Pre-publication copy of the paper eventually published as Dinges, D. F., Whitehouse, W. G., Orne, E. C., Powell, J. W., Orne, M. T., & Erdelyi, M. H. Evaluating hypnotic memory enhancement (hypermnesia and reminiscence) using multitrial forced recall. Journal of Experimental Psychology: Learning, Memory, and Cognition, 1992, 18, 1139-1147.

Evaluating Hypnotic Memory Enhancement (Hypermnesia and Reminiscence) Using Multitrial Forced Recall

David F. Dinges, Wayne G. Whitehouse, Emily Carota Orne, John W. Powell, and Martin T. Orne

The Institute of Pennsylvania Hospital and University of Pennsylvania

Matthew H. Erdelyi

Brooklyn College, City University of New York



Abstract: Two experiments investigated whether hypnosis enhances memory retrieval per se or merely increases a person's willingness to report recollections. Both experiments assessed immediate and delayed (i.e., 1 week) recall for pictorial stimuli. In Experiment 1, following an initial waking baseline recall, subjects of high or low hypnotic ability completed a series of recall trials conducted either in hypnosis or in the waking condition. The classic hypermnesia effect was obtained, but with no supplemental contribution of hypnosis. In Experiment 2, hypnosis was introduced only after 6 waking-recall trials. Hypnosis again failed to enhance retrieval of new correct items, although it increased the production of new incorrect recall among hypnotizable individuals. The findings provide no evidence for alleged hypermnesic properties of hypnosis.



The phenomenon of hypermnesia (i.e., increased recall with repeated retrieval attempts) is well documented in the memory literature (see Belmore, 1981; Erdelyi, 1982, 1984; Erdelyi & Becker, 1974; Payne, 1987), although isolation of the responsible mechanism(s) remains a goal of empirical and theoretical analyses (e.g., Erdelyi, 1982; Erdelyi & Becker, 1974; Payne, 1986; Roediger, Payne, Gillespie, & Lean, 1982). Such is not the case, however, regarding the phenomenon of hypnotic hypermnesia, the fundamental reality of which is still unresolved. The view that hypnosis may improve recall arises from its usefulness in psychotherapy in helping patients come to terms with emotions and autobiographical experiences that they are otherwise consciously unable to remember or disclose. Additionally, much of the current interest in hypnotic hypermnesia has been fueled by the use of hypnosis in law enforcement to "refresh" the recollection of eyewitnesses and victims of crimes (Orne, Soskis, Dinges, & Orne, 1984; Udolf, 1983). Because of obvious difficulties in documenting the historical authenticity of information recalled during hypnosis, neither the therapeutic nor the forensic setting is adequate to evaluate hypnotic hypermnesia. For this reason, a number of experimental studies have been carried out to investigate the alleged memory-enhancing properties of hypnosis. (For reviews, see Orne, Whitehouse, Dinges, & Orne, 1988; and Smith, 1983.)

By and large, evidence in support of the phenomenon of hypnotic hypermnesia has been sporadic and unreliable. There is a clear indication that memory for nonsense syllables is not improved by the use of hypnosis (e.g., Baker, Haynes, & Patrick, 1983; Barber & Calverley, 1966; Dhanens & Lundy, 1975; White, Fox, & Harris, 1940; Young, 1925). On the other hand, the use of meaningful stimuli, such as films, pictures, prose, or poetry, does not ensure positive effects of hypnosis on subsequent memory tests. Indeed, although there have been a few reports of enhanced recall among hypnotically responsive subjects experiencing hypnosis (e.g., Dhanens &



Lundy, 1975; Dywan & Bowers, 1983; McConkey & Kinoshita, 1988; Stager & Lundy, 1985), the majority of studies have found no advantage for hypnosis over normal waking recall (e.g., Lytle & Lundy, 1988; McConkey & Nogrady, 1984; Nogrady, McConkey, & Perry, 1985; Putnam, 1979; Register & Kihlstrom, 1987; Sheehan, 1988; Whitehouse, Dinges, Orne, & Orne, 1988; Yuille & McEwan, 1985). Similarly, the enhancement effects of hypnosis have been found to be no greater than those produced by motivating instructions (Cooper & London, 1973) or by using specialized cognitive retrieval strategies (Geiselman, Fisher, MacKinnon, & Holland, 1985; Mingay, 1986).

Nevertheless, a simple tally of the number of studies that appear to validate the phenomenon versus those that do not is unlikely to decide the issue. Failures to produce hypnotic hypermnesia could be due to any number of variables, ranging from the hypnotizability of the subject sample and the nature of the hypnotic suggestions employed, to the choice of stimulus material or memory test. An observation culled from many recent studies, however, is that hypnotic procedures tend to result in increased reporting of erroneous recall (e.g., Dywan, 1988; McConkey & Kinoshita, 1988; Nogrady et al., 1985; Register & Kihlstrom, 1987; Whitehouse et al., 1988). This point is telling because it raises the possibility that occasional positive findings could reflect a reporting bias rather than enhanced retrieval. As noted by Klatzky and Erdelyi (1985), efforts toward an unequivocal demonstration of hypnotic hypermnesia should seek to distinguish any direct effect of hypnosis on the pool of accessible memories from its effect on the subject's willingness to report recollections and the confidence placed in those reports.

