SIGNS OF ANXIETY DURING THREE VERBAI. ASSOCIATION CONDITIONS Dissertation for the Degree ‘of Ph. D. MICHIGAN STATE UNIVERSITY DONNEL B. STERN 1975 IIIII IIIIIIII l ...,.....e...,.,, 0321 4502 Mg . u LIBR/NL I 3 1293 1 chhir on Si U ‘ ref my fl . fi ) ‘ \ thesis entitled ' I SIGNS OF ANXIETY DURING THREE VERBAL ASSOCIATION CONDITIONS presented by DONNEL BERNARD STERN has been accepted towards fulfillment of the requirements for PH.D. degreein PSYCHOLOGY 0-7639 W MINT“ amnmvm LIBRARY BINDER ABSTRACT SIGNS OF ANXIETY DURING THREE: VERBAL ASSOCIATION CONDITIONS’ By Donnel B. Stern 0n the basis of the concepts of H. S. Sullivan and others, it was predicted that subjects would spontaneously report more psycho- pathology during verbal association when they were uncertain about the experimenter's evaluation of their efforts and when they believed the experimenter disapproved than when they believed the experimenter approved of their efforts. In a factorial design, thirty-six male and thirty-six female undergraduate subjects each participated in one of three verbal association treatments: 1) Uncertainty--subjects given no indication of experimenter evaluation; 2) Positive Evaluation-- subjects led to believe that the experimenter approved their efforts; 3) Negative Evaluation--subjects led to believe that the experimenter disapproved their efforts. Two male and two female undergraduate experimenters, blind to the experimental hypotheses, were each respon- sible for three of the male and three of the female subjects assigned to each treatment. In a paper completed after the initiation of the present study (Stern and Reyher, 1975; see Appendix A), reactions scorable by the symptomatic reaction scale (SRS), previously termed psychopathology, Donnel B. Stern were re-labelled "signs of anxiety.“ Because of the similarity of the procedures and data of the two studies, scorable SRS reactions in the present study were also termed anxiety signs. Experimental hypotheses remained the same. The results provided little support for the hypotheses as initially formulated. There were no differences in number of anxiety signs reported by subjects in the three treatment groups. However, consistent with the hypotheses was the finding that subjects who reported self-esteem reduction under-positive evaluation conditions repeated the sign with significantly lower frequency than did subjects in the no evaluation and negative evaluation grdups. Four signs of anxiety were reported with high enough frequency for further statistical analysis. For females, verbally expressed anxiety, sleepiness. and self-esteem reduction were significantly positively related to total SRS score. while inability to think was unrelated. For males. sleepiness was significantly negatively related to total SR5 score, while the other three signs were uncorrelated with SRS total. These analyses were carried out separately with analysis of variance and point-biserial correlation. The difference between the male and female correlations was significant for sleepiness in both analyses and for self-esteem reduction in the analysis of variance. It was concluded that sex was a moderating variable in the relationship of self-esteem reduction, verbally expressed anxiety. and sleepiness to total SRS score. For women only, these findings indicate that, con- sistent with Sullivanian theory, loss of self-esteem and expressed Donnel B. Stern anxiety bear a vital relationship to total anxiety. It remains to be determined whether. as Sullivan predicted. this relationship is a causal one. For both frequency of report of anxiety signs (regardless of the number of times an SRS category was repeated) and number of SRS categories reported, males and.females showed heterogeneous variances. While males tended to react to experimental conditions with low or moderate anxiety. females reacted with low or high anxiety. The greater number of positive part-whole correlations for females than males may have been due to this distribution, since the presence of any SRS cate- gory in a female's protocol, because.of greater female variance. may have led to more predictive power. Additional analyses revealed a significant curvilinear rela- tionship between number of words spoken by subjects and number of anxiety sign-types (i.e.. the.number of different SRS categories) they reported. Subjects with low and high fluency reported few anxiety signs, while subjects with mid-range fluency reported large numbers of signs. It was shown that even though their mean anxiety scores were nearly equivalent, high-fluency.subjects were effectively less anxious than low-fluency subjects. The curvilinear relationship was specula- tively interpreted as a representation of the dependency of chronic self-esteem level and anxiety during verbal association. The variances representing number of words spoken were heter- ogeneous across the three.treatments. The two equally large vari- ances were obtained in the positive and negative evaluation groups; thus, when evaluative feedback was available, some subjects' fluency Donnel B. Stern rose. Speculatively, security operations change when such feedback ~becomes unavailable,.because.self-esteem maintenance requires that one be able to judge the effect of one's behavior on others. SIGNS OF ANXIETY DURING THREE VERBAL ASSOCIATION CONDITIONS By \ 0v?“ Donnel Becgtern A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1975 This was the easiest decision of all: to Annie. ACKNOWLEDGMENTS To Dr. Joseph.Reyher. my committee chairman, supervisor, mentor.and.friend, I have already expressed my thanks. No matter who I meet.in my future career, I have no doubt that his influence on my professional life will remain greatest. Dr..Don Grumman has also been involved in much of my graduate training. He always.read-my work promptly and carefully and came up with concrete, usable suggestions which he had the knack of communicat- ing noncritically. His support, capacity to listen openmindedly, and appreciation of enthusiasm. all of which he seems to maintain without effort, have been assets to the working relationship I have had with him. Dr. Norman Abeles.has been on both my theSis and dissertation committees, and supervised me.for a year. His willingness to read my dissertation during the last hectic days before he left for Europe was greatly appreciated. He also led me to think critically about ethical issues and psychotherapy research. Dr. A. I. Rabin has had a moderating effect on me throughout graduate.school. During his excellent Rorschach course, I learned to make responsible psychodynamic prediction, which has helped immensely in psychotherapy. As a committee member, Dr. Rabin applied the same critical.faculty to my research. leading to some hard thinking on my part about the conceptualization of my hypotheses and dependent variables. iii ..... Perhaps most.of.all.I thank Barry Kennedy. Linda Whitney. John Pety,.and Laurie Jentleson..who worked indefatigably and without whom there.would have been.no data. -I enjoyed knowing them and will always appreciate their interest in the dissertation. iv TABLE OF CONTENTS /, Page LIST OF TABLES . ................. . . . . . . vi LIST OF FIGURES ....................... ix INTRODUCTION . . ....................... I METHOD . ................ . ..... . . . . . IO RESULTS . .......................... 17 DISCUSSION . . . . . . . ................... 59 SUMMARY . ..................... .. . . . . 69 REFERENCES . . ..... . ........... . . . . . . . 72 APPENDICES A. Anxiety manifestations during verbal association and visual imagery . ...... . . . . . ..... 82 B. Questionnaire . . ............ . . . . . . . 100 C. Symptom descriptions for the scoring of free verbal material (Symptomatic Reaction Scale) . . . . 101 D. Representative transcripts I? ............. 106 E. Raw data . . . . ..... . ........ . . . . . lZl F. Literature review: Experimental psychpathology . . . . 125 LIST OF TABLES Table Page 1. Raw-score means, variances andF a for experimental treatments, experimenters, aid §ex of subject. Variable 1: Number of sign- -types . . . . . . . . . . 19 2. Transformed-score (X = /X + .5) means, variances and F x for experimental treatment, experi- menteyg, and sex of subject. Variable 1: Number of sign— -types ... . . . . . . . . . . . . . . . . . . 20 3. Raw-score means, variances, andF for experimental treatments, experimenters and-QQR of subject. Variable 2: Frequency of signs . . . . . . . . . . . 2l '4. Transformed-score (X = /X + .5) means, variances and F x for experimental treatments, sex of subjEEi and experimenter. Variable 2: Frequency of signs . . . . . . . . . . . . . . . . . . . . . . 22 5. Frequency of signs before and after the subtraction of totals for inability to think, verbally expressed anxiety, and self-esteem reduction . . . . 24 £3. Total number of sign-types before and after the subtraction of totals for inability to think, verbally expressed anxiety, and self-esteem reduction . . . . ... . . . . . . . . . . . . . . . . 25 7'. Proportion of positive evaluation subjects vs. no evaluation + negative evaluation subjects who reported inability to think, verbally expressed anxiety, and self-esteem reduction . . ....... 26 8 - Mann-Hhitney U tests for sign frequency in positive evaluation treatment vs. no evaluation + nega- tive evaluation treatments . . . ....... . . . 27 9. I Analysis of variancefor total number of sign- types by experimental treatment, sex of subject, and experimenter . . ...... . . . . ..... . 31 vi Table I0. II. I2. I3. I4. I5. I6. I7. I8. I9. 