EFFECTS ON RECALL OF HYPNOTIC INDUCTION, MOTIVATIONAL

Journal of Abnormal Psychology1966, Vol. 71, No. 3, 169-180

EFFECTS ON RECALL OF HYPNOTIC INDUCTION,MOTIVATIONAL SUGGESTIONS, AND

SUGGESTED REGRESSION:

A METHODOLOGICAL AND EXPERIMENTAL ANALYSIS 1

THEODORE XENOPHON BARBER AND DAVID SMITH CALVERLEY

Medfield Foundation, Harding, Massachusetts

90 Ss were assessed for recall of syllables (learned 2 mo. previously) under1 of 9 experimental conditions in a 3 X 3 factorial design. The 3 levels of the1st independent variable were: hypnotic induction, eyes closed; no induction,eyes closed; and no induction, eyes open. The 3 levels of the 2nd independentvariable were: recall without suggestions; recall with motivational suggestions;and recall with suggestions to regress to the time of original learning. Recallwas not affected by the independent variables or their interaction. A com-parable number of Ss in each of the 9 experimental groups testified post-experimentally that, during the experiment, they had imagined, felt, and be-lieved they were back in the original learning situation.

A series of experiments has ostensiblydemonstrated that recall of remote memoriesor of previously learned material can befacilitated by first administering a hypnoticinduction procedure and then administeringeither (a) motivational suggestions for height-ened recall or (b) suggestions to regress tothe time of original learning. If these findingsare valid, hypnotic procedures which includemotivational suggestions or regression sug-gestions should prove useful in various situa-tions, such as in psychotherapy, in producingheightened recall of important life-historymaterial. However, the findings cannot asyet be accepted as defintive; all pertinentexperiments conducted up to this time areopen to one or more serious methodologicalobjections. The present paper first specifiesmethodological problems relevant to researchin this area and then presents an experimentthat was designed to avoid inadequacies inmethod.

METHODOLOGICAL CONSIDERATIONSIn this section are discussed, in consecutive order,

methodological problems pertaining to assessmentof the effects on recall of (a) motivational sug-

JThis research was supported by a grant (MH-07003) from the National Institute of MentalHealth, United States Public Health Service. Weare indebted to Frederick J. Ryan of WorcesterJunior College for assistance in the conduct of theexperiment.

gestions given under hypnotic conditions and of (6)hypnotic age-regression suggestions. Following this,consideration is given general problems in methodpertaining to (c) control of experimenter bias and(d) control of bias in selection of subjects.

Motivational suggestions for heightened recall.Stalnaker and Riddle (1932) and Hammer (1954)found that suggestions of high motivation and in-creased ability given subsequent to a hypnotic induc-tion were effective, as compared to a nonsuggestionwaking condition, in facilitating recall of previouslylearned material. However, the possibility was notconsidered that subjects might have performed asproficiently under the waking condition as underthe hypnotic condition if motivational suggestionsfor enhanced recall had been given under both con-ditions. Since the investigations confounded twoindependent variables, namely, (a) hypnotic induc-tion (or the "hypnotic trance state") and (b) moti-vational suggestions intended to elicit a high levelof performance, no conclusions are possible concern-ing the effects of hypnotic induction alone, or ofmotivational suggestions alone, or of the interactionof hypnotic induction with motivational suggestions.

White, Fox, and Harris (1940) reported that sug-gestions for enhanced recall were more effectivewhen given under hypnotic, rather than underwaking conditions, in improving recall of poeticmaterial, but not of nonsense syllables or of motionpicture scenes. In a series of studies, Rosenthal(1944) found that motivational suggestions for im-proved recall of meaningful and nonsense material attimes were more effective and at times were notmore effective when given under hypnotic, ratherthan under waking conditions.

In the White et al. and Rosenthal studies thesubjects were apparently tested for recall with theireyes open under the waking treatment and withtheir eyes closed under the hypnotic treatment.

169

170 THEODORE X. BARBER AND DAVID S. CALVERLEY

Since recall may be more proficient when the eyesare dosed, the possibility was not excluded that theheightened recall that was at times obtained underthe hypnotic treatment was due, not to hypnoticinduction (or "hypnotic trance"), but to the removalof distractions which occurs when the eyes are closed(cf. Barber, 196Sa).

It follows from the above considerations thatfurther research is needed to assess the effects onrecall of the following variables independently: hyp-notic induction; motivational suggestions for height-ened recall; and recall with eyes closed. In thepresent experiment the effects of these variables wereevaluated separately and in all possible combinations.

Age-regression suggestions. Several studies (e.g.,As, 1962; True, 1949), reviewed in detail elsewhere(Barber, 1962b), indicate that recall can be facili-tated by administering a hypnotic induction and thensuggesting to the subject that he regress to the timeof original learning. However, each of these studieslacked an important control; in no instance weresuggestions to regress given to awake subjects. Sincethe regression suggestions were always confoundedwith the hypnotic induction, no conclusions can bedrawn concerning the effects on recall of hypnoticinduction alone, or of regression suggestions alone,or of the interaction of hypnotic induction withregression suggestions. Clearly, further research isneeded to assess the effects on recall of regressionsuggestions given both with and without a precedinghypnotic induction. The present experiment wasdesigned to meet this need.

Control of experimenter bias. There is evidence toindicate that experimenters find it difficult to giveidentically worded suggestions to hypnotic andwaking subjects in the same tone of voice, and thatthe biased tone of voice in which the suggestionsare presented effects the subjects' responses (Barber& Calverley, 1964b; Troffer & Tart, 1964). Tocontrol this factor in our experiment, the suggestionswere administered in exactly the same way to hyp-notic and waking subjects by means of a taperecording of the experimenter's voice (Barber &Calverley, 1964a).