One approach to this problem is to attempt to hold constant the subject's criterion for reporting recollections by using a forced-recall procedure (Erdelyi & Becker, 1974). With such a procedure, the subject is required to provide a fixed "recall" output on every trial, with the limit being set well above the level obtained using standard free recall. This has the effect of causing subjects to offer a



certain number of subcriterion responses (i.e., low-confidence responses that, in the context of free recall, would likely not have been reported) to achieve the required output. Accordingly, any differences in correct recall between subjects or over trials cannot be attributed to a fortuitous relaxation of the report criterion, which could result in the confident reporting of one or more correct guesses. Instead, a difference in correct recall between two forced-recall assessments implies a true difference in accessible memory. Because it virtually eliminates the effects of response bias in memory reports, the forced-recall technique has been instrumental in identifying hypermnesia as a genuine increase in accessible memory produced by repeated retrieval efforts (Erdelyi & Becker, 1974; Erdelyi & Kleinbard, 1978).

An adaptation of this methodology was used in a recent investigation of the effects of hypnosis on memory for complex filmed material (Whitehouse et al., 1988). During each of two identical interrogatory recall tests, one of which was a waking baseline test and the other a hypnosis versus waking treatment recall test, subjects were required to provide a response to every question, even if they had to guess. In addition, confidence ratings were assigned to each response in order to distinguish guesses from confidently supplied responses. Subjects in the hypnosis condition were able to provide additional correct information relative to their baseline performance, but no more so than subjects in the waking condition. Moreover, while subjects in the two treatments did not differ in the level of confidence they invested in their correct recall, subjects in the hypnosis condition, and in particular, those with high hypnotic ability, expressed significantly greater confidence in their incorrect recall than did subjects who recalled only in the waking condition. Thus, by using a modification of the forced-recall methodology developed by Erdelyi and Becker (1974), Whitehouse et al. were able to rule out an explanation of the interest gain in correct recall based on a report-criterion shift. At the same time, however, because the increase in correct recall from baseline to treatment was the same in magnitude for subjects in the hypnosis and waking



conditions, the findings provided no evidence of a contribution to hypermnesia attributable to the use of hypnosis.

Although the investigation by Whitehouse et al. (1988) failed to yield evidence of hypnotic hypermnesia when report criterion was held constant, it is possible that other factors rendered the paradigm nonconducive to the expression of hypnotic hypermnesia. This possibility seems unlikely, however, because the study was a systematic replication, except for the use of a forced-recall procedure, of an earlier experiment by Stager and Lundy (1985), in which highly hypnotizable subjects who were exposed to the identical film stimulus, tape-recorded hypnotic proceedings, and recall questions gave significantly more correct responses than their nonhypnotic counterparts. Nevertheless, it may be prudent to defer conclusions about the reality of hypnotic hypermnesia until other experimental circumstances, potentially more favorable to its manifestation, are investigated.

We report here two experiments, both using the forced-recall procedure of Erdelyi and Becker (1974), that attempt to evaluate hypnotic hypermnesia under potentially more conducive conditions than those of the investigation by Whitehouse et al. (1988). Thus, one objection to the latter study might be that it relied on only a single opportunity to recall during hypnosis. Experiment 1 of the current study, therefore, provided subjects with five consecutive opportunities to recall information in hypnosis and compares their performance with that of subjects who undertook an equal number of recall attempts in the waking condition. Another objection to the earlier study might be that the treatment recall test occurred at a time when the recall level of subjects in both conditions was preasymptotic and possibly unstable. If so, any recall benefit due to hypnosis might have been masked by the equally large gains in correct recall possible for subjects



in the waking control condition. In support of this possibility, Dywan and Bowers (1983) observed that subjects in a hypnotic recall condition reported more than twice as much new information as did unhypnotized control subjects following a full week of daily waking recall attempts. It should be noted, however, that less than one third of the content of this increased output was correct recall. Nevertheless, to assess the potential importance of the extent of prehypnotic recall, Experiment 2 examines the effects of administering hypnosis after subjects have already carried out six recall attempts in the waking condition. Finally, each of the present experiments relies on a multitrial procedure to determine if hypnosis enhances accessible memory beyond that attributable to normal waking hypermnesia resulting from repeated retrieval effort.


This experiment evaluated the effect of hypnosis on multitrial forced recall of pictorial stimuli following a 1-week retention interval. The primary objective was to control report criteria of subjects in hypnosis and waking conditions while providing substantial opportunity (five consecutive trials) for any impact of hypnosis on recall to manifest itself. Additionally, subjects were selected on the basis of their low or high responsivity to suggestions given during prior standardized tests of hypnotic ability. These trait differences in hypnotizability can help to identify whether any effects on recall associated with hypnosis are the result of the actual experiencing of suggestions for enhanced retrieval (i.e., are observed selectively among highly responsive subjects), or whether they arise from motivation or compliance with demand characteristics (Orne, 1962) inherent in the experimental procedures (i.e., are observed equally among subjects of low and high hypnotic ability).





The subjects were 48 undergraduate student volunteers (22 men, 26 women) who were selected from a larger sample (n = 428) for scoring in the extreme ranges of both the Harvard Group Scale of Hypnotic Susceptibility, Form A (HGSHS:A of Shor & Orne, 1962) and the Stanford Hypnotic Susceptibility Scale, Form C (SHSS:C of Weitzenhoffer & Hilgard, 1962). 1 Twenty-five subjects (12 men, 13 women) were classified as low hypnotizable (SHSS:C = 0-4) and 23 (10 men, 13 women) were classified as high hypnotizable (SHSS:C = 8-12).


Initially subjects participated in 1 of 16 group sessions, ranging in size from 9 to 42 subjects (M = 27). During these sessions, the tape-recorded HGSHS:A was administered to assess hypnotic ability, and self-report booklets that evaluated both behavioral and subjective responses to the hypnotic suggestions were completed. Following this, the booklets were collected and subjects were engaged in other activities (i.e., watching a film and completing research questionnaires), which were unrelated to the focus of the present experiment, while the HGSHS:A booklets were unobtrusively scored. Subjects who scored below 6 and above 8 on the HGSHS:A behavioral scale were invited to participate in two individual follow-up sessions, although the basis for their selection for these latter sessions was not disclosed.