20. Analysis of variance for sign frequency by experimental treatments, experimenters, and sex of subject ................... Raw-score means, variances, andF for experi- mental treatments, experimentewg, and sex of subject. Variable 3: Number of words spoken Raw-score means and.variances of number of sign- types reported by subjects.who spoke 000-199 words, 200-399 words, 400-599 words, 600-799 words, 800-999 words,.and 1,000 words or more Raw-score means and variances of sign frequency for subjects who spoke.000-199 words, 200-399 words, 400-599 words, 600-799 words, 800-999 words, and 1,000 words or more ........... Transformed-score (X = /X + .5) means and vari- ances of number of sign-types reported by subjects who spoke 000-199 words, 200-399 words, 400-599 words, 600-799 words, 800-999 words, and 1,000 words or more .............. Analysis of variance of linear and nonlinear components in the relationship of fluency and number of sign-types ................ Comparison of sign- type totals divided by number of words spoken for low (000- 399 words), medium (400-799 words), and high (800 or more words) fluency subjects using the method of adjusted significance levels ................ I Transformed-score (X = JX + .5) means and vari- ances of number of sign-types for subjects who reported at least one anxiety sign ..... . . . . Analysis of variance of the effects of sex and the presence of sleepiness report on number of sign- types ........ . .............. Analysis of variance.of.the effects of sex and the presence of report of inability to think on number of sign-types ................ Analysis of variance of the effects of sex and the presence of verbally expressed anxiety on number of sign-types ................... vii Page 34 35 4O 40 4I 43 45 47 49 49 50 Table Page 21. Analysis of variance of.the.effects of sex and the presence of report of self-esteem reduction on number of sign-types... ............... 50 22. Correlation of sign-type with total score ....... A 55 viii Table Page 21. Analysis of variance of the effects of sex and the presence of report of self-esteem reduction on number of sign-types.. . . . . . . . ........ 50 22. Correlation of sign-type with total score ....... g 55 viii Figure LIST OF FIGURES Distribution of.total number of sign-types by sex . . . . Distribution of sign.frequency by sex ........ . . Distribution of fluency across experimental treat- ment . . . . .................... Raw-score mean number of sign-types for subjects who spoke 000-199.words, 200-399 words, 400-599 words, 600-799 words, 800-999 words, and 1,000 words or more .................... Raw-score mean sign frequency for subjects who spoke 000- 199 words, 200- 399 words, 400- 599 words, 600-799 words, 800- 999 words, and 1,000 words or more ........................ Transformed-score (X = /X + .5) mean number of sign-types for subjects who spoke 000-199 words, 200-399 words, 400-599 words, 600-799 words, 800- 999 words, and 1,000 words or more . . . ...... Interaction effects of sex of subject and presence of self-esteem reduction on number of sign-types . . . Interaction effect of sex of subject and presence of sleepiness on number of sign-types ....... ix Page 29 32 36 38 39 42 52 53 INTRODUCTION ‘ In.a previous study comparing the psychotherapeutic uncovering properties of verbal association and visual imagery (Stern and Reyher, 1974), the writer was surprised.to observe that subjects spontaneously reported behaviors and experiences traditionally known as psychopatho- logical symptoms. ,The symptoms.ranged from reports of sleepiness, sen- sations of heat and cold, numbness, floating and rocking, and expressions of intense.anxiety to obsessive-compulsive and paranoid thought. In all, thirty-one of thirty-six subjects reported some symptoms, and most reported many.varieties. Each subject participated in ten minutes of each of three conditions: .verbal association with eyes open, verbal association with eyes closed,.and visual imagery (eyes closed).. The. order of presentation of experimental conditions was completely counter- balanced, and there were no significant order effects. The experimenter sat behind the subject, who reclined in a lounge chair, and the experi- menter spoke only to give instructions. A significantly greater number of symptoms were reported during both verbal association conditions than during the visual imagery condition. ,. The importance.of.experimental methods which can reliably produce psychopathology has been recently emphasized (Cohen, 1974). If the dimensions of the experimental conditions which are necessary and sufficient elicitors of psychopathology are isolated, new information about precipitating stress, and eventually about etiology, may be obtained. The purpose of the present study is to further establish that a simple verbal association technique reliably produces psycho- pathology and to begin to identify the variables in this procedure most responsible for the reports.of.symptomatic experience. Much of the previous work in human experimental psychopath- ology (i.e., studies in which the symptoms to be observed are produced experimentally) has.been aimed at an understanding of psychosomatic symptoms.(e.g., Alexander, 1962). The full range of human psycho- pathology has seldom been conceptualized as a dependent variable, and most of the experiments of this.type which have been done do not employ concepts useful in understanding the results of Stern and Reyher's verbal.association paradigm. .The sensory deprivation literature, for instance, is full of rich anecdotal accounts of a wide variety of psychopathology, but.only.imagery, or hallucination, has been systema- tically.studied (Freeman, Grunebaum, and Greenblatt, 1961; Kubzansky and Leiderman, 1961; Ruff, 1966). Furthermore, the rubric "sensory deprivation" is an umbrella term.covering dozens of variables, and it is still unclear which of these.variables are essential to the produc- tion of typical sensory deprivation effects (Ruff, Levy and Thaler, 1961; Cohen, Silverman, Bressler and Shmamovian, 1961). Studies of sleep.deprivation have also.reported a broad range of psychopathology (Luby and.Gottlieb, 1966),.but contrary to some investigators' initial notions.(Dement,.1966; Dement,.Greenberg and Kline, 1966), sleep deprivation.is no longer considered to be an important precipitating stress in acute psychosis.(Dement, Cohen, Ferguson and Zarcone, 1970). Discovery of the psychotomimetic effects of certain drugs (LSD, mescal, Sernyl) led some researchers.to assert that if similar substances could be isolated in the bloodstreams of psychotic persons, the riddle of psychosis would be solved.(e.g., Smythies, 1962). The logic of this argument.is questionable, since.it does not follow that the capa- city of.drugs to produce psychotic-like states means that such drugs cause psychosis under natural conditions. Furthermore, careful clini- cal observation soon.showed.that.drug-induced and psychotic symptoms were qualitatively different.(Hollister, 1964). Of the experimental methods which produce a broad range of psychopathology under controlled.conditions, the most clinically rele- vant is the method of posthypnotic.stimulation of hypnotically induced conflict (Moore, 1964; Perkins and.Reyher, 1971; Reyher, 1961; 1962; 1967; Sommerschield and Reyher, 1973; Burns and Reyher, in press). This technique allows the psychodynamically oriented investigator to activate oedipal, aggressive,.and other conflicts and then observe whether.or not the activation has the psychopathological consequences predicted by the theory.. The.effectiveness of the paradigm seems to depend on the intensification of drive, which strains the capacity of defenses.to keep conflict unconscious.. Explanatory concepts from this line of research therefore do not apply equally well to the symp- toms reported during the verbal association technique used by Stern and Reyher. Evidence that.drive intensification was not an important causal factor is provided by other data collected in the study: using a modification of Holtis tenth.edition of a manual for the scoring of manifestations of the primary process on the Rorshach (1970), neither drive blatancy nor amount of primary process content was higher in one experimental condition than in another. If either of these variables had occurred.more frequently during the verbal association conditions, it could have been argued that the greater number of symp- tom types reported during verbal association was due to the presence of a greater number.of.anxiety-provoking derivatives of unconscious conflict. However, since the number of symptom types in a condition did not co-vary with the strength of drive, the intrapsychic hypothesis can be discarded. One conspicuous aspect of Stern and Reyher's study was the lack of.re1evant information available to the subject. Because of the simple, nonspecific instructions and the planned silence of the experi- menter, subjects had no way.to know if they were satisfying the experi- menter's expectations of them. .Also, unlike all previous studies from this laboratory, subjects reported many indications of loss of self- esteem (feeling dumb, stupid,.use1ess,-etc.) and spent much of their half-hour in the experiment.trying to figure out what the experimenter wanted of them. Adding to their discomfort, the experimenter sat behind the subject, outside his field of vision. .One way to organize these facts into a coherent causal hypothesis.of the psychopathology reported by subjects is provided by Sullivan's (1953) concept of.security.operations. Along with many ' other.neo-Freudian theorists, Sullivan posited a central role for self-esteem. Security operations are those activities (or lack of activity, e.g., selective inattention) undertaken by the self to pro- tect and.maintain self-esteem. The.anxiety concerning loss of self- esteem and the activities required to allay it are so important to the self that they supersede all other feelings and psychological functions. Stable, traditional relationships have familiar security operations built into them. A student, for instance, plays his role by performing the tasks set before him by the instructor and derives part of his self-esteem from the assurance that he has satisfied the instructor's demands. Since the