Control of bias in selection of subjects. In someof the investigations in this area, subjects allocatedto the hypnotic treatment, but not to the wakingcontrol treatment, were preselected for "hypnotiza-bility." This selection procedure is open to twoserious objections, as follows:

1. To certify that subjects allocated to the hyp-notic treatment met criteria for "hypnotizability,"preliminary assessment or "training" sessions wereconducted with them. Discrepancies in performanceduring the critical session may thus have been con-tingent, not upon the experimental group being in"hypnosis" and the control group being in the"awake state," but upon the experimental grouphaving participated in the preliminary sessions andhaving acquired useful experience. For instance, inthe preliminary sessions, the experimental group mayhave formed a friendly relationship with the experi-menter and may have become familiar with the

experimental setting. These, and possibly other fac-tors carried over from the assessment or "training"sessions to the experiment proper, could have thegeneral effect of making the experimental subjectsmore at ease or less anxious than the controls andof enhancing their motivation and capability to givea maximal performance on the tests for recall.

2. The explicit criterion for selection of the experi-mental group—that the subjects were "hypnotizable"—appears difficult to differentiate from the inter-related implicit criterion that the subjects werehighly responsive to suggestions with or without"hypnotic induction." By thus confounding the ex-perimental conditions with preexisting differencesamong subjects with respect to "hypnotizability" or"suggestibility," it could not be ascertained whetherthe experimental subjects were more responsive thanthe controls to suggestions for heightened recallbecause they received the suggestions when theywere in "hypnotic trance," or because they weremore responsive to suggestions with or without"hypnosis."

In other experiments in this area, selected "good"hypnotic subjects were used as "their own controls,"that is, were tested under both the hypnotic andwaking conditions. This procedure is open to thefollowing criticism: there is evidence to indicate thatsome "good" hypnotic subjects surmise that theexperimenter expects or wants better results underthe hypnotic condition and, desiring to please theexperimenter, purposely give their best possible per-formance under the hypnotic condition and pur-posely give an inferior performance under thewaking condition (Barber, 1962a; Pattie, 1935;Schrenck-Notzing, 1896; Sutcliffe, 1960; Wolberg,1948, p. 49).

The above considerations suggest that furtherstudies should allocate different subjects to thehypnotic and waking treatments and should avoidconfounding the hypnotic treatment with such vari-ables as prior "training" and preexisting high sug-gestibility. In accordance with these considerations,in the present experiment, unselected and untrainedsubjects from a homogeneous population wereassigned at random to the treatments.

METHOD

Experimental Design

All subjects first learned a list of syllables. Twomonths later they were tested individually for recallof the syllables under one of nine experimental con-ditions. The nine experimental conditions constitutedthe cells of a 3 X 3 factorial as illustrated in Table 1.The first independent variable (Experimental Treat-ments) was at three levels: hypnotic induction, eyesclosed; no hypnotic induction, eyes closed; andno hypnotic induction, eyes open. The second inde-pendent variable (Conditions of Recall) was alsoat three levels: recall without suggestions; recallwith motivational suggestions; and recall withregression suggestions.

RECALL AND HYPNOTIC INDUCTION 171

Subjects

Subjects were 90 freshmen and sophomores (77males and 13 females) at Worcester Junior Collegewho were enrolled in the introductory psychologycourse taught by Frederick J. Ryan. The 90 studentswere randomly assigned to the nine experimentalgroups with 10 to each group.

Original Learning

When assembled in classes during the first weekof October 1964, the students were given 12 syl-lables to learn. The syllables were introduced asfollows:

The instructor in the class (F. J, Ryan) had beendiscussing the history of psychology and during thisparticular class session lectured on Ebbinghaus andthe learning of nonsense material. At the conclusionof the lecture the students were told that, to betterunderstand what was involved in memorizing non-sense syllables, they should try to learn a list them-selves. A 22 X 8-inch white cardboard containing 12syllables was then affixed on the front wall. Thestudents were told they would have S minutes tolearn the list. The list was removed after S minutesand the students were asked to write the syllableson a sheet of paper and to sign their name. After3 minutes elapsed, the papers were collected. On theaverage, the students wrote 10 syllables of which 9were the same as those on the original list.2

The 12 syllables were selected in accordance withthe following considerations. It was desired that thematerial should be standardized, should consist ofequal units for convenience of scoring, and shouldbe of equal difficulty for all subjects. Standardizedlists of nonsense syllables as compiled by Glaze(1928) best meet these criteria. However, since itwas also desired that the material should be as mean-ingful as possible, the 12 syllables were randomlyselected from among those that Glaze had found tobe of 93% association value (Hilgard, 1951, p. 541).These syllables were: KAF, PTJE, HEM, PAV, LEC, NOM,LOY, YAC, FIV, TIE, MUR, and LUN.

Introduction to Experimental Sessions

Two months after the original learning, the stu-dents were informed by the instructor that in orderto fulfill the course requirements, they would par-ticipate in an experiment. They were further toldthat: by participating in the experiment they wouldincrease their understanding of psychology and whatit is like to be a subject in a psychological investiga-tion; the experiment would begin the following day;it would be conducted by a former instructor atthe college (D. S. Calverley) ; some subjects would

be hypnotized and others would serve in a controlgroup and would not be hypnotized; the experimentwould be explained to them after it was completed;and, during the 10-day period the experiment wasin progress, they were not to discuss it with eachother. The instructor concluded by carefully explain-ing to the class why the experiment would be"ruined" if the students talked about it. (At theend of each subject's test-session, the subject wasagain admonished, this time by the experimenter, notto "ruin the experiment" by discussing it with theother students. Each subject stated that he under-stood the importance of this admonition andpromised to adhere to it.)

Experimental Treatments

On each of the 10 experimental days, one subjectwas tested from each experimental group. Each sub-ject was tested individually by one experimenter(DSC) under one of three Experimental Treatments,and also under one of three Conditions of Recall,as illustrated in Table 1. The three experimentaltreatments were as follows:

Hypnotic induction, eyes closed. Subjects allocatedto Groups 1, 2, and 3 were first told:

"You have been assigned to the hypnosis group andyou will be deeply hypnotized. The instructions forthis experiment are on this recording. Listen care-fully to the recording and follow the instructions."

