The first part of the series, what is known as Subseries A, was published in the monograph Extrasensory Perception written in 1934 by J.B.Rhine. This section was all that was completed at the time the monograph was written. In 1936 a brief account of the series and its total results was given in an article by J.B. Rhine in the Journal of Abnormal and Social Psychology, and in 1937 a condensed version of this article was included in the first number of the Journal of Parapsychology.
Another reason for the present undertaking is the fact that almost immediately upon publication, the Pearce-Pratt Series received special attention. It represented a methodological advance over earlier experimental work in parapsychology; and both for the laboratory group associated with the experiment and for those who were attempting to appraise and criticize the evidence for extrasensory perception, the series had to be considered.
Moreover, as new questions were raised about the series, further analysis of the data resulted. Most of these analyses were reported as they were completed, but to the student of today it would be a difficultundertaking to run them all down.
There is the further point that it is now possible to appraise the experiment and its results in the light of the developments of the intervening twenty years, the most productive period of parapsychology. It was considered an advantage to older students as well as new, therefore, for the authors to assemble for re-examination the factual matter that has accumulated around this single experimental series.
Something should be said regarding the general background of the research. First, there is the all- important aspect of personnel. It should not be forgotten that without Prof. William McDougall's appreciation of the problem and his tolerant and courageous interest in seeing it investigated under good conditions in a psychology laboratory, the experiment would not have been possible.
J.B. Rhine was at the time an assistant professor in the department of which Professor McDougall was head; it was generally understood in those days that research in parapsychology was approved by the Department. J.G. Pratt was a graduate student in psychology, especially employed as research assistant to J.B. Rhine. From the viewpoint of objectivity, it should be noted that J.G. Pratt had not at that time shown special interest in the problems of parapsychology, and in fact worked on other problems for his graduate researches. It was not until some years later that he decided to devote his other problems for his graduate researches.
The subject, Hubert E. Pearce, Jr., was at the time a student in the Divinity School at Duke. He had introduced himself to J.B. Rhine approximately eighteen months earlier and had stated that he believed he had inherited his mother's clairvoyant powers. In ESP card tests given by J.G. Pratt and J.B. Rhine during the intervening period he had exceeded the average score to be expected from chance in practically every experimental session under a wide variety of conditions.
During that period he had participated in tests involving nearly 700 runs through the standard deck of ESP cards, averaging approximately 32% successes as compared with the mean chance expectation of 20%. Nothing like this prolonged series of tests had ever been made up to that time, and Hubert Pearce's performance was recognized even then as highly exceptional.
The Distance Series was the first step involving different buildings in the separation of Hubert Pearce from the target card he was attempting to identify. The move was not so much a strictly necessary requirement for the exclusion of visual cues as it was a matter of providing a conspicuously wide margin of safety against the possibility of such cues. The use of different buildings, incidentally, was convenient for the independent recording of the subject's responses and the card sequences. It became easily possible at the same time to provide for duplicate recording and independent checking.
To those of us who had participated in the long series of earlier tests with Hubert Pearce under gradually improving conditions of test and observation, this further advance in experimental conditions was hardly required. The essential safeguards had already been approximated. advance in experimental conditions was hardly required.
The essential safeguards had already been approximated. There is, however, a tendency of the mind, when confronted with so incredible a hypothesis as that of ESP, to exaggerate the possibility of alternative factors such as visual cues, recording errors, the loss of records, and the like.
The revolutionary character of the ESP hypothesis, then, made necessary a range of precautions that were not normally considered a part of the routine of experimental psychology. This atmosphere of critical apprehension concerning the adequacy of the design needs to be taken into account, for it was part of the actual situation in which the experiment was conducted.
Some idea of the state of mind prevailing at the time can be gained from the circumstances leading to the planning of Subseries D. Subseries A, B, and C had been designed on the assumption that no error was possible that could favor in the ESP hypothesis - not unless the two designed on the assumption that no error was possible that could favor in the ESP hypothesis - not unless the two men, J.G. Pratt and Hubert Pearce were deliberately to conspire to produce a fraudulent set of results.
Wisely (and accurately) anticipating that there would be those who would find it easier to suspect collusion than to accept ESP as established, Professor McDougall recommended that J.B. Rhine identify himself with the actual performance of at least a short subseries of the distance tests in order that a theory of collusion would have to involve all three of the participants in the experiment. On the basis of this plan Subseries D was conducted with J.B. Rhine actively officiating with J.G. Pratt.
Actually the primary research objective in the experiment was to compare the effect of short and long distance on the results. In the planning of the test series, this concern with the role of distance was the essentially novel feature of the experimental design. In most of the tests in which J.G. Pratt took part during the preceding period, the target cards had been within a yard of him. It was considered a sufficient first step to introduce a distance of at least a hundred times that unit as one that should reveal any effect of distance on any possible radiant energy that conceivably intermediated in the operation of ESP.
Later in the series this distance was increased
still farther. While, then, for the general public
and the critic especially, the Pearce-Pratt Series came
into focus as a conclusive demonstration of the occurrence
of ESP, to the workers in the Parapsychology Laboratory
it became the first definite step in the testing of
the hypothesis of the non-physical nature of psi, the
hypothesis suggested by earlier experimental work
as well as by the study of spontaneous psi experiences.
A single subject, Hubert Pearce, was tested for his ability to identify ESP test cards manipulated by the experimental assistant, J.G. Pratt, in another building, part of the time at a distance of 100 yards and part of the time at a distance of more than 250 yards from the location of the subject. The experiment was designed to test for the clairvoyant type of ESP; and J.G. Pratt, accordingly, did not know the card order in the test.
Aside from planning the experiment, J.B. Rhine participated only in the independent checking of the results, except for Series D in which he participated with J.G. Pratt as the witness to the operation of the test.