situation was so unusual and unstruc- tured, though, subjects in the.Stern and Reyher study found that security operations appropriate.and effective in other relationships simply did not work in this one.. The experimenter did not interact with subjects, so subjects could not directly appease him or accede to his wishes; and how to complete.the task was so unclear that subjects had no way to know when.they.had.done.so. Subjects could not even see the experimenter's facial.expressions, a source of interpersonal stimu- lation important to security.operations. Thus, self-esteem was threat- ened, and most subjects then reported anxiety and other symptoms. This explanation accounts.for.the.greater.number of symptom types reported during Stern and Reyher’s.verba1 association conditions than during the visual.imagery condition in terms of degree of threat to self- esteem. Subjects may have felt that they had satisfied task demands (and the demands of their.own.security operations) when they reported a visual image and described it to the experimenter. Verbal association, though, is a less novel task; consequently, task demands were less discriminable, and subjects could not perceive when they had given the experimenter what he wanted. One prediction which can be made from this point of view is that, if subjects' fears of experimenter the self that they supersede all other feelings and psychological functions. Stable, traditional relationships have familiar security operations built into them. A student, for instance, plays his role by performing the tasks set before him by the instructor and derives part of his self-esteem from the.assurance that he has satisfied the instructor's demands. Since the situation was so unusual and unstruc- tured, though, subjects in.the.Stern and Reyher study found that security.operations appropriate.and effective in other relationships simply did not work in this one.. The experimenter did not interact with subjects, so subjects could not directly appease him or accede to his wishes; and how to complete the task was so unclear that subjects had no way to know when.they had.done so. Subjects could not even see the experimenter's facia1.expressions, a source of interpersonal stimu- lation important to security operations. Thus, self-esteem was threat- ened, and most subjects then reported anxiety and other symptoms. This explanation accounts.for.the.greater.number of symptom types reported during Stern and Reyheris.verbal association conditions than during the visual imagery condition in terms of degree of threat to self- esteem. Subjects may have felt that they had satisfied task demands (and the.demands of their own.security operations) when they reported a visual image and described it to the experimenter. Verbal association, though, is a less novel task; consequently, task demands were less discriminable, and subjects could not perceive when they had given the experimenter what he wanted. One prediction which can be made from this point of view is that, if subjects' fears of experimenter disapproval were.confirmed,.subjects' psychopathology would be even more severe and frequent than it was during the Stern and Reyher study. In other words, certainty.about-experimenter disapproval is more patho- genic.than uncertainty about the.experimenter's evaluation. This posi- tion can be named the.certajnty hypothesis. However, a case for the opposite point of view (i.e., that gnoertainty.concerning.experimenter evaluation is more pathogenic than certainty of disapproval) can be.made. Several theorists have given uncertainty or concepts closely related to uncertainty a prominent role in their writings. Berlyne (1957; 1970) defines objective (meas- urable).uncertainty as the.presence of novel stimuli, and goes on to define the degree.of novelty of a-stimu1us as the amount of information it contains. .Because.its termination can be used as a reinforcer, uncertainty is aversive. .Subjective (felt) uncertainty can be dealt with by.leaving.the.situation,.by.staying in it and habituating to the uncertainty, or by staying and.gathering.enough information to reduce the novelty.(or.uncertainty)...Berlyne does not speculate about what would happen if a person were.faced with an inescapable situation with a high degree of uncertainty.and no method of obtaining information, but the.uncertainty hypothesis used in the present study predicts that such conditions may.lead.to.anxiety and other psychopathology. In support of this notion, Masserman.(197l) defines neurotogenic conflict (which he labels "uncertainty") as experiences in which "the organism apprehends a failure to predict and control events important to its welfare".(p..l9). Kelly.contends that personality is composed of “personal constructs," or hypotheses about the future based on past disapproval were confirmed,.subjects' psychopathology would be even more severe and frequent than.it.was during the Stern and Reyher study. In other words, certainty.about-experimenter disapproval is more patho- genic than ugpertainty about the.experimenter's evaluation. This posi- tion can be named the.certainty.bypothesis. However, a case for the opposite point of view (i.e., that uncertainty.concerning experimenter evaluation is more pathogenic than certainty of disapproval) can be made. Several theorists have given uncertainty or concepts.closely.related to uncertainty a prominent role in their writings. Berlyne (1957; 1970) defines objective (meas- urable) uncertainty as the presence of novel stimuli, and goes on to define the degree of novelty of.a-stimulus as the amount of information it contains.. Because its termination can be used as a reinforcer, uncertainty is aversive.. Subjective (felt) uncertainty can be dealt with by.leaving the.Situation,.by.staying in it and habituating to the uncertainty, or.by staying and gathering enough information to reduce the novelty (or uncertainty). -Berlyne does not speculate about what would happen if a person were.faced with an inescapable situation with a high degree of uncertainty.and no method of obtaining information, but the uncertainty hypothesis used in the present study predicts that such conditions may lead.to anxiety and other psychopathology. In support of this notion, Masserman (1971) defines neurotogenic conflict (which he labels "uncertainty") as experiences in which "the organism apprehends a failure to predict and control events important to its welfare".(p. 19). Kelly contends that personality is composed of “personal constructs," or hypotheses about the future based on past experience; anxiety is described as "the recognition that the events with which one is confronted lie outside the range of convenience of one's construct system“.(p..495).. Typical cognitive and perceptual distortion and dysfunction during sensory deprivation have been said to be due to the difficulty.subject.has in imposing on the novel environ- ment the.previously learned.cognitive.models normally used to structure (make sense of) experience (Freedman, Grunebaum and Greenblatt, 1961; Bruner, 1961). In explicating this view, Bruner says: One may suggest that one of the prime sources of anxiety is a state in which oneis.conception or perception of the en- vironment with which.one must.dea1 does not "fit" or predict that environment in a.manner.that.makes action possible. If there is anything.to.this.view of anxiety, then it follows that when one prevents an.organism from monitoring the fit- tingness of.his percepts and his cognitive structures, one is cutting him off from one of his principal sources of adjustment.(p..206)...... Preventing an organism from.monitoring the fittingness of certain per- cepts and cognitive structures is exactly what a highly ambiguous situ- ation.like.the verbal association technique used by Stern and Reyher accomplishes. Most.of.the theorists.reviewed would agree that behavior and experience are constantly adjusted and monitored by commerce with the external environment, and that with the lack of ability to predict the effect.of one's actions (uncertainty) may come anxiety and psycho- pathology. The uncertainty hypothesis used here views the greater number of symptoms in Stern.and Reyher's verbal association conditions as a consequence of a higher.degree of uncertainty, i.e., the difficulty in perceiving task demands.. In.other words, the uncertainty experienced by subjects concerning the experimenter's evaluation of them was more pathogenic than a direct communication or disapproval would have been. experience; anxiety is described as "the recognition that the events with which one is confronted lie outside the range of convenience of one's construct system" (p. 495). Typical cognitive and perceptual distortion and dysfunction.during sensory deprivation have been said to be.due to the difficulty subject has in imposing on the novel environ- ment the.previously learned cognitive models normally used to structure (make sense of).experience (Freedman, Grunebaum and Greenblatt, 1961; Bruner, 1961). In explicating this.view, Bruner says: One may suggest that one of the prime sources of anxiety is a state in which oneis.conception or perception of the en- vironment with which.one must deal does not "fit" or predict that environment in a.manner.that makes action possible. If there is anything to this.view of anxiety, then it follows that when one prevents an organism from monitoring the fit- tingness of his percepts and his cognitive structures, one is cutting him off from one of his principal sources of adjustment (p. 206)...... . Preventing an organism from monitoring the fittingness of certain per- cepts and cognitive structures is exactly what a highly ambiguous situ- ation like the verbal association technique used by Stern and Reyher accomplishes., Most.of.the theorists reviewed would agree that behavior and experience are constantly.adjusted and monitored by commerce with the external environment, and