During the first individual session, subjects were informed that they would view a series of 40 slides that they would later be asked to recall. The slides consisted of black-and-white line drawings of common objects from the set developed by Erdelyi (e.g., Erdelyi & Becker, 1974). Presentation rate was 1 slide every 5 s, and subjects were asked to name aloud the object in each slide while



the experimenter recorded the specific label given to every item. Immediately afterward, a forced-recall test (R1) was administered. Following the procedures of Erdelyi, subjects were given a sheet of paper consisting of 40 blank lines and were told to write down the names of all the objects they could recall having seen. When they felt that they could not remember any further items, they were to draw a line under the last entry and begin filling in the remaining blank spaces with their best guesses. Any new items recalled during this guessing section were to be distinguished from guesses by placing a check mark next to those items. The recall test lasted 5 min, and subjects were warned when they had 2 min and 1 min remaining. When subjects had completed the forced-recall test, they were introduced to a second experimenter, who administered the standardized SHSS:C assessment of hypnotic ability. Following this, the same experimenter interviewed subjects regarding their perceptions of the experiment and their experiences with the group and individual hypnosis scales, and confirmed their appointments for the third and final session.

Subjects returned to the laboratory 1 week later to participate in a final individual session. At this time, they were introduced to a new experimenter, who was unaware of both the specific slides the subjects had viewed and their hypnotizability scores. The session began with the experimenter explaining that the purpose of the study was to investigate certain procedures that would help to refresh their memory for the pictures they saw in the previous week's session. In attempting to establish a plausible rationale, reference was made to experimental evidence for the efficacy of trying to recall in the same environmental context in which the slides were originally shown, and to the benefits of repeatedly searching their memories. No mention was made at this point in the session of the number of times they would be asked to recall the slides, or the possible use of hypnosis. Following these introductory comments about the research, subjects were given instructions for completing a forced-recall test (R2), which was identical to the R1 test administered after they had viewed the slides a



week earlier. As before, subjects were told that they would have 5 min in which to fill all 40 spaces on their recall sheet with the names of the objects depicted in the slides. When they felt that they could not remember any further items, they were to draw a line under the last item they were sure of, and fill in the remaining spaces with guesses, placing check marks next to any nonguesses (i.e., remembered items) that might be written below the line.

At the conclusion of the R2 forced-recall test, which provided a waking baseline recall measure, the experimental treatments were introduced. Of the 25 low-hypnotizable subjects, 13 had been randomly preassigned to the waking condition and 12 to the hypnosis treatment; of the 23 high-hypnotizable subjects, 11 were preassigned to the waking condition and 12 to hypnosis.

Subjects in the waking condition were informed that periodically they would be engaged in a task designed to take their minds off of the pictures for a short time so that their recall would improve when they next tried to remember the stimuli. They were then given a modified audiocassette recorder and instructions for a visual reaction time (RT) task. Periodic illumination of a light-emitting diode numeric display at the top of the audiocassette recorder served as the target stimulus, which subjects were required to terminate as quickly as possible by pressing a button. The RT value in milliseconds was registered in the numeric display and read aloud by the subject following each trial. The random interstimulus interval ranged from 3 to 13 s. The visual RT task was administered prior to the R3 forced recall attempt and again prior to R8. On both occasions it was carried out for 10 min. At other times, subjects were instructed to spend time relaxed, with their eyes closed, thinking about the pictures until they were told to stop. These "think" intervals each lasted 2 min and occurred prior to R4, R5, R6, R7, and following the RT task, prior to R8.

Subjects in the hypnosis condition were informed that hypnosis would be used to help them



better remember the pictures. It was explained that hypnosis would (a) take their minds off the pictures for a while, thereby making them easier to remember when asked to do so; (b) allow them to focus their minds on the things they are trying to remember; and (c) facilitate their reexperiencing of the previous session, seeing the slides being shown again in their minds and hearing themselves name the objects. Consistent with common forensic practice, it was suggested that hypnosis would bring additional memories from their subconscious minds, and all their recollections would be more vivid and clear. The hypnotic induction procedure was a modification of the induction used with the SHSS:C, which incorporated a number of direct suggestions for enhanced accurate recall and assurances that, when asked to do so, subjects would be able to open their eyes and write down the names of the pictures while remaining deeply relaxed and hypnotized. Prior to each recall attempt, subjects spent a 2-min period quietly thinking about the pictures. Suggestions aimed at deepening the level of relaxation and hypnosis were given following R4 and again following R6. Following R7, subjects were told that they were about to be awakened from hypnosis, but that they would continue to be able to remember all of the pictures that they recalled prior to and during hypnosis. After terminating hypnosis, subjects were given instructions for the visual RT task, which was administered for a 10-min period and followed by a 2-min think interval; R8 was then carried out in the waking condition.

To summarize, the important procedural difference between the two recall treatments was that low- and high-hypnotizable subjects in the waking condition performed all forced-recall trials (R2-R8) while awake and alert, whereas their counterparts in the hypnosis treatment performed only R2 and R8 in the waking condition; trials R3-R7 were carried out in hypnosis.

Following the final recall trial, subjects were escorted to another room, where they were interviewed by a different experimenter about their impressions of the study, debriefed, thanked, and paid for their participation.




Between-subjects and mixed-design factorial analysis of variance (ANOVA) models were applied separately to correct recall, confident incorrect recall, and items forgotten between successive recall trials. Post-hoc comparisons were computed using the Newman-Keuls procedure. In cases where effects are identified as statistically significant, the associated alpha level is .05 or better.

