SPR_Vol67 Scientific Papers on Telepathy

8/13/2019 SPR_Vol67 Scientific Papers on Telepathy

http://slidepdf.com/reader/full/sprvol67-scientific-papers-on-telepathy 1/8

,lotrrnul ol the Socc/"y, for Psychical lleseerclr I Vo I . 6 7 . s , N o . 8 7 2

EXPERIMDNTAL TESTS FOR TELEPHONE TELEPATHY

by Runr:RT Sntit,DRAI(E nd pAMEt,nSnannt

AI]STRACT

Many people claim to hnve known rvho was calling before they picked up t6etclephonc, or to havc thought for no apparent reason about someone, who thencal led ' We carr ied or t t a scr ies of exper imcnts to tes t whether or not people real ty

could te l l * 'ho was tc lephoning. Each par t ic ipant had four potent ia l ca l lers , an dwhen the teleplrone rang had to guess who was calling before the other personslrcke. By chatrcc tltc sttcccss rnte worrld have been 25Yo. n a total of S?1 trials,involving 63 participants, the overall success rate was 40%, with g5%-confidencelirnits from 30% to 4l'9o, This effcct is hrrgcl.y signifir:nnt stati.stically (p= 4 x l0-r6).Wc obtairtcd .sirrrilar sigrrificant po.sitivc cffccts b<;th when the cnlls were rnadc ntrandolrrl.v :lrosctr irtte's an<l rvhcn thc r:nlls werc rnndc nt fixed times known to thestrbject in ntlvnrrcu. With 37 pnrticil)nnts, we conrpnrcd the success rates wichfamil iar and unfntni l inr cal lers and fount l o very s t r ik ing difference . With famil iarcnllers, 53% of [lte grresses were correct (n = lg0; p = I x 10-16). With unfarnilinrcallers. only 25Yo of thc guesses were correct, exactly at the chance level. Thisdifference between the responses with familiar and unfamiliar callers is highlysignificant (p = 3 x l0-7). We also investigated the effects of distance betweenthe cal lers and pnr t ic ipants . With overseas cal lers , a t leas t 1 ,000 miles away, thesuccess rate was 65% (n=43i p =3 x l0-E) . With cal lers in the UK, the successrate wos lower (35%). In most c8ses, the overseas cnllers were people to whomt'he participants werc closely bonded. For the successful identification of callers,emotional closetress seemed to be more important than physical proximity.

INrnopucttotrrSeemingly elepathic experiences with telephones re common. Many people

hnve found thnt for no nl)l)nrent eason hey start thinking about a particularperson. hen the phone rings and that person s on the line. Or else, when th ete lephone tar ts r inging, hey have an in tui t ion about who is cal l ing, and turnout to be correct. Usually such calls are from people whom the person knowswel l .

Such experiences seem to be the comrnonest kind of apparent telepathyin the nrodern world (Sheldrake, 2001, 2003; Brown & Sheldrake, iOOZj.

Strrprisingly, psycltic esearchers and porapsychologists eem o have ignoredthe phenomenon.Could apparent telephone elepathy merely be a matter of chance coincid-

ence? Perhaps people often have thoughts aborrt others for no particularreason. By chance, lrese houghts may sotnet.inres e followed by a telepholecall from that person. If people only remember the times they are right an dforget the times they are wrong, an illusion of telepathy may be createcl by acombinat ion f chance oincidence nd select ive memory.

Al ternat ively, a person may be expect ing a cal l around a par t icular t imefrom a particular person, but may be trnconsciotrs f this expectation. So whenthe cal l comes here s no need o invoke elepathy because n unconscious x.pecta t ion ould expla in t ins tead. The trouble s that unconscious xpecta t ions

July 2003J Experimental Tests for Telephone Telepath

are elusive. This may be an untes[able hypothesis, because f the expectationof a te lephone al l s unconscious , ow csn anyone prove hat it i s rea l ly there ' /And if it is really there, then might this expectation be a reeult of telepatlri ' .rather than an alternative to it ?

The best way to resolve hese questions s by means of experimental test:that can be evaluated sta t is t ica l ly. We have developed s imple procedure ;which participants receive a call from one of fotrr differenc callers. They knor' ,who the potent ia l ca l lers are , but do not know which one wil l be cal l ing n an; .g iven est , because he cal ler has been picked at randorn by the exper imenter

They have to guess who the caller s before he caller says anything. By chan, ' ,they would be right about one time in four, or 25 per cent of the time. Are thc',right significantly more often than would be expected on the basis of randonrguessing? n this paper we clescribe he results of 571 such tests. The scorr,were very signi{icant ly bove hance evels .

We also compared he partici;rants ' esponses o calls from people hey knev'wel l wi th those rom s t rnngcrs . And wc explorcd he cffects of dis tance , wi l 'ca l lers up to 11,000 r i lcsawny.

MBruoos

Recruiting Part:icipants

In a preliminary experiment, one of us (PS) served as participant and th eother (RS) as experimenter. For strbsequent xperiments, we recruited partici-pants through advertisements n the Part-Time Work section of newspapers - ' ' i 'through a recruitment website called www.holrecruit.co.uft. ur advertisernenl:rread: "Do you knorv who is ringing before you pick up the phone? Good pay frrrfun and simple experiments as part of psychic esearch project."