that with the lack of ability to predict the effect of one's actions (uncertainty) may come anxiety and psycho- pathology. .The uncertainty hypothesis used here views the greater number of symptoms in Stern.and Reyher's verbal association conditions as a consequence of a higher.degree of uncertainty, i.e., the difficulty in perceiving task demands.. In.other words, the uncertainty experienced by subjects concerning the experimenter's evaluation of them was more pathogenic than a direct communication or disapproval would have been. This point of view can.be.appreviated.as the uncertainty hypothesis. One question asked.by.this study was: Which is more patho- genic, inability to predict (uncertainty concerning) the experimenter's evaluation (with the.probable.fear.of-negative experimenter evaluation), or certainty of negative experimenter evaluation? According to the certainty hypothesis, subjects.who.believe the experimenter disapproves of their.efforts spontaneously report more descriptions of psycho- pathology than do subjects given no indication of experimenter evalua- tion...The.uncertainty.hypothesis predicts the opposite, i.e., that uncertainty leads to more descriptions of psychopathology than certainty of negative evaluation. .Both hypotheses, because they concern the importance of self-esteem maintenance, predict that positive evaluation from the.experimenter.does not-1ead.to.reports of psychopathological symptoms.. However, the factors of eye closure, the experimenter's silence, the.room's soundproofing,.and the subject's restriction to the lounge chair add up to.potential sensory deprivation conditions. Also, minimal experimental evidence suggests that the introspection of normal subjects may 1ead.to the perception of “bizarre bodily expe- riences“ (Barber,.l969)...For these reasons, I have not hypothesized an absence of symptoms under conditions of.positive experimenter evalu- ation. Rather, the hypotheses.are couched in comparative terms: .1. More descriptions.of symptoms are contained in the verbal associations of subjects.who believe their efforts cause experi- menter disapproval (certainty condition) than in the verbal associations of subjects who believe their efforts cause experimenter approval. 2. More descriptions of symptoms are contained in the verbal associations of subjects given no indication of experimenter evaluation (uncertainty condition) than in the verbal associations of subjects who believe their.efforts cause experimenter approval. .The study also explored the relative pathogenicity of the certainty.and uncertainty conditions, but insufficient information made it impossible to state a hypothesis regarding this relationship. METHOD Subjects Seventy-two subjects (thirty-six males and thirty-six females) were randomly selected from.a group of undergraduate volunteers.who had signed up to participate in an experiment titled "Thought processes." Experimenters Two male and two female experimenters, selected from a group of volunteers in an undergraduate clinical psychology class, were each randomly assigned nine male and nine female subjects. Of the six sub- jects each.experimenter.conducted-through each of the three experimental conditions, three were male.and three were female. Experimenters were blind to.the experimental hypotheses and to the more general fact that the experiment concerned psychopathology. Materials and experimental setting The experimental room was small, soundproofed, and windowless. A cot and a six channel Grass #5 polygraph were in storage in the room. The room also contained the large black lounge chair used by subjects and the experimenter's straight-backed chair. Procedure Since subjects volunteered (were not a random sample), random assignment to treatment level was essential (Kirk, 1972, p. 13). After IO II the subjects were randomly assigned to groups, they were randomly assigned (separately, by sex) to the experimenters. Subjects were contacted by telephone by the assigned experi- menter.and an appointment.was made. At the appointed time, the experi- menter met the subject and-conducted.him into the laboratory, where the experimenter suggested the subject put his books and/or coat "on the cot" and "sit down in the.big, black chair." After requesting permission to tape record ("If.you would like to participate, this research requires that the session be tape recorded."), the experimenter gave the subject the experimental instructions. Once the experiment was under way, the experimenter spoke only to give instructions. If the subject asked a question about the instructions after the experi- menter had given them, the instructions were repeated. If the subject asked.another.question, the experimenter said "I can't answer your questions." If the subject asked another question, the experimenter did not respond. Each subject attended one session in which he received one of three experimental treatments. In each condition, the subject was asked to associate verbally for ten minutes with eyes closed. Twenty- four subjects (twelve males and twelve females) received each experi- mental treatment. If.they asked,.subjects had been told that the experiment would take about an hour; they had no other estimate of the amount of time they would.spend verbally associating. The following instructions were read to subjects: No evalua- tion group (uncertainty regarding experimenter evaluation): I2 Now, with your eyes closed, I would like you to describe whatever thoughts and feelings occur to you. These thoughts and feelings may be about anything. .Just let your mind wander and tell me the thoughts and feelings that come to your attention without leaving out a thing. Now, go ahead. Positive evaluation condition (certainty of experimenter approval): Now, with your eyes closed, I would like you to describe whatever thoughts and feelings occur to you. These thoughts and feelings may be about anything. Just let your mind wander and tell me the thoughts and feelings that come to your attention without.leaving out a thing. I won't inter- rupt you as long as.you're doing just what I'd like you to do, because I don't want to disturb the natural process. Now, go ahead. Negative evaluation condition (certainty of experimenter disapproval): Now, with your eyesclosed, I would like you to describe whatever thoughts and feelings occur to you. These thoughts and feelings may be about anything. Just let your mind wander and tell me the thoughts and feelings that come to your attention without.leaving.out a thing. I'll interrupt you when you're doing just what I'd like you to do, because I'll want to ask questions and gain more information. Now, go ahead. It will be noted that all three groups received the same verbal association instructions.. The only difference in the three sets of instructions was the meaning.attributed to the experimenter's silence. After ten minutes of verbal association had elapsed, the experimenter interrupted the subject and.asked him to answer some ques- tions about.his experience in the experiment. The questions (Appendix B) were.designed to assess certain potential subtle demand characteris- tics.. For instance, the professor whose laboratory was used for data collection has performed many studies on psychopathology on this campus, and I wanted to make sure that subjects' expectations about the experi- ment had not been previously biased. Subjects were also asked what they thought the purpose of the experiment was, and an assessment of I3 whether or not subjects had believed the evaluations contained in the instructions was made.. Questions were phrased in a way that did not communicate what answer.the experimenter expected to receive. ,After completion.of the questionnaire, the experimenter informed those.in the negative evaluation condition that the experi- menter's silence had been planned and did not actually signify dis- approval of the subjects' efforts. All subjects were told that their efforts had been exactly what.the experimenter had been looking for, and were asked not to divulge the procedures of the experiment. - Finally,.subjects.were requested to read and sign a standard research consent form and sign a statement which, providing the data remained confidential, released the tape recordings for scientific purposes. Each experimenter transcribed the data he or she had collected. Experimenters.were instructed to include punctuation (including begin- nings and endings of sentences) only where the subjects' use of it was definite. No interpretive or "clarifying" punctuation was to be added. Scoring A revised version of the symptomatic reaction scale (Sommer- schield and Reyher, 1973; Burns and Reyher, in press) was used to score the data (Appendix D). All reactions included on the scale are opera- tionally.defined. The data were scored independently by two advanced graduate students in clinical psychology. One of the judges (the author) scored all the data, while the other scored one-third of it (i.e., transcriptions from twenty-four subjects). The protocols chosen 14 for the second judge were selected randomly, with the restriction that the selection contained two transcripts, one from a male subject and one from a female, from.the group of six subjects assigned to each of the four experimenters in each of the three experimental treatments ((2) (4) (3)=(24)). The transcribed data gave no indication of which experimental treatment a subject had been assigned, nor were the scorers aware of the sex of the subject and identity of the experi- menter unless.the.subject presented this information in his/her tran- scribed remarks. Although earlier research (Stern and Reyher, 1975; see Appendix A) and the original plan for the present study chose the sentence as the basic.unit.to.be scored, it was decided to use the "phrase" instead. Subjects'.spontaneous speech was seldom naturally divisible into.sentences, so the sentence method would have required