A standardized 10-minute hypnotic induction pro-cedure was then administered by a tape recordingof the experimenter's voice. This induction pro-cedure, which was patterned closely after the hyp-notic procedures of Marcuse (1959, pp. 52-53) andWeitzenhoffer and Hilgard (1959, pp. 13-18), in-cluded instructions designed to produce positive rap-port and positive motivation to perform well in thehypnotic situation, suggestions of eye-heaviness andeye-closure, repeated suggestions of relaxation,drowsiness, and sleep, suggestions that the subjectwas entering a deep hypnotic trance, and suggestionsthat he would now be able to respond easily tofurther suggestions.8

Upon completion of the hypnotic induction pro-cedure, Groups 1, 2, and 3 were told "Keep youreyes closed" and were then tested immediately forrecall of the previously learned syllables underone of three Conditions of Recall, as describedsubsequently.

No induction, eyes closed. Subjects allocated toGroups 4, S, and 6 were told to keep their eyesclosed during the experiment but were not exposed

2 The reader should note that the students werenot told that they would again be asked to recallthe nonsense syllables at a later time. On the con-trary, the impression was given that, after they hadturned in their list at the end of the class session,the nonsense syllables were a closed matter.

8 A verbatim account of the hypnotic inductionprocedure may be obtained without charge bywriting to the authors or for a fee from the Amer-ican Documentation Institute. Order Document No.8320 from ADI Auxiliary Publications Project,Photoduplication Service, Library of Congress,Washington, D. C. Remit in advance $1.25 forphotocopies or 35-mm. microfilm and make checkspayable to: Chief, Photoduplication Service, Libraryof Congress.


to a hypnotic induction. Instructions for thesegroups were:

You have been assigned to the waking controlgroup and you will not be hypnotized. Theinstructions for this experiment are on this record-ing. I want you to close your eyes and to keepthem closed during the experiment. Now closeyour eyes and keep them closed and listen care-fully to the recording and follow the instructions.

Groups 4, 5, and 6 were then tested immediatelyfor recall of the previously learned syllables underone of three Conditions of Recall.

No induction, eyes open. Subjects assigned toGroups 1, 8, and 9 were not exposed to a hypnoticinduction and were not asked to close their eyes.Instructions were:

You have been assigned to the waking controlgroup and you will not be hypnotized. The in-structions for this experiment are on this record-ing. Listen carefully to the recording and followthe instructions.

Groups 7, 8, and 9 were then tested immediatelyunder one of three Conditions of Recall, as describednext.

Conditions of Recall

Recall without suggestions. Group 1 (Hypnoticinduction, eyes closed), Group 4 (No induction, eyesclosed), and Group 7 (No induction, eyes open)were asked to recall the previously learned materialwithout receiving motivational suggestions or sug-gestions to regress. The instructions, which wereadministered by a tape recording, were:

Now I want you to listen very carefully towhat I say. At the beginning of the semester,during the first week in October, Mr. Ryanshowed you a large card with nonsense syllableson it and he asked you to learn the nonsense syl-lables. Now, during the next 3 minutes, Iwant you to recall as many of the nonsensesyllables as you can. Spell out the nonsensesyllables aloud, one by one. Start now.

The subjects were given 3 minutes for recall.Recall with, motivational suggestions. Subjects in

Group 2 (Hypnotic induction, eyes closed), Group S(No induction, eyes closed), and Group 8 (No induc-tion, eyes open) were given 3 minutes to recallthe nonsense syllables after receiving the followingmotivational suggestions for heightened recall bymeans of a tape recording:

Now I want you to listen very carefully towhat I say. At the beginning of the semester,during the first week in October, Mr. Ryanshowed you a large card with nonsense syllableson it and he asked you to learn the nonsensesyllables. If you really try very hard to rememberthe nonsense syllables, you will be able to recallthem. Even though you may feel that they aredifficult to remember, you'll be surprised how

many of them you will be able to recall if youreally try hard to recall as many of the nonsensesyllables as you can. Try very hard to recall them.Don't assume that you can't remember them.Other students were able to do surprisingly wellon this test. You too can remember them verywell if you really try. Now, during the next3 minutes, I want you to recall as many of thenonsense syllables as you can. Spell out the non-sense syllables aloud, one by one. Start now.

Recall with regression suggestions. Subjects inGroup 3 (Hypnotic induction, eyes closed), Group 6(No induction, eyes closed), and Group 9 (No induc-tion, eyes open) were given suggestions to regress tothe time when they originally learned the nonsensesyllables (2 months previously). The suggestions,given by means of a tape recording, were as follows:

Now I want you to listen very carefully towhat I say. I want you to imagine vividly thatyou are going back in time to the beginning ofthe semester, during the first week in October,when Mr. Ryan showed you a large card withnonsense syllables on it and he asked you to learnthe nonsense syllables. Time is now going back-wards. The days are going backward. It is nowNovember 10, 1964. The days are going backward.It is now November 1, 1964. You are going backin time. It is now October 20th. Time is goingback. It is now the second week in October. Youare going back in time. It is now the first week inOctober. It is the first week in October and youare in Mr. Ryan's class. Mr. Ryan has given you alist of nonsense syllables to learn. It is the firstweek in October, you are in Mr. Ryan's class, andyou have just learned the nonsense syllables. Nowanswer these questions aloud. Answer these ques-tions aloud, clearly, and distinctly.

In whose class are you? (10-second pause)What is the month and the week? (10-second

pause)What have you just learned? (10-second pause)Now spell out the nonsense syllables that you

have just learned. During the next 3 minutesspell out aloud each of the nonsense syllables thatyou have learned. Spell them out aloud one byone. Start now.

Immediately following the 3-minute recall period,Groups 3, 6, and 9 were told: "Now you are backto the present. You are now back to the present."