We initially offered a payment of f l0 per two-trial session, nd later tl0 fo ra l - t r ia l sess ion. We sent detni ls of the tes t proccdure o the people who repl ic<lto these advertisernents, and asked thcm to nominate people to whose cal :;they thought they might respond. We asked hem to check hat these peoplr,would be willing to tahe part, anrl asked thenr to supply us with their contactdetails and telephone numbers. We also asked participants to tell us whciithey would be able to take part in tests, and to check that their callers woulrlbe free to call them at those times. Because t was the responsibility of thr:participants to ensure that their callers wotrld be available, and they did not.

get paid for a t r ia l i f the cal lers were not , n mosL ases he cal lers were ndee, lavai lable . f the randonr ly-se lected al ler wos not avai lable , he t r ia l wa scancelled.

Some par t ic ipants were t rnable o complcte he tcn- t r ia l ser ies or a var ie tyof reasons , ncluding changes n thei r personal i rcunrs tances , uch as s tar t inga ful l - t ime ob, or becnrrsc nc or more of t l re i r cnl lcr. swas unable o cont in t rctaking par t . Ideal ly, all par t ic ipnnts would have completed ll l0 t r ia ls , as thel 'had agreed, o avoid the poss ibi l i tyof opt ional s topping, whereby par t ic ipantswho were not scor ing above hance evels might have decl ined o do nrore es ts .But i f th is happened t a l l , i t was rare . Most par t ic ipnnts vho did not complet i 'a l l 10 t r ia ls fa i led to do so becouse ne or more of thci r cal lers was unable orunwilling to contintre aking 1lart.

l8 . l



Jourrwl of the Society for Psychical Research IVol .67.e , o .872

Participants in the first series of experiments were recruited mainly throughlocal newspapers in the north of England. For the second series, most wererecruited through wwut.lrctrecruit.co.uft. hen we wanted to find participantswith callers overseas, we advertised in free newspapers aimed at youngAustralians, New Zealanders and South Africans living in England (?N? an dSoulhern Cross).

&IlersIn our first series of experiments, we asked participants to nominate all four

callers. This restricted the number of applicants who were able to participate,because most were unable to find four people o whom they thought they mightrespond telepathically and who were able and willing to take part.

In our second series, we asked participants to nominate a minimum of tw ocallers, and we supplied the others, who were strangers to the participants.This procedure had the advantage of enabling us to recruit more participants,and it also enabled us to conrpnre heir responses o familiar and unfamiliarcallers. Most people nominated only two callers, but some nominated three;hence the totnl number of trials with fnrnilinr callers was larger than withunfamil iar ca l lers .

Test ProceduresIn all tests, participants used and-line elephones, nd of course we required

them to use phones without caller identification systems. We avoided the use

of mobile phones because ll models have built-in caller dentification displays.For each trial, there were four potential callers. The participants knew whothey were and also knew that one of them would be selected t random by thethrow of a die. For the throw of the die, we used high-quality casino dice anda ribbed casino-style dice cup, purchnsed n Las Vegas. Each of the potentialcallers was assigned a number from I to 4, and was s6lected y the thrown dieshowing one of these numbers. If the die showed 5 or 6, it was thrown againunt i l a number between and 4 came up .

We used hree different procedures, nvolving progressive implifications. nall cases, hen a trial was taking place, he participant picked up the telephoneand immediately indicated the person guessed y saying that person's name.The caller then revealed his or her identity, so the participants receivedimmediate feedback.

I In Method l, participants took part in two trials per session. The tw oeallers were selected at random by two throws of the die (ignoring 5 and 6). Ifthe die shorved he same number wice , hen the same person was the cal ler nboth trials.

The times of the trinls wcre also selectecl t , andom because we wanted totest the idea that people can tell who is calling even f the calls are at randomtimes. Participants rvere not told at what time the calls would be made,although of course hey knew that they would occur within the test session.

Test sessions were ustrally an hour long, beginning and ending at timesagreed n advance with participants and their callers. To pick the call timesat random, the sess ion was divided nto 6 , and the beginning of one of theseperiods was selected y the throw of a die. ftor example, f the test session wa s

18 6 r 87

July 20031 Experintental Tests for Telephone Telepathy

f rom 10-11 am, the s ix per iods egan at 10-minute n tervals , tar t ing at 10. f0 .Thus i f the die showet l 4 , then the tes t would be at 10.40. The die was thenthrown again to selcct hc t imc for thc tes t wi[h thc othcr cal lcr. I f a I cameup, th is was lO. lO arn . f thc same number canlc up twice , hc die was thrownagain .

The exper imentcr (e i ther I tS or I 'S) te lephoncd he randomly selectcdcallers n advance, usually an hour or two beforehand, and asked them to callat the time selected. We asked callers to think nbout the participant for abouta minute before al l ing. We also ang the cal lers who hacl not been se lected o

tell them that t hey were not involved n this test session.A few minutes after the tests, he experirncnter ang the participant to as k

what h is or her guess had been, and in some cases also asked he cal lers . nno cases did cal lers ancl par t ic i l lants disagree . 'he exper imentcr ecordcd hcresul t , not ing down the date and t i rne of cnch . r ia l , he cnl ler and the gucss .

This method was used n our prel iminary exper iment and in o t r r f i rs t ser ieswi th 17 par t ic ipants , or a to ta l of 198 r ia ls .