that one person (for.the.sake of consistency) break the data into sentences,.which would then have been used for scoring purposes by both judges. The possibility was high that subject's manner of speech would differentially affect the way in which this "sentence-breaker" did his work. The result would have been a potential contamination of results, since an individual's average sentence length might have partially determined.the number of scores he or she received. "Phrases" were defined as speech segments set apart from their contexts by periods, commas, conjunctions, or silences. Rather than dividing the data into phrases previous to scoring, both scorers used the preceding rule to make their own decisions about scorable units. This procedure did away with the possible systematic bias 15 introduced by having one person, unchecked for reliability, decide upon the scoring units. In effect, then, the reliability coefficient to be reported below is.a measure of the degree to which scorers agreed on scorable units, as well as an indication of the similarity in the judges' applications of.the symptomatic reaction scale (SRS). While some of.the subjects described a particular type of SRS reaction only once during the experimental session, others repeat- edly described the same reaction. .Subjects who reported the same type of reaction also.showed variability in the number of words and amount of time spent in description (Stern and Reyher, 1975). Therefore, each subject received two total.scores:. l) the total number of phrases in which.any.SRS reaction was described, termed "anxiety sign frequency," or simply "sign frequency“;.2).the total number of different SRS reac- tions described, called.9number of anxiety sign-types" or "number of sign-types." In prior investigations, both these measures have been shown to be sensitive to experimental treatments (Sommerschield and Reyher, 1973; Burns and.Reyher, in press). If subjects of one sex or subjects who had been in one experi- mental treatment spoke more words than their counterparts, it could be objected that.their protocols would have a correspondingly higher probability of containing reports of.SRS reactions. Stern and Reyher (1975) reported a negligible correlation between number of sentences spoken and.number of SRS categories (r = .05); nevertheless, their use of the.potentially biased.sentence.unit made it necessary to control for fluency in the present study. Therefore, the number of words used 16 by each subject was calculated and was added as a third dependent variable. Several of the symptom categories refer to a subject's expression of a feeling.(e.g., ego alien affect, discomfort, stupidity, etc.). These.categories.were scored only when subjects indicated an association.of the feeling with the experimental conditions. So, for example, the statement "This room feels weird. It's eerie" is scorable as ego alien affect, but the statement "It feels weird to be in college" is not scorable. Only that data subject to independent observation were used, i.e.,.the experimenter made no behavioral observations. Therefore, for instance, blushing was scored only if it was mentioned by the subject. RESULTS General characterization of the data An implicit prediction of this study was that the procedures used were sufficient to produce psychopathology. Although the SRS was an appropriate tool to.use in scoring the data, many of its psycho- somatic.and "higher intensity" reactions, like hallucinations and delusions, were not reported-by.subjects in the present study. In fact, the reactions were sufficiently less frequent and more restricted in range.than those produced by other methods in the same laboratory (Perkins and.Reyher, 1971; Sommerschield and Reyher, 1973; Burns and Reyher, in press; Moses,-1974) that it would be a misnomer to term most of them "symptoms“ or "psychopathology" at all. Instead, they will be labeled "anxiety signs" throughout the remainder of this report. ..Reproductions.of illustrative transcripts appear in Appendix D, and Appendix E contains each subject's raw data. Reliability Interjudgment agreement was calculated differently for the two anxiety sign variables; all reliability coefficients were based on the twenty-four transcripts scored by both judges. For the first variable,.sign frequency, an agreement was defined as an instance in which.both scorers assigned the same type of SRS reaction to the same phrase. A disagreement was indicated when one judge scored a phrase 17 18 which the other did not, and when judges assigned the same phrase different scores. The.percentage of interjudgment agreement was repre- sented by the total.number.of agreements divided by the totalqpf the agreements and the disagreements. The resulting quotient, .83, indi- cates adequate reliability both for the SRS scoring and for judges' selection of the same kinds of basic units to be scored (phrases). .For the second variable, number of sign-types, an agreement consisted of the two scorers' assessments that a subject's total phrases contained one or more.instances of a particular sign. A disagreement was tallied when one scorer assigned a sign-type to a protocol and the other did not. Degree of.interjudgment agreement was represented by the total number of agreements divided by the total of the agreements and disagreements. Percentage of agreement was .97, indicating good reliability. Since adequate.reliability was demonstrated for both measures, the author's scoring was used in the following analyses. Experimental hypotheses A significant difference was not obtained when the pooled scores of the negative evaluation and no evaluation treatment groups were compared to the scores of subjects in the positive evaluation group. This lack of support for the hypotheses resulted for both dependent.variables, number of sign-types (t_= .93, gf_= 70, p_> .10) and sign frequency (t_= 1.60, gf_= 70, p_< .10). Neither were the negative evaluation and no evaluation treatment groups significantly different from one another. Again, both variables, number of sign-types 19 (t_= .25,.gj.= 46, p_> .10).and sign frequency (t_= .13, gf.= 46, p.> .10) failed to.differentiate the two groups. ..Although these analyses provide no support for the experi- mental hypotheses, Tables 1, 2, 3 and 4 (showing both the raw-score means and the transformed-score means which were later used for analy- sis of variance) indicate that, for both anxiety sign variables, dif- ferences were.in the predicted direction. These differences are almost wholly accounted for by group differences on three anxiety signs: Table 1. Raw-score means, variances andF for experimental treat- ments, experimenters, and sex ofmgObject. Variable 1: Number of sign-types. Mean Variance Emax Treatments No evaluation 3.50 8.35 Positive evaluation 2.92 8.34 1.13; gf_= 23, k_= 3 Negative evaluation 3.67 7.36 Sex of subject Female 3.72 10.72 * 2 l3 , gf_= 35,_5 = Male 3.00 5.03 Experimenters JP (Male) 2.56 6.73 BK (Male) 4.06 9.47 2.64; gf_= l7, k_= 4 LJ (Female) 3.72 11.15 LW (Female) 3.11 4.22 * p_< .05. 20 Table 2. Transformed- -score (X =JX+ 5) means, variances andF for experimental treatment, experimenters, and sex of Eflhject. Variable 1: Number of sign- types. Mean Variance Emax Treatments No evaluation 1.85 .61 Positive evaluation 1.68 .57 1.49; gf_= 23, k_= 3 Negative evaluation 1.94 .41 Sex of subject Female ” 1.89 .63 1.47; gj.= 35, k_= 2 Male. 1.76 .43 Experimenters JP (Male) 1.60 .51 BK (Male) 2.01 .57 1.55;<_i_f_=17,k_=4 LJ (Female) 1.91 .62 LW (Female) 1.78 .40 Table 3. Raw-score means, variances, and F , ' r a e treatments exper1mente 5 nd 5 Rme Frequency of signs. 21 for experimental subject. Variable 2: Mean Variance Emax Treatments No evaluation 6.96 57.35 Positive evaluation 4.17 31.36 1.83; gj_= 23, g.= 3 Negative evaluation 6.79 47.74 Sex of subject Female 6.89 66.39 ** 2.67 , _d_f_ 35, k_= Male 5.06 24.91 Experimenter. JP (Male) 3.83 19.79 BK (Male) 7.61 59.31 * 4.68 ; g__= 17, g_= LJ (Female) 7.17 84.62 LW (Female) 5.28 18.10 22 Table 4. Transformed- -score (X = JX+ 5) Means, variances andF x for experimental treatments, sex of subject and experimenEQE. Variable 2: Frequency of signs. Mean Variance Emax Treatments No evaluation 2.39 1.81 Positive evaluation 1.89 1.15 1.30; gf_= 23, §;= 3 Negative evaluation 2.46 1.32 Sex of subject Female 2.36 1.86 1.77; g:_= 35, k_= 2 Male 2.13 1.05 Experimenters JP (Male) 1.84 .98 BK (Male) 2.55 1.70 2.19; gf_= l7, k_= 4 LJ (Female) 2.38 2.12 LW (Female) 2.21 .97 23 inability to think, verbally expressed anxiety, and self-esteem reduction. Tables 5 and 6 illustrate that treatment group totals for both anxiety sign variables are nearly identical after these three signs have been subtracted. If most.of.the treatment group mean differences can be attrib- uted to these three signs, it is possible that individual examination of the signs.will.highlight differences which were overshadowed in the overall tests. Because of the direction of the experimental hypotheses, the tests to follow will compare the subjects of the positive evaluation group to the.combined subjects of the no evaluation and negative evalua- tion groups. Table 7 shows that the two comparison groups show no signifi- cant difference in the proportion.of subjects who reported inability to think, verbal anxiety,.or.self-esteem reduction (p_> .05). However, the positive evaluation subjects who did report these signs reported them consistently less often (Table 8). If it is agreed that a subject who reported.the same sign several times (e.g., one Who says over and over again that he is anxious and uncomfortable) experienced that sign more intensely than a.subject.who.reported it only once or twice, then a treatment difference in frequency of report of a particular sign would indicate that one experimental condition (or more) stimulated a more intense.form of that sign than did the others. In order to test this possibility, the positive evaluation subjects who reported a particular. sign (inability to think, verbally expressed anxiety, or self-esteem reduction) were compared to the combined no evaluation and negative evaluation subjects who reported the same sign. The two groups were 24 Now Pm mm Fm ome m—maOP mm mm m_ i we. mop cowumapm>m w>wummwz mm o_ PP Fm oo_ coFumsz>m m>wuwmoa Fm «m mm mm NmF cowuuspm>m oz cowposumc somuxm z umpxcm tepmm ucm .xumwxcm cowpozvmc . xcmzu ou .xcwgp on »u__wna=_ eaapmm-cpmm a_flwmwwnaww sanwneew to :Mwmmflwmuw mucmspamee mazes adamacwcm we accmzcmcu mucoscmc aucmzcmcm . cmwm ~mpoh d .cowuuacmc emmpmmtepmm one .zumwxzm commmaaxm appmacm> .xcwgp op zawpwnmcw com mpmpou co cowpomcpnzm exp Lance use mgoema mcmwm mo zucmscmcd .m mpnmh 25 po_ mm mm me mom mpouok Fm o_ m mp mm :owpospo>o o>wpomoz oo m 5 N_ on cowaoopo>o o>wuwmoo mo NF m_ m_ om cowpozpo>o oz cowuoooou soopmo copuooooc zuowxco xcrcp nepom oco zuowxco soopwotwpom oommocoxo op zuwpwomcw . moozutcmwm mo .xcvsa op zuwpwoocw wasps moozutcmwm wo conga: Popo» mcwucoooc muoon »__ooco> mpcospooch mcwugoooc muoon noom mo coosoz mcwucoooc mpoon amass: Pogo» loom mo coosoz room mo Loosoz .cowuosoog EooummrmFom oco .xuowxco commocoxo appooco> .xcwno op xuw—woocw cow mpopop mo compoocpoom on» copwo oco ocomoo mooxutcmwm mo amass: Fopop .m open» 26 _ ".mm “NN._ . ".mm m-._ P ".mm “No. u Nx mS\NN oo\N~ oe\_m cowoM=hmwwawflwwmowu ~.