Baseline Recall Performance: Immediate and 1-Week Delay

The mean immediate correct recall level was 52%, which declined to 38% when retested after 1 week. A 2 (low vs. high hypnotizability) x 2 (waking vs. hypnosis treatment assignment) x 2 (R1 vs. R2 trial) repeated measures ANOVA found only a significant main effect of trials, F(1, 44) = 111.91, MSe = 6.07. Thus, there were no preexisting differences in either immediate or delayed baseline correct recall among the groups prior to the treatment phase. Confident errors (i.e., excluding incorrect guesses supplied to meet the fixed-output requirement of the forced-recall procedure) increased from a mean of 0.29 at R1 to 1.27 at R2, F(1, 44) = 21.22, MS, = 1.11, but again, there were no intergroup differences on either baseline trial.

Treatment Recall Performance

Hypermnesia. Figure 1 depicts mean correct recall performance, calculated without regard to confidence, for groups of subjects representing the four Hypnotic Ability X Treatment combinations on each recall trial. To evaluate the influence of the treatment phase, a mixed-model ANOVA was conducted on the number of correct responses given on each of trials R3 through R8. This analysis identified a significant gain in correct recall over trials, F(5, 220) = 20.89,



MSe = 2.59. However, effects associated with hypnotizability and waking versus hypnosis treatment, as well as their interactions with the trials factor, were unreliable. A similar analysis based on only confidently reported correct recall yielded comparable results. Although Figure 1 suggests that highly hypnotizable subjects in the hypnosis condition produced quantitatively more correct recall than subjects in the remaining conditions, it should be noted that this slight superiority was evident as well during the two waking baseline trials and thus was not conferred by the hypnotic procedure. Overall, the findings are clear that the hypermnesia associated with repeated recall attempts was not enhanced by the use of hypnosis.

Insert Figure 1 about here

Reminiscence and forgetting. Cumulative new correct recall (i.e., correctly recalled pictures not reported during the R2 delayed baseline test) was examined to determine whether hypnosis differentially influenced the rate of recovery over trials of additional correct items (i.e., reminiscence). It is possible, for example, that hypnosis accessed more items from memory but that these gains were not maintained across trials (i.e., greater intertrial forgetting occurred), resulting in comparable net recall performance between hypnosis and waking treatments. Figure 2 portrays the relevant data, beginning with the first treatment recall trial, R3. As the figure indicates, the rate at which items were forgotten, despite having been reported during the immediately preceding trial, was stable throughout the treatment and posttreatment (R8) trials. Although it appears that subjects in the hypnosis condition experienced less intertrial forgetting than waking subjects over the final three recall attempts, this difference was not statistically reliable. On the other hand, there is clear evidence that with each trial, subjects in the two treatment conditions continued to produce previously unreported correct recall. Of course, due to the incremental nature of



this cumulative recall measure, the trials effect was highly reliable, F(5, 220) = 148.68, MSe = 1.49. More important, the Treatment X Trials interaction was also significant, F(5, 220) = 4.33, MSe = 1.49. Post hoc tests found that the two treatments did not differ significantly on any trial; however, the overall rate of reminiscence over trials was reduced for subjects in the hypnosis condition compared to that for subjects in the waking condition. The data suggest, therefore, that the comparability of hypnosis and waking conditions with respect to their hypermnesia functions is the result of an inferior rate of new item recovery in hypnosis being offset by slightly, but nonsignificantly, better intertrial retention than that which occurred among waking subjects. Nevertheless, the finding that hypnosis was associated with poorer reminiscence is at odds with its purported retrieval-enhancement properties.

Insert Figure 2 about here

Intrusions. Confident errors also increased over trials, F(5, 220) = 17.21, MSe = 1.57. However, as indicated in Figure 3, highly hypnotizable subjects produced more confident errors overall, F(1, 44) = 5.38, MSe = 66.73, and exhibited a greater increase in error rate over trials, F(5, 220) = 2.61, MSe = 1.57, than did low-hypnotizable subjects. Although the figure suggests that the use of hypnosis differentially affected the tendency for high- and low-hypnotizable subjects to report such pseudomemories, the three-way interaction involving treatment, hypnotic ability, and trials fell short of significance, F(5, 220) = 1.90, MSe = 1.57.



Insert Figure 3 about here


The findings of this experiment indicate that when a forced-recall procedure is adopted to eliminate the possibility of differential productivity, hypnosis provides no recall advantage over normal waking recollective efforts. The failure to observe an effect of the hypnotic procedure is striking in view of the rather extensive opportunity (i.e., five trials) for hypnotic hypermnesia to emerge. In fact, there was evidence that the use of hypnosis diminished the recovery of new items over trials, such that the overall rate of reminiscence was significantly lower than that achieved by subjects who recalled exclusively in the nonhypnotic condition. The data also show that following a 1-week retention interval, hypnotizable individuals were particularly prone to accepting erroneous recall as being accurate, whether or not hypnosis was used. Furthermore, although the tendency to develop confidence in the accuracy of incorrect information with repeated recall opportunities was characteristic of both low- and high-hypnotizable subjects, it was more pronounced among hypnotizable persons.

Thus, despite the use of a multiple-recall paradigm congenial to the expression of hypermnesia, no unique recall enhancement could be credited to the use of hypnosis. Furthermore, hypnosis decreased reminiscence and appeared to increase pseudomemories reported by low-hypnotizable subjects. Rather than documenting an enhancement effect of hypnosis on memory retrieval, the present findings highlight liabilities associated with hypnotic recall relative to normal waking recall.