2 In Method 2, we simplified this procedure by trsing ixecl imes for the twotr ia ls in a sess ion, or cxample a t 10.15 and 10.30 am. Thc f i rs t ca l ler to beselected was asked o cal l a t 10.15, nd the second with a I in 4 chance hatit would be the same person again) o cal l a t 10.30. These people were cal ledin advance of the test session and told they had been selected and when tocall. Those who had not been selected were called and told they had not beenselected. he experimenter ound out and recorded what the guesses ad beenas in Method 1. We used this method for the last 5 part,icipants n our firstseries of experiments and the first 3 in the second series, who took part in atotal of 87 trials.

3 In Method 3 , there wns only one r ia l per sess ion. he exper imenter PS )selected he cal ler nt . random lcss t l ran 15 minrr tcs leforc the prearrangedtes t t ime. Soy, for example , hc tes t t ime was 2.30 p.h . , then the cal ler wa sselected at random ab 2.L5, and notified before 2.2O.We used his method with37 participants n our second eries of experiments.

To start with (Method 3A) the experinrenter PS) also called he three peoplenot se lected o te l l thenr so . Star t ing on 15.11.01, his s tep was el iminated(Method 38) by te l l ing the cal lers hat i f they had not been not i f ied at leas t 5minutes before he test time, then they had not been selected.This implilicationmade it possible o carry ouI a series of separate, one-session rials in rapid

success ion, ypical ly a t 15-minute ntervals . The exper imenter ound out andrecorded what the guess harl been as n Method I-

In the second series of experiments, we tested 34 participants using thismethod, n a to tnl of 268 t r ia ls .

StatisticsFor the testing of the hypothesis that the proportion of correct guesses

would be above he chance evel of O.25, r 25Yo, e used he exact binomialtes t (Siegel & Caste l lan , 9B8) . The nul l hypothes is was that the probabi l i ty ofa correct guess s 0.25 or 25Yo.

For combining the resrr l t s of di fferent exper iments we used the Stouffermethod (Rosenthal , 991) .



Jonrnol of the Socie6' for Psychical Researclt IVol . 67 . s , No .872

For the comparison of results with familiar or unfamiliar callers, and th eresults in first and second rials, we used the Fisher exact test (Siegel &Caste l lan , 988) .

We calcrrlated xact 95%-confidence imits for the true probability of a correctresponse s descr ibed y Hahn & Meeker 1991) .

RusuursA Preliminory Experiment

In a preliminary experiment, started in April 1999, PS was the participant

and RS the experinrenter. PS did not claim to have any special psychic powers,but she had noticed that she sometimes seemed to know who was ringingbefore she answered he telephone.

There were 5 potential callers, of whom only 4 took part in any given trial,depending orl their availability. PS was informed before each trial which 4r:allers would be tnking part. The potential callers were Angie and Cathie (hersisters), Muriel (her nrother), her close riend Polly, and RS. Apart from RS ,who was in l .ondon,25O niles awoy, the others ived within a lO-mile radius ofPS's home n Rnnrsbottottt , renter Mnnchester.

We conducted 30 trials according o Method l. In each rial there was a 1 in4 chance of PS naming the caller correctly by chance. n other words, f PS werejust guessing without the help of telepathy, she would have been right about25o/o f the time, an average of 7.5 correct guesses n 30 trials. In fact 13 out of30 of her guesses were correct (43%), significantly above chance (p = O.O2).

The results with individual callers are shown in Table 1. The highestproportion of PS's correct guesses 67%) was with RS as caller (p = 0.01), whowas furthest away.

ToblePam Smorl'e Responces o Calls front 5 Differe.nt Callers (her aisters Angie and Cathie, her

molher Muriel, her friend I'olly and IIS, Iluperl). Only'4 callere tooh part in any giuen

lriol, and f 'S rlns lold in aduance tuhich 4 eollers we.re nuolued (Method I),

Cnl le r Calls Right % right

Angie

Cathie

MurielPolly

Rupert

43

6II

I

I

I4

6

2633

t 75067

Total 30 l 3 4 3

The First Experimental SeriesTo find out if these results were replicable, w€ carried out a series of

exper iments with 2l pnr t ic i l lonts recrui tec l hrough local newspapers n thenorth of England (in Greater Manchester, Lancashire and Yorkshire). Weasked all of them to take par t in l0 t r ia ls , and we then asked some who had

1 88

July 20031 Experimental Tests for Telephorc Telep,th-,

completed 0 tr ia ls to do a fur ther 10 or 20 t r ia ls . Al l the exper imencs rthis series nvolved two trials per session. Most followecl Method 1, as in thc:experiment described above. With 5 of the participants we usecl Methocl :: ,which involved ixed t imes for the cal ls , known to the par t ic ipants n advBnce,rather than randomly selected imes not known to the participants n advancgas n Method 1.

Nine participants completed the prearranged number of trials, 8nd somr-agreed to do a second series of 10 trials. Two did a third series as well. A,shown in Table 2, these participants took part in a total of 146 trials. In l'(40%), heir guesses were correct. This result is highly significant, tatisticall-,(p = 0.00003).

Out of these 9 participants, all but one made more han 25%ioorrect guesses.The p values or al l the par t ic ipants ' esul ts are shown n Table 2 . As an al ternat ive way of ca lcula t ing he signif icance f the ovaral l resul t , he incl iv i t l rur lresul ts were combined by the Stouffer method. I ty th is more conservat iv, .procedure , he resul t was s t i l l very s ignif icant p = 0.001) .' fable 2

S'corea of 9 I'articipanta Wlrc Contpleted At kal,t l0 Telephone Telepathy Triale Eor/r,Following Method 1. one participant ruae male (m) and the othere female (f ).