Assessment of Hypnotic Subjects' Responseto Barber Suggestibility Scale

Upon completing the above, subjects in Groups4-9 (No induction) were told that the experimentwas over. Subjects in Groups 1-3 (Hypnotic induc-tion) were given additional suggestions of relaxation,drowsiness, and sleep for 2 minutes, then assessed onresponse to the Barber Suggestibility Scale (Barber,196Sc; Barber & Calverley, 1963; Barber, Karacan,& Calverley, 1964) and, finally, told that the experi-ment was over.


Postexperimental Subjective Reports (Imag-ined, Felt, and Believed)

Immediately after it was stated that the experi-ment was over, all 90 subjects were given a 3-itemdittoed questionnaire to answer, which was wordedas follows:

Please answer the following questions by placinga check mark (V) next to the appropriate answer.

1. During the experiment, when you were askedto recall the nonsense syllables, to what extentdid you imagine that it was the first week inOctober, that you were in Mr. Ryan's class, andthat you had just learned the nonsense syllables:

[A][B][C][D]

-I thoroughly imagined this.-I imagined this to a great extent.-I imagined this to some extent.-I did not imagine this.

Item 2 was identical with Item 1 with the excep-tion that the word feel (or felt) was substituted forthe word imagine (or imagined).

Item 3 was also identical with Item 1 with theexception that the word believed was substitutedfor the word imagined.

Each of the three items (imagined, felt, andbelieved) was scored as follows: Weights of 3, 2, 1,and 0 were assigned to A, B, C, and D answers,respectively.

RESULTS AND DISCUSSION

First the authors shall present the resultspertaining to the primary question posed inthis experiment, namely, what were the ef-fects on recall of the two major independentvariables (Experimental Treatments andConditions of Recall)? Next, we shall turn

to three supplementary questions: (a) Weresuggestions to regress effective in producingregression as indicated by subjects' answersto questions concerning where they were inplace and time and what they were doing?(b) What were the effects of the independentvariables on postexperimental subjective re-ports? (c) With respect to Groups 1-3, didthe better hypnotic subjects (those obtaininghigher scores on the Barber SuggestibilityScale) differ from the poorer hypnotic sub-jects in recall, in degree of regression, or inpostexperimental subjective reports?

Effects of Independent Variables on Recall

Each of the nonsense syllables proffered bythe subject during the experimental sessionwas scored as correct if it was the same asone of the nonsense syllables that the subjecthad written down during the original learningsession 2 months previously. The mean recallscores (average number of correct nonsensesyllables) for each experimental group arepresented in Table 1 together with an analy-sis of variance of the scores. This table showsthat all experimental groups correctly re-called an average of less than one nonsensesyllable. The number of syllables correctlyrecalled did not differ significantly among(a) subjects who had received a hypnoticinduction and whose eyes were closed, (b)subjects who had not received an inductionand whose eyes were closed, and (c) sub-

TABLE 1MEAN NUMBER OF NONSENSE SYLLABLES CORRECTLY RECALLED (RECALL SCORES)

AND ANALYSIS OF VARIANCE OF RECALL SCORES

Conditions of recallRecall without suggestionsRecall with motivational suggestionsRecall with regression suggestions

Average

Source of variance

Experimental treatments (A)Conditions of recall (B)A X BErrorTotal

Experimental treatments

Hypnotic induction,eyes closed

.4

.2

.5

.4

(Group 1)(Group 2)(Group 3)

No induction,eyes closed

.9 (Group 4)

.1 (Group 5)

.4 (Group 6)

df

224

8189

No induction,eyes open

.1 (Group 7)

.7 (Group 8)

.2 (Group 9)

.3

MS

.14

.141.30.58

Average

.5

.3

.4

F

2.24


jects who had not received an inductionand whose eyes were open (nonsignificantmain effect for Experimental Treatments).Furthermore, the recall scores did not differsignificantly among (a) subjects who recalledwithout suggestions, (b) subjects who re-called with motivational suggestions, and (c)subjects who recalled with regression sug-gestions (nonsignificant main effect for Con-ditions of Recall). The interaction of theindependent variables (Experimental Treat-ments with Conditions of Recall) was alsononsignificant. These nonsignificant effects re-flected the following: The overwhelming ma-jority of subjects (88%) correctly recalledeither none or one of the nonsense syllables;the 11 subjects (12%) who recalled eithertwo or three of the syllables were distributedmore or less evenly over the nine experi-mental groups. (No subject correctly recalledmore than three syllables.)4

4 A series of additional statistical analyses werealso performed on the Recall scores. These includedthe following:

1. Two analyses of covariance of the scores pre-sented in Table 1. In the first covariance analysis,the covariate was the total number of syllables thatthe subject had written down during the originallearning session 2 months previously. In the secondcovariance analysis, the covariate was the number ofsyllables that the subject had recalled correctly dur-ing the original learning session.

2. An analysis of variance and two analyses of Co-variance of the number of nonsense syllables givenby the subjects during the experimental session whichwere either correct or two-thirds correct. In theseanalyses, a syllable received a score of 1 if at least2 of the 3 letters were identical with and in thesame positions as the letters in one of the syllablesrecalled by the subject during the original learningsession. (In the covariance analyses, the covariateswere the same as in Paragraph 1 of this footnote.)

These analyses yielded the same results as theanalysis presented in the body of this paper. That is,the main effects and the interaction of the inde-pendent variables did not significantly affect recallof the previously learned material. These nonsignifi-cant effects reflected the fact that, irrespective of thescoring method that was used, the overwhelming ma-jority of subjects gave no more than two correctsyllables and the few subjects that scored above 2were distributed more or less evenly over the nineexperimental groups.