The purpose of this experiment was to assess the impact of a single hypnotic recall opportunity administered only after subjects had attempted to recall numerous times in the waking condition. Presumably, given an ample number of recall attempts, the pool of items in memory that are accessible by normal waking retrieval mechanisms should be virtually exhausted. If, however, hypnosis provides alternate access routes to information in memory, there should be a measurable increment in correct recall occasioned by the use of hypnosis even though waking recall is nearly asymptotic.



Forty-one undergraduate student volunteers (18 men, 23 women), derived from the same parent sample as subjects in the previous experiment, served as subjects. Selection was based on subjects' having obtained HGSHS:A scores ranging from 0 to 5 or 9 to 12. A subsequent, individually administered hypnotizability assessment was made with the SHSS:C. To sample a broader spectrum of hypnotic ability than was used in the previous experiment, subjects were not excluded if their scores on the SHSS:C fell in the middle range of hypnotic ability. Rather, the scores on the two instruments were averaged to provide a stable estimate of hypnotic ability, yielding 13 subjects who scored between 0 and 4, 12 between 5 and 8, and 16 between 9 and 12.


The procedure for the first two sessions of the present experiment was identical, in every respect, to that



described for Experiment 1. Thus, the initial hypnotizability screening with the tape-recorded HGSHS:A took place during Session 1. The second session, conducted approximately 1 week later, involved presentation of the stimulus slides, followed by an immediate forced-recall test (R 1), and the administration of the SHSS:C. The subject was then thanked and paid, and an appointment for the final individual session was confirmed.

Upon returning to the laboratory the next week, the subject was introduced to a new experimenter, who explained that they would be using several memory retrieval strategies to help the subject remember the pictures that were shown in the previous session. These strategies involved (a) returning to the same setting, (b) repeatedly searching memory, and (c) performing an interesting diversionary task. Subjects were not told in advance the number of times they would be asked to recall the slides, or about the eventual use of hypnosis, although they had consented to participate in hypnosis at an unspecified point during the session. Instructions were then given for completing the R2 forced-recall test. Briefly, subjects were given 5 min in which to fill all 40 spaces on their recall sheet with the names of objects depicted in the slides. When they felt that they could not remember any further items, they were to draw a line under the last item they were sure of and fill in the remaining spaces with guesses, using check marks to identify any remembered items that were written below the line.

At the conclusion of the R2 forced-recall test, subjects performed the 10-min visual RT task described previously. This was followed by another 5-min forced-recall test (R3). Preceding each of the next four forced-recall tests (i.e., R4 through R7), subjects were instructed to relax, with their eyes closed, and to think about the pictures. These think intervals were each 2 min in duration.

Prior to R8, subjects were informed that the final memory facilitation technique that would be used was hypnosis. The experimenter explained that hypnosis helps to (a) take one's mind off the task for a time to provide a fresh perspective, (b) enhance concentration, and (c) allow one to mentally reexperience the



pictures being presented while hearing oneself name the objects. It was further suggested that hypnosis could bring forth additional memories from the subconscious with great vividness and clarity. A hypnotic induction procedure, adapted from the SHSS:C, was administered, and suggestions for enhanced recall were given. Subjects then completed the R8 forced-recall test during hypnosis. Upon termination of hypnosis, subjects were extensively interviewed about their experiences, thanked, and paid for their participation.

Results and Discussion

Figure 4 depicts correct recall (independent of confidence) and confident recall errors over the course of the investigation. Correct recall on the immediate (R1) forced-recall test was 53%; following a 1-week interval, correct recall performance dropped to 37% on the R2 baseline test. These data compare favorably with those obtained in Experiment 1. Also in line with the findings of that experiment, recall improved significantly during the waking trials from R2 through R7, F(5, 200) = 46.3, MSe = 2.63, reflecting the basic hypermnesic process engendered by repeated recall attempts. In addition, confident recall errors increased with time and number of recall attempts, F(6, 240) = 18.13, MSe = 2.45, beginning at a very low level on the immediate recall test (M = 0.22, SD = 0.57) and increasing progressively (M = 3.17, SD = 3.63 by the final prehypnosis trial, R7).

Insert Figure 4 about here

The focal data for the present experiment concern the amount of new correct recall and new confident incorrect recall (i.e., new intrusions) associated with the use of hypnosis on R8 compared with that obtained during the preceding waking recall trials. Descriptive statistics for each trial subsequent to the delayed recall baseline (R2) test are provided in Table 1. These data indicate that the amount of newly reported correct information decreased in an irregular pattern over waking recall trials. Furthermore, the mean number of new correctly recalled items



obtained with hypnosis was significantly less than the number obtained on the immediately preceding trial, t(40) = 2.05. Because of the trend of diminishing gains in new correct recall established over the preceding waking trials, however, the further decrease at R8 cannot be attributed with any certainty to a decremental influence of hypnosis on recall.2 At the very least, however, it indicates a lack of benefit from the use of hypnosis.

Insert Table 1 about here

Table 1 also shows clearly that the tendency to report new confident incorrect items decreased systematically over repeated waking trials (i.e., R3 through R7). In opposition to its lack of effect on the retrieval of new correct items, hypnosis increased the reporting of new intrusions. Using even a very conservative contrast (i.e., the number of new intrusions reported during R8 versus the mean number of new intrusions for Trials R3 through R7), the tendency for hypnosis to generate novel confident recall errors was statistically reliable, t(40) = 2.55.

