

Inst

ruct

edta

rget

s94

541

63

226

296

111

53

122

93

102

411

(5)

12

(5

)61

V

olit

iona

lta

rget

s18

35

308

233

61

941

020

314

88

310

24

21(1

)11

(

05

)60

Sum

mer

tria

ls19

55

374

287

61

951

058

401

33

310

25

24(4

)12

(2

)

62

Win

ter

tria

ls82

530

82

336

285

100

22

107

018

7(2

)9

(2

)56

Prin

ceto

nta

rget

s10

65

504

394

62

811

125

406

02

310

25

14(4

)13

(4

)

62

Tar

gets

else

whe

re17

15

243

180

61

971

000

234

89

310

23

16(1

)9

(05

)

59

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05










VII FIDO Scoring




























Scoringmethod

Effectsize


Trialsp 05

Trialsp 05

Trialsp 50




















Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50


distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52


2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



















Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References




































































TA

BL

E2

Bin

ary

PRP

Dat

aS

umm

arie

s(S

cori

ngM

etho

dB

)

Sub

set

T

rial

sM

ean

scor

eE

ffec

tsi

ze

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Prob

abili

ty(o

ne-t

aile

d)

Tri

als

p

05

T

rial

sp

0

5

Tri

als

p

50

All

tria

ls41

15

364

279

61

351

060

564

78

310

29

47(1

2)11

(3

)

59

For

mal

tria

ls33

65

447

347

61

521

083

635

51

310

210

44(8

)13

(2

)

62

Non

-for

mal

tria

ls75

496

92

046

62

780

910

20

399

655

3(4

)4

(5

)44

Ab

initi

o27

75

345

263

61

611

033

437

86

310

26

31(5

)11

(2

)

59

Ex

post

fact

o59

594

27

546

417

120

35

792

33

102

914

(2)

24

(3

)75

Sin

gle

perc

ipie

nt21

65

489

382

61

941

098

561

31

310

28

34(6

)16

(3

)

60

Mul

tiple

perc

ipie

nt12

05

404

312

62

511

049

341

63

310

24

12(3

)10

(3

)

63

Inst

ruct

edta

rget

s12

55

653

516

62

671

140

577

14

310

29

23(5

)18

(4

)

65

Vol

itio

nal

targ

ets

211

532

22

446

191

106

63

549

23

102

425

(3)

12

(1

)60

Sum

mer

tria

ls24

45

466

363

61

831

099

566

37

310

29

35(5

)14

(2

)

65

Win

ter

tria

ls92

540

73

156

286

104

33

017

13

102

313

(2)

14

(2

)57

Chi

cago

targ

ets

316

189

957

65

871

189

533

05

310

28

10(1

)32

(3

)

81

Pri

ncet

onta

rget

s10

65

504

394

62

861

110

406

02

310

25

14(3

)13

(3

)

62

Tar

gets

else

whe

re19

95

267

199

61

941

051

281

02

310

23

20(3

)10

(2

)

58

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05







TA

BL

E3

For

mal

Ab

Initi

oD

ata

Sum

mar

ies

(Sco

ring

Met

hod

B)

Subs

et

Tri

als

Mea

nsc

ore

Eff

ect

size

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Pro

babi

lity

(one

-tai

led)

T

rial

sp

0

5

Tri

als

p

05

T

rial

sp

5

0

All

tria

ls27

75

345

263

61

611

034

437

86

310

26

31(5

)11

(2

)

59

Sing

lepe

rcip

ient

194

537

02

846

197

106

33

949

43

102

524

(6)

12

(3

)56

M

ultip

lepe

rcip

ient

835

321

243

62

750

974

221

50

135

(1)

6(1

)

64

Inst

ruct

edta

rget

s94

541

63

226

296

111

53

122

93

102

411

(5)

12

(5

)61

V

olit

iona

lta

rget

s18

35

308

233

61

941

020

314

88

310

24

21(1

)11

(

05

)60

Sum

mer

tria

ls19

55

374

287

61

951

058

401

33

310

25

24(4

)12

(2

)

62

Win

ter

tria

ls82

530

82

336

285

100

22

107

018

7(2

)9

(2

)56

Prin

ceto

nta

rget

s10

65

504

394

62

811

125

406

02

310

25

14(4

)13

(4

)

62

Tar

gets

else

whe

re17

15

243

180

61

971

000

234

89

310

23

16(1

)9

(05

)

59

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05










VII FIDO Scoring




























Scoringmethod

Effectsize


Trialsp 05

Trialsp 05

Trialsp 50




















Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50


distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52


2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



















Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References


























































TA

BL

E3

For

mal

Ab

Initi

oD

ata

Sum

mar

ies

(Sco

ring

Met

hod

B)

Subs

et

Tri

als

Mea

nsc

ore

Eff

ect

size

99

Con

fide

nce

inte

rval

SD

z-

scor

eC

ompo

site

z-sc

ore

Pro

babi

lity

(one

-tai

led)

T

rial

sp

0

5

Tri

als

p

05

T

rial

sp

5

0

All

tria

ls27

75

345

263

61

611

034

437

86

310

26

31(5

)11

(2

)

59

Sing

lepe

rcip

ient

194

537

02

846

197

106

33

949

43

102

524

(6)

12

(3

)56

M

ultip

lepe

rcip

ient

835

321

243

62

750

974

221

50

135

(1)

6(1

)

64

Inst

ruct

edta

rget

s94

541

63

226

296

111

53

122

93

102

411

(5)

12

(5

)61

V

olit

iona

lta

rget

s18

35

308

233

61

941

020

314

88

310

24

21(1

)11

(

05

)60

Sum

mer

tria

ls19

55

374

287

61

951

058

401

33

310

25

24(4

)12

(2

)

62

Win

ter

tria

ls82

530

82

336

285

100

22

107

018

7(2

)9

(2

)56

Prin

ceto

nta

rget

s10

65

504

394

62

811

125

406

02

310

25

14(4

)13

(4

)

62

Tar

gets

else

whe

re17

15

243

180

61

971

000

234

89

310

23

16(1

)9

(05

)

59

N

umbe

rsin

pare

nthe

ses

indi

cate

num

ber

oftr

ials

wit

hne

gati

vez-

scor

es

p

05










VII FIDO Scoring




























Scoringmethod

Effectsize


Trialsp 05

Trialsp 05

Trialsp 50




















Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50


distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52


2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



















Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References





























































VII FIDO Scoring




























Scoringmethod

Effectsize


Trialsp 05

Trialsp 05

Trialsp 50




















Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50


distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52


2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



















Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References













































































Scoringmethod

Effectsize


Trialsp 05

Trialsp 05

Trialsp 50




















Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50


distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52


2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



















Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References





































































Scoringmethod

Effectsize


Seriesp 05

Trialsp 05

Trialsp 05

Trialsp 50


distributive2 00745 2 0912 819 1 (0) 6 (9) 4 (6) 52


2 00394 2 0483 685 2 (0) 6 (8) 4 (5) 53



















Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References





































































Participants

Database

Trials

Series

Agents



Ex post facto 59 7 4 13 16 5792 754 3 3 102 9

Ab initio 277 42 13 26 30 4378 263 6 3 102 6

Initial trials 168 29 9 21 23 4582 354 2 3 102 6


TOTAL 653 88 39 59 69 5418 212 3 3 102 8





























Acknowledgments



Appendix A










References















































































Acknowledgments



Appendix A










References



























































References





















































information and uncertainty in remote perception research · 2016-12-07 · information and...

Documents