Participant Begu n Trials R igh t % right p

PD n7JBfL PfL EfJP fS H /

MJo /MJa /G B /

1 2 . 1 . 9 93 0 . 1 r . 9 9

r8 . 1 .00

9. 2 .O0

6 . 3 . 00

9. 3 .00

4 . 9 . 00

1 . 1 . 0 0

27 . 3 . 01

l ll 0

30

l 0

l 9

30

l 0

1 0

1 6

45

L 4

3

7

7 4

6

3

3

3650

4 7

30

37

4 7

60

30

1 9

0.280.08

0.008

0 .47

0 . 1 8

0.008

o.o2o.470.80

TotaI 1 46 5 9 40 0.00003

Twelve of the or iginal 21 par t ic ipants did not complete he ini t ia l l0 t r ia lsThey withdrew for a varicty of rensons, most commonly because hey could no tpersuade ll 4 potent ia lcul lers o ngree o be nvai lable t the same imes .somccompleted nly 2 t r ia ls . Al together, hese 2 par t ic ipants ook par t in 55 tr ia ls ,and 16 out of 55 guesses were correct (29%). This restrlt is not statisticallysignificant at the p = 0.05 cvel.

I t could be arguecl hat thc resul ts rom nl l 2 l par t ic ipants shotr lc l e conr-bined. In this case, he [otal score was 75 correct guesses ut of Z0l (37o,,o),h ighly s ignif icant resul t ( r r=0.00007) . Tlre 95%-conf ic lence imits of r l r i ssuccess ate are f rom 3l% Lo44%.

r B 9



Journal of the Society for Psychical Research IVo l . 6? . s , o . 872

Of the two participants who completed 30 trials, one, LP, had 8 potentialcallers altogether, of whom only 4 took part in a given trial, depending ontheir availability (she was, of course, nformed in advance which 4 would beparticipating in each trial). This did not permit a meaningful comparison ofher success ates with different callers. But SH had the same 4 callers hrough.out, and her results with these callers are shown in Table 3. She was muchmore successful with some callers than with others, which was also the casewith PS (Table ).

Table 3SH'c Reeponlel to &lls from her Friends Gayle, Jayne and Kay, and her Mother Emma

(Method I)

Caller Calls Right % right

Emma

Gayle

Jayne

I(ay

8

I

7

7

5

2

5

2

63257 l29

Total 30 l4 47

Sam Bloomfield's DataIn November and December 1999, at RS's request, Sam Bloomfield, a

student at University College, London, carried out some telephone telepathytests following Method 1. Through an adver[isement in a local newspaper inNorth London (the Corrr en New Journal) he recruited four participants, whonominated 4 callers each. Unfortunately none of the participants was able tocomplete the prearranged series of 10 trials. Altogether, they carried out 18trials, and made 9 correct guesses 50%). This positive result is statisticallysignificant (p = 0.02).

The Second Experimental ,SeriesIn March 2001, we started a second series of experiments, in which the

participants were asked to name only two or three out of the four potential

eallers. We did this for two reasons. First, many people could not lind fourfnmiliar people nble to take pnrt in the trials. Hence t was easier to recruitparticipants when they had to find only two or three familiar people.

Second, pontaneotrs elepathic experiences with telephone calls are usuallywith very familiar people, such as mothers, best friends and spouses. Ou rhypothesis was that if participants were able to identify callers telepathically,this effect would show up with familiar callers but not with unfamiliar callers.

The unfamiliar callers were PS and CM, who lives in Ircndon, or in someeases RS. Although the participants had not met the unfamiliar callers, theyknew their names, and of course knew which 4 potential callers would betaking par t in a given tr ia l .

With the first three participants n this series, we used Method 2, in which

19 0

July 20031 Experinrentol Tests for Telephone Telepathy

there were two trials per session at fixed times. With all the other participantswe used Method 3, in wlr ich therc was only 1 l . r ia lper sess ion t a f ixet l t ime.and the cal ler was selccted t random only l5 minutes before he t r ia l .

As in the f i rs t ser ies of exper iments . some pnrt ic ipnnts were unable tocomple t e he lO t r i a l s t hcy had o r ig inn l l yag rce< l . o< lo .S ix t een pa r t i c i pan t sdid complete 10 tr ia ls , anr l sonre went on to take par t in fur ther t r ia ls ( ' l ' able4). All 16 of these larticillants made more than 25Yo orrect gucsses. n total.they were right 98 times out of 232 trials (42o/o), ith a very high statisticalsignificance (p = 7 x 10-e),

As shown in Table 4, with these 16 participants, the success ate withfamiliar callers was 75 out of 138 (54%), and with unfamiliar callers 23 out of94 (24%), a highly significant differen ce (p = 4 x 1C4).

TableScorea of I6 Participanls Who Oontplete<l At Least IO Tblephone Telepathy Triale in {ierioe 2.

There were 9 female (f) and 7 male (m) participants. The figures for the total numbere(T) of right (rl) guelles are ilrown firet, together ruilh lho p ualuen. Then the figuren forright ond wrong (wr) gucsses are slrcun separately for lamiliar (F) and unfamiliar (IJF)callere, The first 2 participants in this liel were tested by Method 2 and all others byMethod 3.