Effects related to hypnotizability were evaluated by correlations computed separately for new correct and new incorrect recall. Values for these coefficients are given on a trial-by-trial basis in Table 1. It is evident from the table that there was no significant relation between hypnotic ability and the magnitude of either new correct or new incorrect items reported during any of the waking recall trials. In contrast, hypnotizability clearly influenced performance during the sole hypnosis trial. The notable absence of correlation between hypnotizability and the number of new correct items reported during R8 (r = -.03) is consistent with the results of Experiment 1 and suggests that the retrieval of new correct items was not dependent on the hypnotic process, since subjects with greater ability were no more successful than those with little hypnotic aptitude. On the other hand, there is a modest, but significant, positive relationship between hypnotic ability and the production of new confident incorrect items (r = .33), a finding that is also concordant with the results of Experiment 1. Thus, during hypnosis, hypnotizable subjects were more likely than their relatively unhypnotizable counterparts to produce recall errors that they felt confident were correct.



General Discussion

The purpose of the present experiments was to investigate the possible enhancement of memory by hypnosis, using a forced-recall procedure that prevented subjects from merely altering their recall output between hypnosis and waking conditions. Although one earlier investigation by Dywan and Bowers (1983) also employed forced-recall methodology, the focus of their report was on the amount of new information reported with confidence following suggestions for hypnotic hypermnesia. Accordingly, their data do not distinguish between a potential increase in accessible memory versus an increase in productivity engendered by hypnosis; such a discrimination was the raison d'etre of the present Experiment 1. The current research also used multiple recall trials because the literature on waking hypermnesia clearly indicates that recall can be improved with repeated retrieval efforts (Erdelyi & Becker, 1974; Payne, 1987). Hence, we anticipated that hypermnesia would develop over trials in both waking and hypnosis treatments. However, if hypnosis were to add anything unique to the memory retrieval process, its contribution should combine with those gains attributable to multiple retrieval opportunities to produce a significantly greater amount of correct recall than that observed in the waking condition.

Because the timing of the application of hypnosis could potentially influence whether hypnotic hypermnesia can be demonstrated, we sought to assess the impact of hypnosis when introduced early and maintained for several trials, as well as when it followed extensive testing in the waking condition. This issue was deemed important because an earlier study by our laboratory (Whitehouse et al., 1988), using a variation of the forced-recall procedure, failed to find evidence of hypnotic hypermnesia when hypnosis was used for a single trial after a waking baseline recall test. Thus, in Experiment 1 of the current study, subjects received the hypnotic treatment immediately after a waking baseline recall and remained in hypnosis for an additional four trials. Although recall



improved significantly across trials, at no point did it surpass the magnitude of hypermnesia attained by waking control subjects. This finding suggests, therefore, that the lack of an effect of hypnosis in the investigation by Whitehouse et al. (1988) was not likely to have been a consequence of employing hypnosis for too brief a period of time. Even with several additional hypnotic recall opportunities, there was clearly no advantage for subjects in hypnosis.

Experiment 2 addressed the possibility that the use of hypnosis at a time when waking recall has approached asymptote would have beneficial effects for memory retrieval. In this experiment, hypnosis was introduced after subjects had attempted to recall in the waking condition over the preceding six trials. Here we were concerned primarily with the retrieval during hypnosis of new correct information, which had never been reported previously, even as a guess. In comparison to the amount of new correct recall reported during the immediately preceding waking trial, hypnosis elicited significantly less, not more, of such information. However, given a pattern of diminishing yields in new correct recall with successive prior waking trials, it is not certain whether this result was due to some detrimental effect of hypnosis. On the other hand, there was no indication that recall was improved in any way by the use of hypnosis. Accordingly, the findings of both experiments strongly support the essential conclusions of our prior research: When control is exercised over potential changes in the willingness of subjects to report information, there is no indication of a contribution of hypnosis to memory retrieval.

Recently, Erdelyi, Finks, and Feigin-Pfau (1989) reported evidence that, under some circumstances, a forced-recall task may produce a processing bias during retrieval. Their findings suggested that forced-recall subjects, who were instructed to guess if necessary, exerted less retrieval effort than free-recall subjects during the early portion of a recall trial, leading to correspondingly less correct recall in the former group. Could a processing bias, induced by the forced-recall instructions, have interfered with the ability of hypnosis to enhance recollection?



Perhaps, but given that the forced-recall procedure was common to both treatment conditions, why should we expect greater interference in hypnosis than in the waking condition?

Similarly, several studies contrasting hypnotic and waking memory using free recall, which apparently does not engender substantial intratrial variation in retrieval effort, have also failed to find an advantage for hypnosis (e.g., Lytle & Lundy, 1988; Nogrady et al., 1985; Sheehan, 1988). Nor is the finding of the present experiments of increased intrusions, associated with the use of hypnosis, consistent with a possible reduction in retrieval effort. Although they represent incorrect recall, the subject tenders such intrusions as true recollections. Finally, the forced-recall procedures administered here are the standard ones used in Erdelyi's laboratory (e.g., Erdelyi & Becker, 1974; Erdelyi & Kleinbard, 1978), which actually consist of a free-recall component followed by a guessing component (see Procedure section of Experiment 1 or 2). Erdelyi et al. (1989) demonstrated that this procedure minimizes differences in retrieval processing strategies on free versus forced-recall tasks. Thus, there is little reason to believe that the methodology employed in the current research was responsible for the lack of superiority of hypnosis over waking recall.

Findings from both experiments revealed that the hypnotic procedure introduced liabilities for memory retrieval. Thus, in Experiment 1, using a cumulative recall index, hypnosis was associated with a decreased rate of new item recoveries (reminiscences) over trials, relative to the nonhypnotic treatment. In Experiment 2, where hypnosis was introduced for the first time on the eighth and final recall trial, nearly twice as much new incorrect information (intrusions) was reported relative to new correct recall. In addition, the magnitude of new incorrect recall reported in hypnosis was substantially greater than that reported during any of the preceding waking trials.