\~ N_\K o~\~_ cowoaspa>o o>aowmoa copooaooc » xcweo soopmoumpom mcwpcowwwxmm op apw__oo:w mcmugoooc mm mcwpcoomc aw o Co cowpco oco mo cowpcoooco mo cowpcoooco .cowuooooc Emopmotmpom ooo .zuomxco commocoxo appooco> .xowgu op xumpwnocw oopcoooc ogz muoowoom cowpo:_o>o o>wpomo= + cowpmopo>o o: .m> mpoonoom cowpoopo>o o>mpwmoo mo cowucoooco .n mpnoh 27 .moooo. v.m .3. No.m eo._ -.F .m .3, mm_ m.nm mmm .m mmo .m om mmz mm mm mam _o mo mpopoh mom pm mm m.omm mo mm «Nu cu Fm cowuoopo>o o>wuomoc + corpoopo>o oz no o. a m.eo __ a NNN _~ NF cowoe=_a>o a>wowmoa xocooooce :mwm zocozoocw : zooooooce capo zoooooocm c zocoooosm :mwm xocoooocm : wo mzcoa N cm_m w mo mxcoc w :mwm w mo mzcoc w omen w cowoozooe zuowxce xcwzo oo zowpcneco EoopmoteFom oommocoxo xFPooco> .mpoospoocp cowooormoo o>wpomoc + cowpoopo>o o: .m> ucospoocp cowuoopo>o o>mpwmoo ow xocoooocm cmwm so; momma a zoopwzztoooz .m opnoh 28 ranked according to the number of times they had expressed the anxiety sign in question, and sum of the two groups' ranks were compared by means of the Mann-Whitney y_test. A significant difference was not obtained for inability to think or verbally expressed anxiety, but sub- jects in the positive evaluation group expressed self-esteem reduction with significantly less frequency (p.< .00005) than subjects in the combined no evaluation and negative evaluation group (see Table 8). Thus, it appears that the effect.of the positive evaluation instructions was not to decrease the number.of subjects who experienced self-esteem reduction, but to decrease the degree of threat to those subjects who did experience it. This finding gives some support to the experimental hypotheses, but because of the lack of significance of the overall analyses, the support should be considered minimal.~ Number of anxiety sign-types Means, variances, and.Hartley's Emax statistic are shown in Table l. The variances of males and females are significantly hetero- geneous.(p.<..05), so the raw scores are unsuitable for analysis of variance.. Figure 1 indicates the source of this difference: while males tended to receive.sign-type totals in the mid-range, females showed few mid-range totals and a correspondingly greater number of totals in the high range. The minimum transformation of the scores necessary to support the assumption of homogeneity of variance was X. = /Xf:—T5 (Table 2). The square-root transformation is suggested by Kirk (1968) for situations in which the dependent variable is a frequency count of events having a 30 low probability of occurrence--a characterization which fits the data of the present.study.. The.above.modification of the simple square-root transformation (i.e.,XI =./X).is recommended by Kirk when any X < 10, as is the case in the data here. Notice, in Table 2, that the relation- ship between means and variances of the raw scores, an indication of departure from normality, has.also been decreased by the transformation. . A 3.x 2 x 4 factorial analysis of variance (Winer, 1962) was performed on the transformed scores (Table 9). There were no signifi- cant effects. Freguency of anxiety signs In contrast to the sign-types variable, in the tabulation of which an SRS category could be added to a subject's sum only once, "frequency" refers to the number of a subject's phrases containing any SRS score, regardless of how.many times the subject had already described the particular SRS category.. Table 3 contains the means, variances, and Emax of the raw scores; again the data show lack of homogeneity of variance. Female variance once more exceeds that of males (p_< .01), and there is a significant difference between the subjects of different experimenters (p_< .05). Figure 2 shows that, again, females had greater representa- tion in the high range and a lower frequency than males in the mid-range. The modified square-root transformation was employed once more, for the same reasons noted above, and Table 4 shows the transformed scores to be more appropriate to analysis of variance. Inspection of the means Table 9. Analysis of variance for total number for sign-types by 31 experimental treatment, sex of subject, and experimenter. Source §_§_ _d_i_"_ _M§_ _|: Treatments (A) .8232 2 .4116 .65 Experimenters (B) 1.6226 .5409 .86 Sex of subjects (o) .3348 1 .3348 .53 A x B 1.9816 5 .3303 .52 A x c .2781 2 .1391 .22 BAX c .5255 3 .1755 .28 A x B x c .l.5ll8 5 .2520 .40 Error 30.2273 48 .5297 Totals ' 37.305 71 32 1 04 A I \ I \ 91 I ‘ ' \ I \ l \ 8+ I ‘\ I \ ’ 1 7.. l \ I \ ’ \ at I ‘ . . ___. \ \ \ 5.. \ \ \I 4.. \ 55 \ \ \ / ” '\ / 3.. \___ \ . \ 2., \ \ 1.. \. 6 0 1 t t t : t : e z 0 1-2 3-4 5-6 7-8 9-10 11-14 15-21 22-30 Frequency of anxiety signs Figure 2. Distribution of sign frequency by sex. 33 and variances in Table 4 also shows that the transformation again accomplished a decrease in departure from normality. The 3 x 2 x 4 factorial analysis of variance performed on the transformed scores again shows no significant effects (Table 10). Number of words spoken Table 11 indicates a significant departure from homogeneity for the variances of the experimental treatments (p_< .05). Figure 3 shows that these variances differed because subjects in the no evalua- tion treatment exhibited more fluency in the mid-range (400-799 words) than subjects in the two evaluation conditions, and the evaluation subjects showed correspondingly more fluencies in the high-range (800 or more words). N0 common transformation (i.e., square-root, logarithm, reci- procal) made these data amenable to analysis of variance, and rather than distorting them by applying a more exotic transformation, it was decided to use nonparametric.meth0ds. The Mann-Whitney U_test showed that the medians 0f the sexes were.nearly identical (p_> .10; z_= -.35), and the.Kruskal-Wallis.one-way analysis of variance by ranks showed no significant effect for.experimenta1 treatment (p_> .10; x2 2 = .14, gf_= 2) 0r experimenter (p_> .10; x = 3.43, gf_= 3). Lack of significant effects obviates the need for re-analysis of the anxiety sign data while controlling for fluency. .Interactions between the factors were graphed; visual inspection revealed no need for further analysis. 34 Table 10. Analysis of variance for sign frequency by experimental treatments, experimenters, and sex of subject. Source SS d1: M_S_ f_ Treatments (A) 4.64 2 2.32 1.34 Experimenter (B) 4.93 3 1.64 .95 Sex of subject (C) .96 1 .96 .54 A x B 1.21 6 .20 .12 A x C .85 2 .43 .24 B x C .32 3 .11 .06 A x B x C 6.92 6 1.15 .66 Error 83.09 48 1.73 Totals 102.91 _ 71 35 Table 11. Raw-score means, variances, andF ax for experimental treatments, experimenters, and $39 of subject. Variable 3: Number of words spoken. Mean Variance Emax Treatments N0 evaluation 481.88 75862.64 Positive evaluation 608.75 243907.07 3.30"; __1: = 23, [<_. = Negative evaluation 590.21 250159.74 Sex of subject Female 555.97 142625.46 1.68; gf_= 35, l_<_= Male 561.81 239977.36 Experimenter JP (Male) 468.89 139798.79 BK (Male) 678.61 291311.19 2.084; gj.= 17, k_= LJ (Female) 473.33 159420.59 LW (Female) 520.28 158324.92 * p < .05 **p< .Ol 36 .mpcoanocu Popcoewcooxo mmocoo Aocozpe mo comuoowcumwo .m ocamwd :ozoom mocoz mo Loosoz ooopx mmmroow monroom mmmrooe monsoom mmpuooo \V11.1.1.111‘fl ..F / ..N \ \ 4 \ w m \ I \ :3 \ \\\\ cowuo=~o>o o>wuomoc . 1 r 1 r 1 1 5.5335 3.558 I / \ cowpoopo>o o: / \ staapqns go aaqwnN 37 Relationship of the measures The Pearson product-moment correlation coefficient for the relationship of number of sign-types and frequency of signs was .91. Lack of independence in this relationship was expected, since each time a subject reported a previously unmentioned SRS category, the total frequency was also.increased.. However, a correlation of this magnitude indicates that it was not.characteristic for subjects to report one anxiety sign over and over again; rather, the rate of increase in vari- ety of signs almost.kept.pace with increase in frequency. Figures 4 and 5 depict the relationships between the two anxiety sign variables and the number of words spoken by the subjects. The points on the horizontal axis represent intervals of 200 words, and subjects were assigned to these groups on the basis of the number of words they spoke during.the experimental session. Because of the low g_and similar scores of those falling in ZOO-word intervals above 999 words, all subjects who.spoke 1,000 words or more were collapsed into one group for purposes of analysis. Tables 12 and 13 present the means and variances of the fluency groups. Hays (1966) presents a regression method by which the §§-1inear regression and §§deviations from linear regression (the strength of curvilinear relationship) can be separately partitioned. The mean square of each sum of squares can then be calculated, and an §_is derived to represent the strength of both linear and curvilinear rela- tionships in the data. .Because it is an analysis of variance method, the usual assumptions must.be satisfied. When broken into fluency groups, both the number of sign-types (p_< .05; Emax = 7.84, gj_= 10, Mean number anxiety sign-types 1+» 38 000-199 200-399 400-599 600-799 800—999 >1000 (n=12) (11321) (1514) (E7) (IE7) (n=11) Number of words spoken Figure 4. Raw-score mean number of sign-types for subjects who spoke 000-199 words, 200-399 words, 400-599 words, 600-799 words, 800-999 words. and 1000 words or more. 18 I7 16.L I5 141_ 13. 