To ascertain whether the independent variables af-fected the degree to which the subjects tried to givenonsense syllables (Attempt scores), an additionalanalysis of variance was performed on the totalnumber of nonsense syllables elicited from the sub-

These results are in harmony with recentexperiments presented by Lyon-James (1957)and by Leonard (1963). In the Lyon-Jamesexperiment, 12 selected, good hypnotic sub-jects were assessed for recall of meaningfulmaterial under hypnotic conditions, and anadditional 12 selected, good hypnotic sub-jects were assessed for recall under wakingconditions, with motivational suggestions forhigh recall given under both conditions. Thetests for recall were given immediately afterthe subjects had learned the material, after 3weeks, and again after 9 weeks. The hypnoticand waking groups did not differ significantlyin either immediate or delayed recall of thematerial. In the Leonard experiment, fivehighly selected, good hypnotic subjects whowere regressed to the time when they hadoriginally learned a list of paired associates(2 weeks previously) did not relearn theassociates more proficiently than 10 unse-lected nonhypnotic subjects who were notgiven suggestions to regress to the time oforiginal learning.

The results obtained in our experiment,and in the recent experiments of Lyon-Jamesand of Leonard, do not exclude the possibilitythat hypnotic induction, or eyes closed, ormotivational suggestions, or regression sug-gestions, or some combinations of these varia-bles may at times facilitate recall. As noted inthe introduction to the present paper, severalstudies (As, 1962; Rosenthal, 1944; Whiteet al., 1940) found that recall was at timesfacilitated by motivational suggestions orregression suggestions given subsequent to ahypnotic induction procedure. The latterstudies differed from each other, and alsofrom our study and from the studies of Lyon-James and of Leonard, in the methods used toselect subjects, in the wording of instructionsand suggestions, and in many other methodo-logical aspects. Clearly, further research isneeded to determine which of the many pro-cedural differences among these investiga-tions were important and which were ex-traneous in producing both the positive andthe negative results. Future studies, how-

jects during the 3-minute experimental recallperiod, irrespective of whether the syllables werecorrect or incorrect. This analysis also yielded non-significant F ratios.


TABLE 2

NUMBER (AND PERCENTAGE) OF SUBJECTS PASSING REGRESSION QUESTIONS

Group 3 (Hypnotic induction, eyes closed)Group 6 (No induction, eyes closed)Group 9 (No induction, eyes open)

Question

1. In whose classare you?

10 (100%)9 (90%)9 (90%)

2. What is the monthand the week?

10 (100%)6 (60%)9 (90%)

3. What have youjust learned?

10 (100%)10 (100%)8 (80%)

1-3

10 (100%)6 (60%)8 (80%)

ever, should take into consideration the ex-perimental controls discussed in the methodo-logical introduction to the present paper. Forinstance, since there is evidence to indicatethat experimenters may bias the outcome byvarying the tone of voice in which they ad-minister suggestions to "hypnotized" and"awake" subjects (Barber & Calverley, 1964b;Troffer & Tart, 1964), further researchshould consider controlling experimenter biasin the same way as in the present experiment,namely, by using a tape recording to ad-minister instructions and suggestions.

Effectiveness of Suggestions to Regress

Group 3 (Hypnotic induction, eyes closed),Group 6 (No induction, eyes closed), andGroup 9 (No induction, eyes open) weregiven suggestions to regress to the time (2months previously) when they had originallylearned the nonsense syllables. Following thesuggestions to regress, the subjects in thesegroups were asked three questions: (a) Inwhose class are you?; (b) What is the monthand the week?; and (c) What have you justlearned? These questions provided three inter-related criteria for regression, as follows: Inreply to the first, second, and third questions,respectively, the subject who is regressedstates that (a) he is in Mr. Ryan's class, (b)it is the first week in October, and (c) he hasjust learned a list of nonsense syllables. Werethe suggestions to regress effective in produc-ing regression as indicated by these criteria?Table 2 presents the pertinent data. The finalcolumn in this table shows that 100%, 60%,and 80% of the subjects in Groups 3, 6, and9, respectively, answered each of the threequestions as if they were regressed. A chi-square test showed that the differences be-tween the groups in the number of subjects

passing each of the three criteria for regres-sion fell short of significance at the 5%level of confidence. The import of these ex-perimental results is as follows:

1. Sixty percent and 80% of the subjectsin the two groups that did not receive a hyp-notic induction met each of the three criteriafor regression (Table 2, final column). Thisoutcome, indicating that suggestions to regressare effective in producing regression in themajority of "waking" subjects, does not con-tradict any previous study known to us. Al-though in several previous studies nonhyp-notic subjects were asked to simulate or topretend they were regressed, in no prior in-vestigations were nonhypnotic subjects giventhe same suggestions to regress as were givento the hypnotic subjects (see Barber, 1961,1962b, for a review of the literature).

2. It needs to be emphasized that all priorexperiments failed to give regression sug-gestions to nonhypnotic subjects and, further,that all prior experiments included, among thecriteria for regression, statements from thehypnotic subjects that they were in anotherplace and at a previous time. Clearly, no con-clusions concerning the effectiveness of hyp-notic induction (or "hypnosis") in producingacceptance of suggestions to regress can bedrawn from any previous experiment until itis replicated with the addition of a nonhyp-notic group that is given identical suggestionsto regress and is assessed on the same criteriafor regression as the hypnotic group.

3. There was a strong trend in the presentexperiment for regression suggestions to bemore effective with subjects who had receiveda hypnotic induction. This trend could havebeen produced by any one or a combinationof the many variables that were included in

176 THEODOKE X. BAKBER AND DAVID S. CALVERLEY

TABLE 3

MEAN SCORES AND ANALYSIS OP VARIANCE or POSTEXPERIMENTAL SUBJECTIVE REPORTS

Conditions of recallRecall without suggestionsRecall with motivational suggestionsRecall with regression suggestions

Average

Source of variance

Experimental treatments (A)Conditions o{ recall (B)A X BErrorTotal

Experimental treatments

Hypnotic inductioneyes closed

1.0 (0.6)1.3 (0.5)1.3 (1.1)1.2 (0.7)

df

224

8189

"0.51"0.5'|0.6 |

_0.5J

No induction,eyes closed

1.2 (1.0)1.2 (1.0)1.3 (0.8)1.2 (0.9)

"0.8"0.90.4O.VJ

No induction,eyes open

1.8 (1.2)1.8 (1.4)1.2 (1.4)1.6 (1.3)

MS

1.34 (2.21 (

80) [1.21]23)