In both experiments, confident reporting of memory intrusions increased with hypnosis, especially among the more hypnotizable subjects. This finding is consistent with a large body of literature on the effects of hypnosis, which shows that the more dramatic outcomes occur among individuals who



experience hypnosis vividly. It is all the more striking, therefore, that the highly hypnotizable subjects in these experiments did not demonstrate a hypnotic hypermnesia distinct from that produced by repeated waking recall attempts. All in all, it is clear that introducing hypnosis to refresh memory not only fails to increase accurate memory reports but also significantly augments the likelihood that pseudomemories will be reported with some confidence.




Baker, R. A., Haynes, B., & Patrick, B. S. (1983). Hypnosis, memory, and incidental memory. American Journal of Clinical Hypnosis, 25, 253-262.

Barber, T. X., & Calverley, D. S. (1966). Effects on recall of hypnotic induction, motivational suggestions, and suggested regression: A methodological and experimental analysis. Journal of Abnormal Psychology, 71, 169-180.

Belmore, S. M. (1981). Imagery and semantic elaboration in hyperm nesia for words. Journal of Experimental Psychology: Human Learning and Memory, 7, 191-203.

Cooper, L. M., & London, P. (1973). Reactivation of memory by hypnosis and suggestion. International Journal of Clinical and Experimental Hypnosis, 21, 312-323.

Dhanens, T. P., & Lundy, R. M. (1975). Hypnotic and waking suggestions and recall. International Journal of Clinical and Experimental Hypnosis, 23, 68-79.

Dywan, J. (1988). The imagery factor in hypnotic hypermnesia. International Journal of Clinical and Experimental Hypnosis, 36, 312-326.

Dywan, J., & Bowers, K. S. (1983). The use of hypnosis to enhance recall. Science, 222, 184-185.

Erdelyi, M. H. (1982). A note on the level of recall, level of processing, and imagery hypotheses of hypermnesia. Journal of Verbal Learning and Verbal Behavior, 21, 656-661.

Erdelyi, M. H. (1984). The recovery of unconscious (inaccessible memories: Laboratory studies of hypermnesia. In G. Bower (Ed.) The psychology of learning and memory: Advances in research and theory (Vol. 18, pp. 95-127). San Diego, CA: Academic Press.



Erdelyi, M. H., & Becker, J. (1974). Hypermnesia for pictures Incremental memory for pictures but not words in multiple recall trials. Cognitive Psychology, 6, 159-171.

Erdelyi, M. H., Finks, J., & Feigin-Pfau, M. B. (1989). The effect of response bias on recall performance, with some observations on processing bias. Journal of Experimental Psychology: General, 118, 245-254.

Erdelyi, M. H., & Kleinbard, J. (1978). Has Ebbinghaus decayed with time? The growth of recall (hypermnesia) over days. Journal of Experimental Psychology: Human Learning and Memory, 4, 275-289.

Geiselman, R. E., Fisher, R. P., MacKinnon, D. P., & Holland, H. L. (1985). Eyewitness memory enhancement in the police interview: Cognitive retrieval mnemonics versus hypnosis. Journal of Applied Psychology, 70, 401-412.

Klatzky, R. L., & Erdelyi, M. H. (1985). The response criterion problem in tests of hypnosis and memory. International Journal of Clinical and Experimental Hypnosis, 33, 246-257.

Lytle, R. A., & Lundy, R. M. (1988). Hypnosis and the recall of visually presented material: A failure to replicate Stager and Lundy. International Journal of Clinical and Experimental Hypnosis, 36, 327-335.

McConkey, K. M., & Kinoshita, S. (1988). The influence of hypnosis on memory after one day and one week. Journal of Abnormal Psychology, 97, 48-53.

McConkey, K. M., & Nogrady, H. (1984). Hypnosis, hypnotizability, and story recall. Australian Journal of Clinical and Experimental Hypnosis, 12, 93-98.

Mingay, D. J. (1986). Hypnosis and memory for incidentally learned scenes. British Journal of Experimental and Clinical Hypnosis, 3, 173-183.



Nogrady, H., McConkey, K. M., & Perry, C. (1985). Enhancing visual memory: Trying hypnosis, trying imagination, and trying again. Journal of Abnormal Psychology, 94, 195-204.

Orne, M. T. (1962). On the social psychology of the psychological experiment: With particular reference to demand characteristics and their implications. American Psychologist, 17, 776-783.

Orne, M. T., Soskis, D. A., Dinges, D. F., & Orne, E. C. (1984). Hypnotically induced testimony. In G. L. Wells & E. F. Loftus (Eds.), Eyewitness testimony: Psychological perspectives (pp. 171-213). New York: Cambridge University Press.

Orne, M. T., Whitehouse, W. G., Dinges, D. F., & Orne, E. C. (1988). Reconstructing memory through hypnosis: Forensic and clinical implications. In H. M. Pettinati (Ed.), Hypnosis and memory (pp. 21-63). New York: Guilford Press.

Payne, D. G. (1986). Hypermnesia for pictures and words: Testing the recall level hypothesis. Journal of Experimental Psychology: Learning, Memory, and Cognition, 12, 16-29.

Payne, D. G. (1987). Hypermnesia and reminiscence in recall: A historical and empirical review. Psychological Bulletin, 101, 5-28.

Putnam, W. H. (1979). Hypnosis and distortions in eyewitness memory. International Journal of Clinical and Experimental Hypnosis, 27, 437-448.

Register, P. A., & Kihlstrom, J. F. (1987). Hypnotic effects on hypermnesia. International Journal of Clinical and Experimental Hypnosis, 35, 155-170.

Roediger, H. L., Payne, D. G., Gillespie, G. L., & Lean, D. S. (1982). Hypermnesia as determined by level of recall. Journal of Verbal Learning and Verbal Behavior, 21, 635-655.