Participant Trials

EKfA P m

FL rn

SR rn

vc fE L fAC nlD L mC M /L H frAfS A /G C fE B /T M mHH nt

r2 . 3 .01

2 . 4 . O l

1 4 . 5 . 0 1

1 6 . 5 . 0 r

1 4 . 1 1 . 0 1

r 9 . 1 2 . 0 r2 6 . 1 1 . 0 1

2 8 . r 1 . 0 1

11 . 2 . 0 1

20. r.o2

20. r.o2

5. 2 .02

tL . 2 .O2

19. 2 .02

4. 3 .02

5 . 3 .02

36l 0t 71 1

1 0

l 1

10

l 4

1 6

30

L 7

1 0

1 0

1 0

l 0

1 0

l 54

7

5

3

6

4

6

I

1 1

8

3

3

5

7

3

o.o2o.220 . 1 I

o . r2o.470.03

0.22

0 . l

0 . 03

0 . 1

0 .04

0 . 4 7

o .47

0.08

0.004

o . 4 7

l 04

3

4

3

6

3

6

5

8

6

3

3

4

6

I

l 02

6

2

4

4

2

I

5

I

4

4

7

I

I

2

5

0

4

I

0

0

I

0

3

3

2

0

0

I

I

2

l l

4

4

4

3

I

4

7

3

1 l

5

3

0

4

2

5

7 x l 0 - o



lournal ol llrc .Socrcly l<tr Psyclttcul llcsearcit IVol. t i7.s , Ntt . 872

There were 21 participants who did not complete 10 trials. Altogether, theyrvere correct in 36 out of 90 trials (40%), again a very significant result (p =0.001). Again scores were higher with familiar people: 26 out of 52 correct(50%). With unfamiliar people only 10 out of 38 (26%) were correct. Thisdifference s statistically significant p = 0.02).

Taking the overall results for the second series of experiments, including 16participants who completed 0 or more trials (Table 4) and the 21 who did not,37 participants took part in 322 trials and nrade 134 correct guesses 42%).This overall result is extremely significant statistically (p = 5 x l0rtr). The95o/o-conlidence inrits of this success ate are from 36%o o 4?%. Altogether,l0l out of 190 (53%) guesses with familiar callers were correct (p = 1 x 10-16).With unfamiliar callers, only 33 out of 132 (25%) guesses were correct, exactlyat the chance level (Figure 1). This difference between the responses withfamiliar and trnfarniliar callers s highly significant (p = 3 x 10-?).

familiar

unfamiliar

37 participants EK LH

Figure I. The percentage f correct guesses n telephone elepathy rials with familiarand unfamiliar callers. he data or all 37 participants nvolued total of 190 trials.The results or two participants, EK and LH, who tooh part in 30 or more rials each,

are also shown separately. The horizontal line shows he leuel expected y chance,nantely 25'h.

Conrparisons of Tesls tuith Calls at Randonrly-Selected and at Fixed TimesIn the trials conducted by Method 1, the times at which the calls were made

were selected at random, and the participants did not know when the callswould come. This procedtrre was designed o simulate the real-life situation, inwhich people do not usually receive calls at preananged times. In the othertrials. following Methods 2 and 3, the calls were made at fixed times, known tothe participants n advance. The use of lixed times simplified the experimentalprocedure, although it rvas more artificial. In order to find out what effect hi sdifference n procedure had on the results, we compared the overall results

July 2003 j l , , x l s c r t t t r t r r l u l ' l ' e r t s l o r ' l ' e l e p h o n e ' I ' e l e p i a t i ,

from Method I with those from Methods 2 and 3 (Table 5). There was nsignif icant di fference. he success ate was almost dent ica lwi th both ,ng1[er l : ,39.9% correct with random times, and 40.3% orrect with fixed times known Ithe par t ic ipants n advance. hus there appears o be no disadvantage n us i r rfixed times.

Table 5

Comparison of Parlicipants' Success in Trials ruith Calls at llandomly-Selected Tinc,(Method 1) with Calls at Fixed Timcs Known to the I 'artieipants in Adtnnce (Ifiethods ;and 3)

Method R i g h t Wrong Totals % right

Random imes

Fixed times

79

1 4 3

l l 9

2121 9 8

35 5

39.9

40 .3

Totals 22 2 3 3 1 5 5 3 4 0 . 1

Comparison of Results in ltirst and Second Trials in Two-Trial SecsronsIn Methods 1,and 2, part icipants received two calls during each experimer: '

al session, and the callers were noti l ied before the session began. Hence th,f irst caller could conceivably have given the part icipant some clue as to t l i ,identi ty of the next caller, even though we had specifically requested callu,

not to do so. For example, when the same caller had been picked for both call , ;he or she migh t have ind ica ted tha t he or she would be r ing ing aga in soon . Oiperhaps the { i r s t ca l l e r migh t have consciously or unconsciously ind ica ted th ;someone e l se would be t lo ing the nex t call . If so, pa r t i c ipan t s would have be t : 'choos ing f rom among 3 poten t i a l ca l l e r s , r a the r than 4 , and thus the chances ctsuccess by random gt res . s ing o t r l r l have been h igher.

We compared resu l t s wi th thc f i r s t and sccond cn l l e r s in all exper imenitha t invo lved Methods 1 and 2 . fhe da ta a re shown in Table 6. If fo r an 'reason there had been a leakage of information from the frrst callers, th ,success rate on the second calls should have been higher than with the first . I ,was, indeed, sl ightly higher, 43yo as opposed to 35%. However this difference i ,not stat ist ical ly significant (by the l- isher exact test , p = 0.10)

Table 6

Compariwn of Participants'.Scores n the Fire and .Sccond rials, Followtng Methods I ar..2. The dala represenl otals rom 24 parlicipants including the preliminary experimenltuith PS} In a few eases he second rial did not ahe place, and hence here were morrfirst than second riala.