.57 (.53)

.86 (.92)

[.84-[.46][.91]

"0.8"1.3'0.70.9J

Average

1.3 (0.9) [0.7'1.4 (1.0) [0.9"1.3 (1.1) [0.6"

F

1.55 (3.04) [1.32]

Note.—Scores on the Imagined item are listed first, scores on the Felt item are listed next in parenthesis, and scores on theBelieved item are listed last in brackets.

the hypnotic induction procedure, for in-stance, by denning the situation to the sub-ject explicitly as "hypnosis" and implicitly asa situation in which high response to sug-gestions was expected, or by suggesting tothe subject that he would find it easy to re-spond to further suggestions (cf. Barber,1965b; Barber & Calverley, 1964c). Furtherstudies are needed to determine whether,with larger subject samples, suggestions toregress are significantly more effective whengiven with, rather than without a precedinghypnotic induction. If a significant effect isfound, further research would then be neces-sary to ascertain which of the many varia-bles included in a hypnotic induction pro-cedure are instrumental and which extraneousto producing the effect.

An additional question needs to be askedwith respect to the present experiment,namely, was regression related to recall? Toanswer this question, product-moment corre-lations were computed separately for Groups3, 6, and 9 between their Recall scores andtheir Regression scores. (The Regressionscores were computed by assigning 1 pointfor each of the three questions—for example,What is the month and the week?—that thesubject answered in the regressed manner.)The correlations were .00, -.68 (p<.0$),and .22, respectively, for Groups 3, 6, and 9.These coefficients indicate that, with respect

to Groups 3 and 9, Regression scores werenot related to Recall scores and that, withrespect to Group 6, the more the subjectsmet criteria for regression, the fewer the num-ber of nonsense syllables they correctly re-called. Stated in more general terms, thesecorrelations indicate that subjects who meetcriteria for regression, that is, who state thatit is now a previous time, do not manifestenhanced recall for material they had learnedat the previous time.

Postexperimental Subjective Reports

After it was stated that the experiment wasover, all subjects were given a questionnaireto answer concerning the extent to which,when they were recalling the nonsense sylla-bles during the experiment, they had (a)imagined, (b) felt, and (c) believed that theywere back at the time when they had origi-nally learned the syllables. It was expectedthat subjects given suggestions to regresswould score markedly higher on each of thesethree questions than subjects not given suchsuggestions. The mean scores on the threepostexperimental questions, together withanalyses of variance of the scores, are pre-sented in Table 3. This table shows that,contrary to expectation, as many subjects whowere not given suggestions to regress as were,stated postexperimentally that, when theywere recalling the syllables during the experi-


ment, they had imagined, felt, and believedthat they were back in Mr. Ryan's class,that it was the first week in October, and thatthey had just learned the nonsense syllables.

With respect to the Imagined item (Table3, first score in each cell), the majority ofsubjects in each of the nine experimentalgroups stated that, during the experimentalrecall period, they had imagined that theywere back in the original learning situationeither "to some extent" (scores of 1) or "toa great extent" (scores of 2). Sixteen sub-jects (18%) testified that they "thoroughlyimagined" that they were in the originallearning situation (scores of 3); these 16subjects were distributed more or less evenlyover the nine experimental groups.

With respect to the Felt item (Table 3,scores in parentheses), subjects in most ofthe experimental groups stated on the averagethat they felt they were back in the originallearning situation "to some extent" (scores of1). With respect to the Believed item (Table3, scores in brackets), in each of the experi-mental groups, about as many subjects testi-fied that they had not believed that theywere back at the previous time (scores ofzero) as stated that they had believed thatthey were back "to some extent" (scores of1). Eleven percent and 10% of the subjects,respectively, stated that they had "thoroughlyfelt" and "thoroughly believed" that theywere back in the original learning situation;these subjects were scattered more or lessequally over the nine experimental groups.

Unstructured interviews conducted with thesubjects after they had answered the Post-Experimental Subjective Reports question-naire provided a tentative explanation forthese surprising results. The interview dataindicated that, when attempting to recallthe nonsense syllables during the experiment,most subjects in each of the nine experi-mental groups tried to place themselves inimagination in the original learning situa-tion, that is, they tried to visualize the in-structor, the list of nonsense syllables, andthemselves learning and writing down thelist. Since the subjects had typically at-tempted to enhance their experimental recol-lection of the syllables by trying to placethemselves in fantasy in the original learn-

ing class, they were able to testify postexperi-mentally, more or less truthfully, that, duringthe experiment, they had to some degreeimagined, felt, or believed that they wereback in the original learning situation.

Three additional statements should be madewith respect to the imagined, felt, and be-lieved items which comprised the Post-Ex-perimental Subjective Reports questionnaire:

1. These items were significantly intercor-related, but the correlations were not as highas one might expect. The overall product-moment correlations were .47, .39, and .65,respectively, between imagined and felt, imag-ined and believed, and felt and believed.

2. Whether or not the subjects stated post-experimentally that, when recalling the syl-lables, they had imagined, felt, or believedthat they were back in the original learningsituation, was not related to the number ofsyllables that they had correctly recalled.The overall correlations between the Recallscores and the imagined, felt, and believeditems were .07, —.05, and .13, respectively.

3. The degree to which the subjects inGroups 3, 6, and 9 met criteria for regressionduring the experiment was not related totheir postexperimental testimony. The prod-uct-moment correlations between the Regres-sion scores and the imagined, felt, andbelieved items of the Post-Experimentalquestionnaire were in no case significantlydifferent from zero.