121 11+. 10 (71 \J (D A l #01 N Mean anxiety sign frequency w i d L I D Figure 5. 39 J I 000-199 2001399 400-599 5001799 800-999 >999 (n=12) (n=21) (n=l4) (n=7) (n=7) Number of words spoken Raw-score mean sign frequency for subjects who spoke 000-199 words, 200-399 words, 400-599 words, 600-799 words, 800-999 words, and 1000 words or more. 40 Table 12. Raw-score means.and.variances of number of sign-types reported by.subjects who spoke 000-199 words, 200-399 words, 400-599 words, 600-799 words, 800-999 words, and 1,000 words OY‘ more. Fluency n Mean Variance 000-199 12 1.57 2.43 200-399 21 3.52 5.15 400—599 14 3.85 5.75 5004799 7 7.43 18.58 800-999 7 2.29 5.55 > 999 .11 2.18 ’ 2.37 Table 13. Raw-score means and variances of sign frequency for subjects who spoke 000-199 words, 200-399 words, 400-599 words, 600- 799 words, 800-999 words, and 1,000 words or more. Fluency n Mean Variance 000-199 12 2.58 9.18 200-399 21 5.43 17.64 400-599 14 7.29 42.12 600-799 7 18.00 122.10 800-999 7 3.29 24.60 > 999 II 2.72 5.29 41 _k= 6) and frequency of types.(p < .01; Em x = 23.08, _c_1_1:= 10, I_<_= 6) a exhibit significant lack of homogeneity of variance. Again the modi- fied square-root.transformation (X. = /X—:_75) was applied to the number of sign-types reported by.each subject. The resulting means and vari- ances appear in Table 14, and a graph of the transformed means is pre- sented in Figure 6. It.is clear that the transformation preserves the form of the relationship, and the assumption of homogeneity of variance of the transformed scores.cannot be rejected (p_> .05; Emax = 3.64, if = 10, k = 6). An appropriate.transformation for the frequency data was not discovered; therefore, the analysis of variance technique was applied only to the data representing number of sign-types. Table 14. Transformed-score (X = ix + .5) means and variances of num- ber of sign-types reported by subjects who spoke 000-199 words, 200-399 words, 400-599 words, 600-799 words, 800-999 words, and 1,000 words or more. Fluency fl_ Mean Variance 000-199 12 1.3799 .2863 200-399 21 1.9506 .3300 400-599 14 1.9824 . .4600 600-799 7 2.6623 .9811 800-999 7 1.5135 .5776 > 999 11 1.5611 .2694 Mean sign-type frequency 3.0. 2.8. 2.6. 2.4. 2.2 2.0. 9° 8 l.%. Figure 5, 42 l 41_ A A I ' 000-199 200-399 400-599 600-799 800-999 >999 Number of words spoken Transformed score (X, = % x+.5 ) mean number of sign-types for subjects who spOke 000-199 words, 200-399 words, 400- 599 words, 600-799 words, 800-999 words, and 1000 words OY‘ more. 43 While the data contain no linear component, the curvilinear relationship is highly significant (p_< .001; see Table 15). Thus, number of sign-types rises with the subject's fluency from 000-599 words, shows a sharp peak between 600 and 799 words, and drops to a low level at 800 words or more. Table 15. Analysis of variance of linear and nonlinear components in the relationship of fluency and number of sign-types. Source §§_ g: M_S_ 1; Between groups 9.4065 5 Linear reg. .0032 1 .0032 .0076 Dev. from lin. reg. 9.4033 4 2.3508 5.5855** Error 27.7753 66 .4208 Totals 37.1818 71 ** p_< .001. The increase in number of sign-types from zero to 799 words shows that, in this fluency range, number of sign-types is a function of number of words spoken. This linear relationship is conceptually plausible if it is assumed that the experimental treatments produced at least some anxiety in all subjects; that is, the more words spoken, 'the more anxiety would be likely to be expressed. From this point of *view, the near-equality of sign-type means in the lowest and highest fluency levels (see Table 14) suggests that high-fluency subjects were 44 effectively less anxious. This possibility was tested by dividing each subject's sign-type total by the number of words he spoke. Since fluency varied greatly, the range of the resulting quotients was wide. When the subjects were separated into three fluency groups (000-399 words, 400-799.words, and 800.w0rds or more), it Was found that no standard transformation was sufficient to make the variances homogene- ous. Therefore, Ryan's method of adjusted significance levels (Kirk, 1968), which employs the Mann-Whitney U, was used. Ryan's method allows multiple group comparisons by ranks and makes it possible to control the error rate for the entire analysis at alpha, The results of the analysis (Table 16) indicate that the median quotients of both the low-fluency and medium-fluency groups are significantly larger than the quotient of the.high-fluency group (p_< .001), while the difference between the low and medium-fluency groups does not reach the .03 level stipulated by Ryan to preserve a .05 error rate. Therefore, if it is plausible that an increase.in fluency should raise the probability of anxiety.sign reports.instead of lowering it, as was the case for the l1igh-fluency.group, then-it can be concluded that high-fluency subjects vvere significantly less anxious than low-fluency subjects. Additional analyses No single anxiety sign was reported significantly more fre- quently inone treatment or byone sex. By and large, males and females showedlow and high frequencies in the same categories. Categories whicfl1 showed the most sex difference were sleepiness, reported by 45 Table 16. Comparison of sign-type totals divided by number of words spoken for 10w (000-399 words), medium (400-799 words), and high (800 or more words) fluency*subjects using the method of adjusted significance levels. Required. ~ Steps probability .' between for ordered Obtained Comparison significance groups 0_ 5. probability High fluency vs. 10w fluency .02 3 521 4.415 < .0001 High fluency vs. medium fluency .03 2 324.5 Y 3.817 < .001 Medium fluency vs. low fluency .03 - 2 443.5 1.721 < .10 *n - "high fluency - 18 4medium fluency = 2' 33 n-low fluency g 46 fifteen females and eight males, and warm sensation, reported by seven females and one male. It was decided that any anxiety sign reported by ten or more subjects.would be analyzed in.more detail. It was thought that signs reported by.a very low number of subjects would provide too little information for.valid analysis, and since one of the purposes of the analysis performed on individual sign-types was to evaluate the possi- bility of sex differences, signs with a frequency less than ten could not satisfy the.standard-statistical requirement of five subjects (in this case,.five males and five females) per cell. Four sign-types were reported by ten or more subjects (see Table 17): sleepiness, inability to think, verbally expressed anxiety, and self-esteem reduction. In order to compensate for the lack of homogeneity of male and female variances and.to make.results comparable to the previous outcomes, the. same transformation of the scores (X. = JX7;_TS) was used in all analyses. Each analysis was conducted as follows: First, all subjects (eleven) who reported no anxiety signs were disregarded. Five of these subjects were female, six were male. Thus, a group of sixty-one remained--thirty males and.thirty-one females. These subjects were then divided into two groups: those whose transcripts contained the sign in question, and those whose transcripts did not receive that score. The total numberofsign-types reported by males and females in each of these groups was then compared. This method of analysis (allowed.the determination.of whether or not the presence of a particular 'bype of anxiety sign discriminated among those subjects who reported a high total of sign-types from those who reported a low total. The 47 E .5 pm 58 mm. wn.p mm mp mp coppooooL soopmoumpom pcoooc poz opo oz: mpoonzom om. pm.~ mm m_ up coppooooc Emopmoumpom oopcoooc oz: mpoonzom .5. km bm popoh mm. mm._ mm op mp appozco> zpopxzo mmocoxo poz opo oz: mpoonzom mm. mm.~ mm up up ap—ozgo> apopxco oommocoxo oz: mpoonzom ._..m ._..m. pm. Pope. mm. mm.p mp u pp zzpzp op zppppzozp pcoooc poz opo oz: mpoowzom mm. ~_.~ mo om mp zzpzp op zppppzozp oopcoooc oz: mpoonzom .6. pm .pm :38 RN. om.p mm mp mm mmozpooopm pcooog poz opo oz: mpoonzom No. mp.N mm mp m mmozpooopm oopcoooc oz: mpoonzam .hm Popoz mo. mo.