Sheehan, P. W. (1988). Memory distortion in hypnosis. International Journal of Clinical and Experimental Hypnosis, 36, 296-311.

Shor, R. E., & Orne, E. C. (1962). Harvard Group Scale of Hypnotic Susceptibility, Form A. Palo Alto, CA: Consulting Psychologists Press.

Smith, M. C. (1983). Hypnotic memory enhancement of witnesses: Does it work? Psychological Bulletin, 94, 387-407.

Stager, G. L., & Lundy, R. M. (1985). Hypnosis and the learning and recall of visually presented material. International Journal of Clinical and Experimental Hypnosis, 33, 27-39.

Udolf, R. (1983). Forensic hypnosis: Psychological and legal aspects. Lexington, MA: Lexington Books.

Weitzenhoffer, A. M., & Hilgard, E. R. (1962). Stanford Hypnotic Susceptibility Scale, Form C. Palo Alto, CA: Consulting Psychologists Press.

White, R. W., Fox, G. F., & Harris, W. W. (1940). Hypnotic hypermnesia for recently learned material. Journal of Abnormal and Social Psychology, 35, 89-103.

Whitehouse, W. G., Dinges, D. F., Orne, E. C., & Orne, M. T. (1988). Hypnotic hypermnesia: Enhanced memory accessibility or report bias? Journal of Abnormal Psychology, 97, 289-295.

Young, P. C. (1925). An experimental study of mental and physical functions in the normal and hypnotic states. American Journal of Psychology, 36, 214-232.

Yuille, J. C., & McEwan, H. (1985). Use of hypnosis as an aid to eyewitness memory. Journal of Applied Psychology, 70, 389-400.



Author Notes

This research was supported in part by Grants MH19156 and MH44193 from the National Institute of Mental Health, U.S. Public Health Service, and in part by a grant from the Institute for Experimental Psychiatry Research Foundation.

We wish to thank Mary F. Auxier, Barbara R. Barras, Lindsay S. Carota, Noel F. Carota, Stephen R. Fairbrother, and Ann M. Whitehouse for their many contributions to the execution of these experiments.

Correspondence concerning this article should be addressed to David F. Dinges, Unit for Experimental Psychiatry, The Institute of Pennsylvania Hospital, 111 North 49th Street, Philadelphia, Pennsylvania 19139-2798.




1 Both scales are standardized research instruments widely used to evaluate the capacity to respond to hypnotic suggestions. Each begins with the establishment of rapport with the subject, followed by a hypnotic induction and administration of 12 suggestions aimed at inducing a range of motor, perceptual, and memory alterations historically identified as phenomena of hypnosis. The HGSHS:A is a tape-recorded scale designed as a convenient preliminary assessment for use with groups of subjects. The SHSS:C is individually administered and includes some suggestions of greater difficulty than those found on the HGSHS:A. With either scale, scores can range from 0 (no suggestions passed) to 12 (all suggestions passed); the higher the score, the greater the person's judged ability to experience hypnosis.

2 Because the present experiment lacks an independent control condition, the question of whether a single hypnosis trial would benefit retrieval following multiple waking recall opportunities was also evaluated by comparison with the performance of subjects in the waking condition of Experiment 1. This was deemed justifiable because both subject samples were obtained from the same hypnotizability screening sample and the two experiments were run concurrently. Most important, the two groups of subjects were exposed to identical procedures throughout both sessions of the study, with the single exception of the R8 trial. Thus, we compared the amount of new correct recall (M = 0.76 item) obtained with the use of hypnosis on Trial 8 of Experiment 2 with that produced on Trial 8 of Experiment 1 by subjects in the waking condition (M = 1.00 item). This slight difference was not statistically significant, once again indicating a failure of hypnosis to enhance recall.



Table 1

Number of New Items Reported During Consecutive Waking Trials (R3-R7) and During Hypnosis (R8) and Correlations With Hypnotizability

New recall R3 R4 R5 R6 R7 R8


1.81 2.00 1.07 1.22 0.76
SD 1.65 1.45 1.72 0.91 1.01 0.89
r 0.13 -0.09 0.07 0.12 -0.11 -0.03
M 1.15 0.73 0.66 0.63 0.32 1.42
SD 1.33 0.92 1.33 1.09 0.65 1.91
r 0.07 0.19 0.10 -0.24 -0.14 0.33*

* p <.05.



Figure Captions

Figure 1. Mean number of correctly recalled pictures on each of Trials R1 through R8 as a function of hypnotic ability (low vs. high) and recall treatment (waking vs. hypnosis).

Figure 2. Mean cumulative new correct recall and intertrial forgetting over Trials R3 through R8 as a function of recall treatment (waking vs. hypnosis).

Figure 3. Mean number of confidently reported but incorrectly recalled pictures (intrusions) on each of Trials R1 through R8 as a function of hypnotic ability (low vs. high) and recall treatment (waking vs. hypnosis).

Figure 4. Mean number of correctly recalled pictures (independent of confidence) and confidently reported but incorrectly recalled pictures (intrusions) on each of Trials R1 through R8.












The preceding paper is a reproduction of the pre-publication submission to the Journal of Experimental Psychology: Learning, Memory, and Cognition (Dinges, D. F., Whitehouse, W. G., Orne, E. C., Powell, J. W., Orne, M. T., & Erdelyi, M. H. Evaluating hypnotic memory enhancement (hypermnesia and reminiscence) using multitrial forced recall.) It was eventually published in Journal of Experimental Psychology: Learning, Memory, and Cognition, 1992, 18, 1139-1147. American Psychological Association © 1992. It is reproduced here with the kind permission of the American Psychological Association. No further reproduction or distribution of this article is permitted without written permission of the publisher. Please consult the published version to cite quotations.