Tr ia l s R i g h t Wr ong Totals % right

First

Second

5 l

60

94

? 8

1 4 5

1 3 8

3543

Totn l s l l l r72 283 39

nffi

tctE8cc,o(uo.

t92



Journal of the Society for Psychical llesearch [Vol.67.s, No. 87 2

Only in the case of one participant, LP, was there a striking and statisticallysignificant difference (p = 0.005) between the success ate with lirst and secondcnlls. With the lirst cnlls she wns right 3 tinres out of 15, slightly below thechance level; with the second calls she was right 11 times out of 15, a 73%osuccess ate. This raises the possibility of leakage of information in thisparticular case. But if this participant 's data are excluded rom the resultssummarized in Table 2, the overall result is hardly affected: he success at eis 39% rather than Aoyo, and the result is still very significant statistically(p = 0.0007).

In any case, a possible eakage of information from first callers cannotpossibly explain the success ates on first calls themselves Table 6). A total of5l out of 145 guesses with lirst calls were correct 35%), ignificantly above hechance evel (p = 0.003).

This possible problenr did not arise with Method 3, when there was only on ecall per session. The great majority of the data in Table 4 were obtained withMethod 3; only 2 pnrticipants were tested occording o Method 2 (EK and AP).If the data from these participants are excluded, the results from the otherparticipants show the same percentage success s before (42%), and the effectsare still highly significant statistically p = I x f O-?).

TableC,omparieon of Parlicipants'Scores in Telephone Telepathy TYiale with Femiliar Callere

Orrcreeoc and in Britain. Ten of lhese participante were female (f) and one male (m).

Participant Overseascal lers in :

Overseas Britain

r ight wrong right wrong

T A f

D A f

HH nr

R H f

E L f

M M /

c M/K M /

t<P fM T /

Jwr

Nigeria

Yemen

Iceland

South Africa

Greece

South Africa

AustraliaSouth Africa

South Africa

South Africa

South Africa

4

4

I

I

2

4

43

1

2

2

I

2

I

I

2

0

3I

I

0

3

2

0

0

1

4

0

10

0

0

0

3L

1

1

1

0

34

0

I

0

TOTAI^S% right

2865

1 5 8

35

15

1 94

July 20031 Experimental Teste for Telephone Telepthy

The Effects of DistanceIn order to find out if the distance hnd any effect on the participants' ability

to identify callers, we recruited participants with friends or family membersoverseas..Table compares heir success ates with overseas allers and withfamiliar callers iving in Britain. The oversens allers were at least 1,000 milesaway, and in one case he cal ler was n Austra l ia , 1 ,000 miles away.

These participants were very successful with overseas callers, with 28correct guesses out of 43 (65%), an extremely significant result (p - 3 x fO{).With callers in the UI(, the success ate was lower (35%). n most cases, heoverseas callers werc people to whom the participants were closely bonded,such as mothers and boyfriends, whereas this was not the case with most ofthe British callers. This result implies that for the succeesful dentification ofcallers, emotional closeness as more mportant, han phyeical proximity.

DrscussroNCombining the results of all our experiments, and adding in the trials

conducted by Sam Bloomfield, there were 63 participants altogether. Theymade 231 correct guesses n 571 trials, a success ate of A0yo,well above hemean chance expectation of 25% (Table 8). The 95o/"-conlidence imits of thigresult are from 36Yo o 45%. This effect was robust and repeatable and ishugely significant statistically (p = 4 x l0-t6). Not all participants scored atlevels above chance, but the great majority did so .

Table 8

Summary of Dala from AII Telephone elepathy raile Described n thie Paper

Experiments Calls Right lo right p

Preliminary

Seriea 1

Series 2

S. Bloomlield

3020L322r8

l 376

r3 4I

43374250

o.o20.00007b x l0-t t

o.o2Totals 671 23r 40 4 x l0-r'

These results rule out the hypothesis that apparent telephone telepathy ismerely a matter of chnnce coincidence nd selective memory. Ttrey rule outthe hypothesis that it depends on unconscioue expectations about the times atwhich familiar people are likely to call. They aleo rule out normal seneorycues: in most cases he callers were miles from the participants, sometimegthoueands of miles away.

Possible ArtefoctsOne possible objection o the data presented n this paper ig that Bome of the

participants failed to complete the 10 trials they undertook to do. This raisesthe possibility of 'optional stopping', whereby participants might stop if theirresults were poor, and continue if they were good, thus creating artefactual

19 5



Journal of the Society for Psychical Research IVoI .67 . s , o . 872

positive results. In fact most participants who stopped did so because theircallers were unable or unwilling to continue. Nevertheless, there could havebeen an element of optional stopping in some cases. n the first series, the factthat the 12 participants who did not complete all 10 trials (and were thusexcluded from the summary of data in Table 2) scored on avera ge only 29Yomight support this view. But in the second series, the 21 participants wh ocompleted ewer than 10 rials had an average success ate of 4oyo, which doesnot fit with the idea of optional stopping. Nor do Sam Bloomfield's data, wherethe 4 participants had a success ate of 50%. Even if there was some optionalstopping by some of the participants in the 55 trials excluded from the firstseries of experiments, this could not explain the highly significant positiveresults rom the remaining 516 trials described n this paper.