The discrepancies between the Regressionscores and the scores on the Post-Experi-mental Subjective Reports questionnairemerit further comment. The reader will re-call that all subjects in Group 3, 60% inGroup 6, and 80% in Group 9 answeredeach of the three questions asked during theexperimental regression period (for example,What is the month and the week?) as if theywere regressed, that is, as if they were backin the original learning class. However, a farsmaller percentage of subjects in each ofthese groups testified postexperimentally that,during the experimental recall period, theyhad "thoroughly" imagined, felt, or believedthat they were back in the original learningsituation. The discrepancies between thesubjects' experimental and postexperimentalstatements may be due to various factors,

178 THEODORE X. BARBEK AND DAVID S. CALVERLEY

such as (a) the questions asked during theexperimental regression period were notworded in the same way as the questionsasked postexperimentally, and (b) the ex-perimental regression questions were answeredorally whereas the postexperimental questionswere answered by writing. Although theseand other factors may have played a role, itseems that the most important factor inproducing the differences in the experimentaland postexperimental reports was as fol-lows: The questions asked during the ex-perimental regression period followed immedi-ately upon repeated emphatic suggestionsthat the subject was in the original learningsituation, for example, ". . . It is the firstweek in October, you are in Mr. Ryan'sclass . . . you have just learned the non-sense syllables. . . ." These suggestionscarried the strong connotation that if the sub-ject were to state that it was not the firstweek in October, that he was not in Mr.Ryan's class, etc., then he would be behavingin a negativistic manner characteristic of"poor" subjects, and he would be subvertingthe experiment and disappointing the experi-menter. In contrast, the postexperimentalquestions were answered after it was statedthat the experiment was over and withoutany explicit suggestions as to how the sub-ject was to answer. Whatever suggestive fac-tors were present in the postexperimentalperiod were thus far less obvious, and pre-sumably far less potent, than the explicit,direct, suggestive influence that was exertedduring the experimental regression period.

The results obtained with the Post-Experi-mental Subjective Reports questionnaireraise a serious question pertaining to previousstudies concerned with "hypnotic age regres-sion." Prior studies usually included amongthe criteria for regression the hypnotic sub-ject's postexperimental testimony that, dur-ing the experiment, he felt or believed thathe had returned to an earlier chronologicalage. However, nonhypnotic subjects werenever asked postexperimentally if, during theexperiment, they too had felt or believed thatthey had returned to a previous age (seeBarber, 1962b, for a review of this litera-ture). The results of the present investiga-tion strongly indicate that no conclusions

concerning the effectiveness of hypnotic in-duction (or "hypnosis") in producing ageregression can be deduced from any previousexperiment in this area until the experimenthas been replicated with the addition of anonhypnotic control group. Subjects in thecontrol group should be instructed to recallevents from a previous time and should beasked postexperimentally whether, when re-calling the events, they had felt or believedthat they were back in the previous time. Ifsubjects are assigned at random to hypnoticand nonhypnotic groups, investigators maybe surprised to find that waking control sub-jects who did not receive suggestions to re-gress give postexperimental subjective reportswhich are very similar to those given by hyp-notic subjects who received regression sug-gestions.

Differences in Performance of "Good" and"Poor" Hypnotic Subjects

At the close of the experiment, the subjectswho had been exposed to a hypnotic induc-tion (Groups 1, 2, and 3) were given addi-tional suggestions of relaxation, drowsiness,and sleep and then were assessed on responseto the Barber Suggestibility Scale (BSS). Inaccordance with previously established norms(Barber, 196Sc), subjects obtaining Objec-tive scores of 0-2, 2-4, 4-6, and 6-8 on theBSS were classified as poor, fair, good, andexcellent hypnotic subjects, respectively.Two-thirds of the hypnotic subjects wererated as either excellent or good. Specifically,of the 30 subjects in the three hypnoticgroups, 4 (13%) were rated as poor, 6(20%) as fair, 11 (37%) as good, and 9(30%) as excellent.5

Was response to the BSS related to pro-ficiency of recall of the nonsense syllables

6 Investigators who use the concept "depth ofhypnosis" would most likely have categorized theexcellent subjects as "deeply hypnotized," the goodsubjects as in "medium hypnosis," etc. When theconstruct "hypnotic depth" is used in this manner,however, it says no more than that subjects whohad received a hypnotic induction procedure showeda certain degree of suggestibility and it fails totake into account the possibility that the subjectsmay have shown the same degree of suggestibilitywithout a hypnotic induction (Barber, 1964; Barber& Calverley, 1965).


(Recall scores) or to the imagined, felt, orbelieved items that comprised the Post-Ex-perimental Subjective Reports questionnaire?To answer this question, Pearsonian corre-lations were computed separately for Groups1, 2, and 3 between scores on the BSS andRecall scores and between scores on the BSSand scores on the imagined, felt, and be-lieved items. None of the correlations betweenthe BSS and Recall scores were significantlydifferent from zero. However, response to theBSS was consistently related to response tothe imagined, felt, and believed items. Over-all, scores on the BSS were correlated .36with the imagined item, .54 with the felt item,and .57 with the believed item (ps < .05).

With respect to the hypnotic group thathad received suggestions to regress (Group3), there was no relationship (r = .00) be-tween scores on the BSS and Regressionscores.

In summary, these results indicate that:(a) The better hypnotic subjects (those ob-taining higher scores on the BSS) did notrecall more of the nonsense syllables thanthe poorer hypnotic subjects, (b) The betterhypnotic subjects did not differ from theothers in response to suggestions to regress.(Irrespective of their scores on the BSS, allhypnotic subjects in Group 3 who receivedregression suggestions responded to questions[e.g., "What is the month and the week?"] asif they were regressed.) (c) The better hyp-notic subjects more often than the poorersubjects testified postexperimentally that,during the experimental recall period, theyhad imagined, felt, and believed that theywere back in the original learning situation.6

6 If, in accord with many previous experiments,we had nonrandomly allocated good hypnotic sub-jects to the hypnotic treatments and poor hypnoticsubjects or unselected subjects to the nonhypnotictreatments, would we have obtained a different setof results? The data presented above indicate that,if we had used this nonrandom method to allocatesubjects, we still would have found that the hypnoticand nonhypnotic groups did not differ significantlyin Recall scores and in Regression scores althoughthey may have differed on the imagined, felt, andbelieved items (Post-Experimental Subjective Re-ports). However, as pointed out in the methodo-logical introduction to this paper, if we had allo-cated good subjects to the hypnotic treatment butnot to the nonhypnotic treatment, it would have

REFERENCES

As, A. The recovery of forgotten language knowl-edge through hypnotic age regression: A casereport. American Journal of Clinical Hypnosis,1962, S, 24-29.