There were, in all, four subseries, A, B, C, and D, totaling 74 runs through the pack of 25 cards; and the series extended from August, 1933 into March, 1934. The testing days were not consecutive, though within a given series extended from August, 1933 into March, 1934. The testing days were not consecutive, though within a given subseries they were more or less so. They were selected, however, at the mutual convenience of Hubert Pearce and J.G. Pratt. Subseries C was begun in October, 1933, and four runs were added to it in March, 1934, with Subseries D following thereafter. Specific dates may be found in Table 1.
Subseries A was done with the 100 yards distance. Subseries B at 250 yards, and the other two subseries back at 100 yards. The 74 runs represent all the ESP tests made with Hubert Pearce during this experiment under the condition of working with the subject and target cards in different buildings. It was, in fact, the only distance test involving different buildings done at the Duke Laboratory at the time.
Series A was set up with an advance commitment on termination point. It was agreed that 300 trials were to be given Hubert Pearce. The following Subseries, B, was intended to be a duplication with only the additional distance involved, but the experimenters were interested in the big shift of scoring level from day to day which was shown at the longer distance.
It was decided to allow Hubert Pearce to continue further so as to see what would happen. Subseries C was intended to be a repetition of Subseries A consisting of 300 trials designed to discover whether the lower scoring rate of Subseries B at the longer distance was a result of the altered situation or whether Hubert Pearce had declined in scoring ability. Subseries D, as has been stated, was intended as introducing a check on J.G. Pratt, and its length was agreed upon in advance (150 trials, or six runs).
In actual operation the experiment proceeded as follows, regardless of which subseries was involved: At the time agreed upon, Hubert Pearce visited J.G. Pratt in his research room on the top floor of what is now the Social Science Building on the main Duke campus. The two men synchronized their watch and set an exact time for starting the test, allowing enough time for Hubert Pearce to cross the quadrangle to the Duke Library where he occupied a cubicle in the stacks at the back of the building. From his window J.G. Pratt could see Hubert Pearce enter the Library.
J.G. Pratt then selected a pack of ESP cards from several packs always available in the room. He gave this pack of cards a number of dovetail shuffles and a final cut, keeping them face-down throughout. He then placed the pack on the right-hand side of the table at which he was sitting. In the center of the table was a closed book on which it had been agreed with Hubert Pearce that the card for each trial would be placed.
At the minute set for starting the test, J.G. Pratt lifted the top card from the inverted deck, placed it face-down on the book, and allowed it to remain there for approximately a full minute. At the beginning of the next minute this card was picked up with the left hand and laid, still face-down, on the left-hand side of the table, while with the right hand J.G. Pratt picked up the next card and put it on the book.
At the end of the second minute, this card was placed on top of the one on the left and the next one was put on the book. In this way, at the rate of one card per minute, the entire pack of 25 cards went through the process of being isolated, one card at a time, on the book in the center of the table, where it was the target or stimulus object for that ESP trial.
In his cubicle in the Library, Hubert Pearce attempted to identify the target cards, minute by minute, and recorded his responses in pencil. At the end of the run, there was on most test days a rest period of five minutes before a second run followed in exactlythe same way. Hubert Pearce made a duplicate of his call record, signed one copy, and sealed it in an envelope for J.B. Rhine.
Over in his room J.G. Pratt recorded the card order for the two decks used in the test as soon as the second run was finished. This record, too, was in duplicate, one copy of which was signed and sealed in an envelope for J.B. Rhine. The two sealed records were delivered personally to J.B. Rhine, most of the time before J.G. Pratt and Hubert Pearce compared their records and scored the number of successes.
On the few occasions when J.G. Pratt and Hubert Pearce met and compared their unsealed duplicates before both of them had delivered their sealed records to J.B. Rhine, the data could not have been changed without collusion, as J.G. Pratt kept the result from the unsealed records and any discrepancy between them and J.B. Rhine's results would have been noticed. In Subseries D, J.B. Rhine was on hand to receive the duplicates as the two other men met immediately after each session for the checkup.
Thus, from day to day as the experiment proceeded, Hubert Pearce was kept informed, as he had been in all his earlier experiments, as to the rate of success achieved. The practice of expressing enthusiastic congratulations should be mentioned as a part of the procedure. If, as rarely happened, the scoring rate was low, favorable emphasis was placed on the overall performance, the general average maintained, and the high standing of the subject in the comparative scale of ESP subjects. Throughout the series the paramount objective of high-order performance was held before the subject with all the vigor and expectation that could be communicated.
Since they were one series of tests carried out under essentially the same conditions, the four subseries (totaling 74 runs, or 1850 trials) may be pooled. Mean chance expectation is 20%, or 370 hits. The total number of successes actually scored for the series is 558, which is better than 30%. The theoretical standard deviation derived on a conservative basis is 17.57. This total of 558 hits is 188 above the theoretical expectation and it gives a crit ical ratio of 10.70. The probability that a critical ratio so large as this would occur on the basis of random sampling is less than 10-22.
In the determination of the critical ratio given
above, allowance is made for the slight correction
applicable when, as in this experiment, the balanced ESP is
used; thatis, when there are five of each symbol in each pack.
The variance of scores obtained with the 5 X 5 ESP deck
depends upon the frequency with which the subject calls the
different symbols. The largest variance results when the subject
always calls exactly five of each symbol, and the standard
deviation of 17.57 was obtained on this assumption.
However, the subject rarely called five of each symbol in a
run, and the exact standard deviation would therefore be smaller
than the one used here, which makes the estimate of statistical
significance a conservative one.
However, the subject rarely called five of each symbol in a run, and the exact standard deviation would therefore be smaller than the one used here, which makes the estimate of statistical significance a conservative one.