The most serious objection o the positive results we obtained s that theremight have been a leakage of information from callere o participants throughtelephone alls hemselves, r even by e-mail.

One opportunity for a deliberate or unintentional leakage was present n thetrials conducted by Methods 1 and 2. The experimental sessions nvolved twotrials, and the callers knew in advance whether they would be involved in bothor only one of these trials. Hence callers in the first trials could have wittinglyor unwittingly conveyed hints as to whether they would or would not be callingin the second rial. In this case, he success ate in the second rials should havebeen signilicantly higher than in first trials. This was not the case, as discussedabove. Only with one participant was there a strikingly higher success atein the second trials. In any case, n the first trials, where no such leakage ofinformation could have occurred, the participants' success rates were wellabove chance.

What about deliberate cheating? Perhaps participants and their callerssimply lied about the guesses, alsely reporting incorrect guesses s correct. Orperhaps after potential callers had been nformed that they had been picked,they rang or e-mailed the participant to pass on this information. Even ifcallers who knew they had not been selected old this to the participant, thechoice would have been narrowed, and hence he chance of successful uessingincreased.

The cheating hypothesis is implausible for three main reasons. First, it isvery improbable that a large majority of the participants would have cheated. tis perhaps conceivable hat a few might have done so, but a few cheats could nothave produced the pattern of results we observed n which most participantsscored above chance evels.

Second, we know that we ourselves did not cheat, and that the unfamiliarcallers involved in Series 2 did not cheat. If some of the familiar callere hadcheated by informing participants that they had nol been selected, he chanceof guessing an unfamiliar caller would have been increased. Yet the scoreswith unfamiliar callers were not above hance evels Table a) .

Third, ss we describe n a separate paper (Sheldrake & Smart, 2003), wecarried out further series of experiments in which the participants were lilmedcontinuously on time-coded videotape, starting 15 minutes before each trial.We selected he caller at random only after the filming had started. Hence ifthe participant had received any other telephone calls or e-mails before th e

July 20031 Experimental Teste for Telephone Telep<t

tes tca l l , th is would have been observed. he videotapes were evaluated bl i r.by an independent observer, nd any tr ia ls in which the par t ic ipant receivc ,an unrela ted cal l or was off camera, owever br ief ly,were disqual i f ied . lso , lparticipants spoke their guess o the camera before picking up the telephr:r 'nncl hence could not l ie aborr t he resul ts . 'he data from these videota ;t r ia ls gave extremely signif icant posi t ive esul ts . The overal l success a te r' ;133 corrcct gt resses rr t of 292 tr ia ls (46%:p= | x l0- t2) .The 95%-conf idcn, .limits of this sr.lccess ate are from 39/o a 5l%.

One f inal poss ibi l i ty for a leakage of informat ion remains . In al l t r i ldescr ibed n the present pnper, he par t ic ipants picked up the te lephone ci , .making their guesses. t is therefore possible hat they heard characterislbackground noises , e lec t ronic hisses or o ther sounds that enabled thenridcnt i fy the cal lcr. Btr t in our f i lmed exper iments h is poss ibi l i ty was el iur iated because he participants made their guesses before they picked up {- ;te lephone. f background noises and hisses could expla in the resul ts in :,trnfilrned rials, the positive eflect we observed hould have disappeared n tl ,f i lmed tr ia ls , but i t d id not .

The videotaped r ia ls showed he same pat terns of response s the experments repor ted in th is paper, wi th scores wel l above chance with famil rcallers, with a success ate of 60% (N = LO2; = | x 10-t3). With unfamili,callers the success ate was not significantly different from chance.

TelepathyThe results of the experiments reported in this paper do not seem to b,explicable n terms of artefacts, nformation leakage or sensory clues. TheE,findings support the hypothesis of telepathy. The positive Ecores with familiarcnl lers and chance- level cores with unfamil iar cal lers Figure l ) a lso supporrth is explanat ion, s incc te lepathy typical ly takes place between people wh ,sharc socia l and enrot ional boncls , nd not wi th s t rangers Gurney, Myers ,L 'Podmore, 886; Stevenson, 970; Schouten, 982;Sheldrake, 999, 2003) .

Some people might wnnt to argue that these resul ts suppor t ESP or 1, . ,in general , but not te lepathy n par t icular. Perhaps he ant ic ipat ion of wir 'is calling before picking up the phone is gnore a matter of precognition o,clairvoyance han telepathy. But in this case t would be hard to explain wh-rprecognition or clairvoyance clid not work with unfamiliar callere, but onl 'with familiar ones. As far as we know, precognition and clairvoyance do nr)depend on socia l bonds , whereas e lepathy seems o, and hence he differenc, ,in resul ts wi th famil iar and t rnfamil iar ca l lers appears o provide a means oidis t inguishing between hese diffcrent kinds of ESP.

An explanat ion n terms of te lepathy would be consis tent wi th the fact thaithe cal ler s a lmost ncvi tably focusing his or her in tent ion on the recipient ol ^the cal l .Typical ly,when A decides o cal l B, f i rs t he or she hinks about B, of ter rin response o some need; hen A may need o look up B 's te lephone number.then A picks up t l re phone and c l ia ls he number. Al l th is t ime A's in tent ion "cl i rec ted owards B. Meanwhi le , B may star t th inking aboutA, or may have;r r rin t r r i t ion hat A is ca l l ing when the te lephone ings .