BARBER, T. X. Experimental evidence for a theoryof hypnotic behavior: II. Experimental controlsin hypnotic age-regression. International Journalof Clinical and Experimental Hypnosis, 1961, 9,181-193.

BARBER, T. X. Experimental controls and the phe-nomena of "hypnosis": A critique of hypnoticresearch methodology. Journal of Nervous andMental Disease, 1962, 134, 493-501 (a)

BARBER, T. X. Hypnotic age-regression: A criticalreview. Psychosomatic Medicine, 1962, 24, 286-299. (b)

BARBER, T. X. "Hypnosis" as a causal variable inpresent-day psychology: A critical analysis. Psy-chological Reports, 1964, 14, 839-842.

BARBER, T. X. Effects of "hypnosis" on learning andrecall: A methodological critique. Journal ofClinical Psychology, 1965, 21, 19-25. (a)

BARBER, T. X. Experimental analyses of "hypnotic"behavior: A review of recent empirical findings.Journal of Abnormal Psychology, 1965, 70, 132-154. (b)

BARBER, T. X. Measuring "hypnotic-like" suggesti-bility with and without "hypnotic induction";psychometric properties, norms, and variables in-fluencing response to the Barber SuggestibilityScale (BSS). Psychological Reports, 1965, 16(Monogr. Suppl. 3—V16), 809-844. (c)

BARBER, T. X., & CALVERLEY, D. S. "Hypnotic-like"suggestibility in children and adults. Journal ofAbnormal and Social Psychology, 1963, 66, 589-597.

BARBER, T. X., & CALVERLEY, D. S. Comparativeeffects on "hypnotic-like" suggestibility of recordedand spoken suggestions. Journal of ConsultingPsychology, 1964, 28, 384. (a)

BARBER, T. X., & CALVERLEY, D. S. Effect of £'stone of voice on "hypnotic-like" suggestibility.Psychological Reports, 1964, IS, 139-144. (b)

BARBER, T. X., & CALVERLEY, D, S. Toward a theoryof hypnotic behavior: Effects on suggestibility ofdefining the situation as hypnosis and definingresponse to suggestions as easy. Journal of Ab-normal and Social Psychology, 1964, 68, 585-592.(c)

BARBER, T. X., & CALVERLEY, D. S. Empirical evi-dence for a theory of hypnotic behavior: Effectson suggestibility of five variables typically in-cluded in hypnotic induction procedures. Journalof Consulting Psychology, 1965, 29, 98-107.

been impossible to state whether any differences ob-tained between hypnotic and nonhypnotic groups onPost-Experimental Subjective Reports were due tothe presence of "hypnosis" in one group and itsabsence in the other or to one group but not theother consisting of subjects who were highly sug-gestible with or without "hypnosis."


BARBER, T. X., KAEACAN, I., & CALVERLEY, D. S."Hypnotizability" and suggestibility in chronicschizophrenics. Archives of General Psychiatry,1964, 11, 439-451.

GLAZE, J. A. The association value of nonsensesyllables. Journal of Genetic Psychology, 1928, 35,255-267.

HAMMER, E. F. Post-hypnotic suggestion and testperformance. Journal of Clinical and ExperimentalHypnosis, 19S4, 2, 178-185.

HILGARD, E. R. Methods and procedures in the studyof learning. In S. S. Stevens (Ed.), Handbook ofexperimental psychology. New York: Wiley, 1951.Pp. 517-567.

LEONARD, J. R. An investigation of hypnotic age-regression. Unpublished doctoral dissertation, Uni-versity of Kentucky, 1963.

LYON-JAMES, S. C. The effect of poslhypnotic sug-gestion upon the study and recall of written ma-terial. Unpublished doctoral dissertation, Univer-sity of Denver, 1957.

MARCUSE, F. L. Hypnosis: Fact and fiction. Balti-more: Penguin, 1959.

PATTIE, F. A. A report on attempts to produceuniocular blindness by hypnotic suggestions.British Journal of Medical Psychology, 1935, IS,230-241.

ROSENTHAL, B. G. Hypnotic recall of materiallearned under anxiety and non-anxiety producing

conditions. Journal of Experimental Psychology,1944, 34, 369-389.

ScHRENCx-NoiziNG, A. F. VON. Ein experimentellerund kritischer Beitrag zur Frage des suggestivenHervorrufung circumscripter vasomotorischer Ver-anderungen auf der ausseren Haut. ZeitschriftHypnotismus, 1896, 4, 209.

STALNAKER, J. M., & RIDDLE, E. E. The effect ofhypnosis on long delayed recall. Journal of Gen-eral Psychology, 1932, 6, 429-440.

SUTCLIEFE, J. P. "Credulous" and "sceptical" viewsof hypnotic phenomena: A review of certain evi-dence and methodology. International Journal ofClinical and Experimental Hypnosis, 1960, 8, 73-101.

TROFEER, S., & TART, C. Experimenter bias in hyp-notist performance. Science, 1964, 145, 1330-1331.

TRUE, R. M. Experimental control in hypnotic ageregression states. Science, 1949, 110, 583-584.

WEITZENHOFFER, A. M., & HILGARD, E. R. Stanfordhypnotic susceptibility scale. Palo Alto, Calif.:Consulting Psychologists Press, 1959.

WHITE, R. W., Fox, G. F., & HARRIS, W. W. Hyp-notic hypermnesia for recently learned material.Journal of Abnormal and Social Psychology, 1940,35, 88-103.

WOLBERG, L. R. Medical hypnosis: I. The principlesof hypnotherapy. New York: Grune & Stratton,1948.

(Received April 23, 1965)

Documents

EFFECTS ON RECALL OF HYPNOTIC INDUCTION, MOTIVATIONAL