Spontaneous Ases f te lepathy cnn occur over dis tnnces f hundreds or eventhousands of rn i l e s e . g .Gurney, Myer s & Podmore , 1886 ; S t evenson , 970 ;

19 6



i irl

Journal of the Society for Psychical llesearclt IVol. i ? . l , No . l7 2

Sheldrake, 1999, 2003), suggesting hnt tclepathic influences do not ftrll offwith distance. However, most experimental research on telepathy has beenconducted ver re lnt ivcly shor t dis tnnccs , f tcn wi thin the sanrc btr i l r l ingor inadjacent bui ld ings , as in the class ic ard-guess ing tudies at Duke Univers i ty,and in most dream telepathy exper iments and ganzfeld exper imcnts [or anexcellent review, see Radin, 1997). One exception was a telepathy experimentconducted n Russia by L. L. Vasiliev with trials at distances up to 1,700 km(Braude, 979).

In experimental research on dogs hat know when their owners are cominghome, we have done ests at distances rom 5 to 45 miles, with no indication of

any decline with distance n telepathic nfluence of the owners' ntentions onthe dogs Sheldrake & Smart, 1.998, 000a, 2000b).

Using telephones, t is relatively easy to carry out telepathy experimentsover any dis tance up to a maximum of 12,500 miles , a t the ant ipodes . n ou rt r ia ls wi th overseas n l lers f rom 1,000 o 11,000 miles away, there w{ls nosuggestion hat the telepathic effect ell off with distance (Table 7), in agree-ment with previous observntions nd research.

If telephone elepathy occurs, hen why are people not right every time? Inthe experiments described n this pnper, the average ailure rate was 60%.Participants esponded more to some amiliar callers han others (Tables an d3). Some participants were more sensitive, others ess so (Tables 2 and 4). Bu teven the most sensitive participants with the most effective callers were no talways right.

The artificial nature of these tests may have weakened the influence oftelepathy. In real life, telepathy does not involve a conscious choice betweenfour equally probable callers, who have no emotional need to call. Nor does thappen on demand. Nevertheless, espite he unnatural conditions mposed bythis experimental design, elepathic communication still seems o have takenplace to a very significant extent.

ACT<NOWUEDGEMENTSWe are grateful to Drs Jan van Bolhuis, of the Free University of Amsterdam,

for his help with statistical analysis. This work was made possible by grantsfrom the Lifebridge Foundation, New York, the Institute of Noetic Sciences,California, and the Bial Foundation, Portugal.

20 Willow RoadLondon NWg ITJ

173 Kay BrowRamsbottomBury BL? 9AY

Ruppnr SHnlnnernara@dircon. co. uk

Peunr"e Suanrpam@telepet . demon. co. uk

REFERENCES

July 20031 Exlll"rinrental Tcnte for Tclcphone Telepath;

Gurney, 8., Myor.s, F. W. Il. nrrd Porlnrorc. F. (188C) Ithonlanma of the Liuine. lnndonKegan Paul, Trench, Triibncr & Co.

I- In l rn , G. '1. nnt l Meeker, W.Q. ( l$91) Sla t in l icol In leruole : A (iui t le ftrr l r rerc l i t i l tneNcw York: John Wilcy & Sons.

Radin, D. (1997) Tlrc oonscioun (Iniueree. snn l, 'rnneisc': IlnrpcrEdge.Rosenthaf, R. (1991) Meta-Analylie ltrocedures for .Socra/ Ilesearch (reuieed,

Newbury Pnrk, CA: Snge Publications.edi l ion ,

Schouten, S. A. (1982) Analysing spontaneous cnscs: n replication based on the Rhinrcollection. EJP 4, ll3-158.

Sheldrake, R. (1999) Dogs lhat Know lfihen Their Owners are Coming Home, and, OtheUnexplained Poruers of Animalt lnndon: I{utchinson.

Sheldrake, R (2000) Tclcpatlric telcphone catls: two surveys. JS4R 64, zz4-212.Sheldrake, R. (2003) The Sense of Being Stared At, and Other Aspecte of the Extende

Mind. London : Hutcl'rinson.Sholdrnko, R. nnd Snrnrt., P. ( l t )98) A dog thnt neems to know when hin owner ir

returning: prcliminnry investigntions. JSPII GZ, 22O-2X2.Slre ldrnke, R. nnd Smnrt, P. (2OOOn) A rlog t l ra t secms to know when his owncr i s comin

home: videotaped expcriments nnd obscrvatione. Journal of Se.ienlifie Exploration l4233-55.

Sheldrake, R. and Smart, P. (2000b) Testing n return.anticipating dog, Kane. Anthrozoos13, 203-2t2 .

Sheldrake, R. and Smart, P. (2003) Videotaped experiments on telephone telepathy. JI,67, tg7-246.

Siegel, S. and Castellan, N.J. (198S) Nonparantelric Statintrce for the lkhsuioral ScrenceeNew York: McGraw-Hill.

Stevenson, I. (1970) Telepathic Impresaione. Chnrtottc.svillc: University Press of Virginia.

Braude, S. (1979) ESP and Peychohinesis:Temple University Press.

Brown. D. and Sheldrake. R. (2001) TheCalifornia. JP 65, 145-156.

A Philosophical Examination. Philadephia:

anticipation of telephone calls: a survey in

19 8 19 9

Documents

SPR_Vol67 Scientific Papers on Telepathy