~ pm 1 1 moposoz om. Rm._ om 1 1 mopoz oozopco> zoo: .m oposoem oposm .zmpm zpopxoo ozo pmmo_-po oopcoooz oz: mpoowzom Lop moozpucmpm mo amass: po moozopco> ozo mzoos Am. + x» n xv ocoomuomscopmzocp .mp opzoz 48 frequency data were not used in these analyses, since, for this purpose, they were contaminated: if the analysis were being performed on self-esteem reduction, for instance, and a subject had reported it ten times (which one subject did), the very process of dividing the subjects into the groups who did and did not report the sign would have unacceptably increased the probability of a Type I error. Since a report of each sign equally influenced a subject's sign-type total, this variable did not.contain the same source of bias. Table 17 con- tains the transformed.sign-type means and variances for males and females, for subjects who reported sleepiness vs. those who did not, and for subjects who did and did.not report inability to think, verbally expressed.anxiety, and self-esteem reduction. Again, the table does not include subjects wh0.reported no anxiety signs. Four two-way conventional analyses of variance for unequal cell gfs (Kirk, 1968) were performed,.one for each of the high frequency anxiety.signs. The results, presented in Tables 18-21, show two highly significant main effects for presence of the anxiety sign in question and two significant sign-type presence X sex interactions. Interpreta- tion of the main effects, though, should be cautious, since the proba- bility that a subject's.transcript contains a particular anxiety sign must increase with an increase in the number of different signs reported by that subject. For example, since a subject with a sign-type total of twelve has a higher probability of reporting self-esteem reduction than a subject with a sign-type of one, it is not surprising that the total number of sign-types reported by subjects whose transcripts con- tained self-esteem reduction was significantly higher than the total 49 Table 18. Analysis of variance of the effects of sex and the presence of sleepiness report on number of sign-types. Source §_§_ _d_f_ _M§_ 5 Sex of subject (A) .2433 l .2433 .88 Report of sleepiness (B) .799 1 .2433 2.89 A x B 4.04 1 4.04 14.51** Error 15.7577 57 .2765 Totals 20.84 60 ** .9. < .001. Table 19. Analysis of variance of the effects of sex and the presence of report of inability to think on number of sign-types. Source §_S_ _d_f_ M_S_ f_ Sex of subject (A) .2433 1 .2433 .72 Presence of inability to think (B) 1.074 1 1.074 3.18 A x B .29 1 .29 .86 Error 19.92 57 .3374 Total 20.84 60 50 Table 20. Analysis of variance of the effects of sex and the presence of verbally expressed anxiety on number of sign-types. Source §_§_ 9: L15. 5. Sex of subject (A) .2433 l .2433 .80 Presence of verbally ** expressed anxiety (B) 2.93 l 2.93 9.64 A x B4 .32 l .32 1.05 Error 17.35 57 .304 Totals 20.84 60 ** E. < .005. Table 21. Analysis of variance of the effects of sex and the presence of report of self-esteem reduction on number of sign-types. Sex of subject (A) .2433 1 .2433 .9212 Presence of self-esteem ** reduction (B) 4.324 1 4.324 16.3726 A x 8 1.2187 1 1.2187 4.5145" Error 15.054 57 .2641 Totals 20.84 60 'k p_ < .05. *9: p < .001. 95 51 reported by those whose transcripts did not contain self-esteem reduction. Nevertheless, the fact that only two of the four analyses showed significant.effects for presence of a particular sign-type indi- cates that other factors than these were operative. A significantly higher number of sign-types was present in the transcripts of subjects who reported self-esteem reduction (p_< .001; see Table 21) and expressed anxiety verbally (p_< .005; see Table 20) than was present in the transcripts of subjects who did not report these signs. In addition, there was.a significant Sex X Presence of sign- type interaction for self-esteem reduction (E.< .05; see Table 21). Thus, although both sexes showed a rise in sign-type total when self- esteem reduction was present, females.showed a significantly greater increase than males. The relationship is depicted graphically in Figure 7. The null hypothesis that inability to think was equally dis- tributed across subjects with varying sign-type totals could not be rejected (p_> .10; see Table 19). The analysis for sleepiness showed no main effect, but there was a highly significant Sex X Presence of sleepiness interaction (p_< .001; see Table 18 and Figure 8). Again, it was females who showed a sharp increase in sign-type totals when 5 leepiness was present, while males did not. In fact, males who reported sleepiness actuallyshoweda decrease in sign-type total (Figure 8).‘I‘ *The usual procedure in analyses of variance of the type pre- sented directly above, after selecting the two groups on the basis of a diclwotomous variable (e.g., report of sleepiness vs. no report of sleepi- ness;),.is to subtract from each subject's total score any effect of the selee .10 SELF-85119311 ** ‘ *** reduction .27 .47 .03 .477 < .08 * .2 < .10. ** p < .05. *** 56 relationship for females (r_= .47; p_< .01). No relationship between self-esteem reduction and total score was obtained for males (r_= .03). The difference between male and female correlations, which is simply another way to represent the Sex X Presence of self-esteem reduction interaction (Figure 7), neared accepted significance levels (p_< .08). These results are essentially in agreement with the previous analyses of variance, which showed a significant main effect for the presence of the sign_and for the interaction. The correlations for sleepiness also replicate the analysis of variance results, showing.that the opposite-signed relationships for males (r_= .55; p_< .Ol).and females (r_= .37; p_< .05) cancelled each other, producing no main effect (r_= .03), but a significant difference between the correlations (p_< .05). e The major difference between the two types of analyses is the lack of a.significant.correlation between verbally expressed anxiety and total score (r_= .20;.p_> .10). It may be inferred that the sig- nificant relationship obtained by the analysis of variance (Table 20) depended.on the inclusion of the sign in the total score. Failure to obtain the result using the more stringent correlative method requires that the finding be interpreted more tentatively than results for sleepiness and self-esteem reduction. A statistically valid statement regarding the direction of causation in these relationships of sign-type to total score cannot be made. However, the experimental treatments were devised to cause anxiety by.reducing_self-esteem--that self-esteem reduction and verbally expressed anxiety were among the three most frequently reported 57 sign-types, and that the presence of self-esteem reduction, and perhaps verbally expressed anxiety, discriminated subjects with high- sign-type totals from those with low totals indicates that the treat- ments were effective. The data contain a pattern: reports of sleepiness, self- esteem reduction, and, tentatively, verbally expressed anxiety accom- panied high total numbers of sign-types in females; the same relation- ships do not hold for males, who actually showed a significantly lower number of sign-types when sleepiness was reported. It can be concluded that, although quantity of anxiety was not different across the sexes, the quality of the experience of anxiety did show a sex difference. Debriefing_of subjects The questionnaire administered to subjects after the experi- ment (Appendix B) did not indicate that subjects responded to the evaluative content of the experimental instructions given to them in the positive and negative evaluation groups. In fact, most subjects wrote that they had felt anxious about experimenter evaluation, even those subjects in the positive evaluation group. The questionnaire also indicated that no subjects had preconceptions about the laboratory or previous experiments which had taken place there. Because of the qualitative nature of answers to the questionnaire, statistical analysis of these data was not attempted. Debriefing of experimenters A careful debriefing of experimenters, who were blind to the experimental hypotheses, established that three of the four, because 58 of the subjects' responses, had guessed that the experiment in some way concerned anxiety. Their unanimous impression was that negative evaluation was the most stressful of the three conditions for subjects (and for experimenters as well). None had any notion of how anxiety would be measured. It can be concluded that the general knowledge the experi- menters had concerning the purposes of the experiment did not influence the results, since there were.no experimenter effects. The significant departure from homogeneity of experimenter variances in the frequency data was not caused by a deviation of the one totally blind experimenter from the other three. DISCUSSION An earlier study (Stern and Reyher, 1974) asserted that treatments almost identical to those used here produced spontaneous subject reports of psychopathology. These results led to the present research, with its initial emphasis on psychopathology. However, after the present study.was already in progress, a reassessment of Stern and Reyher's results produced a second paper (Stern and Reyher, 1975; see Appendixll).in.which the reactions previously termed psycho- pathology were, because they occurred with lower frequency and more restricted range than symptoms.0bserved in other studies from the same laboratory (Sommerschield and Reyher, 1973; Burns and Reyher, in press), redefined as "signs of anxiety." The symptomatic reaction scale, of course, does.contain many symptoms which are accurately described as anxiety signs; the re-labelling was decided upon as a way of focusing attention on the data itself and away from the argument about whether or not the data truly represented psychopathology. .The position taken here is that the SRS categories described by subjects in the present study and that of Stern and Reyher are reactions to subjective distress, either conscious or unconscious, caused by either internal or external stressors. It is less important to establish whether or not these signs of anxiety are psychopathology than it is to describe them accu- rately and understand their development and relationships to one another. 59 60 .Since the Symptomatic Reaction Scale is, as its title suggests, a list of psychopathological symptoms, is it valid to term scorable reactions of relatively low intensity "anxiety signs"? As long as Reyher's (1958, 1961,.1967; Perkins and Reyher, 1971; Sommer- schield and Reyher, 1973; Burns and Reyher, in press) rationale for defining the reactions in.the SRS as symptoms is accepted, it is also conceptually accurate to label as anxiety signs less intense manifesta- tions of the same SRS.categories. Each symptom on the SRS is concep- tualized as the result of ego's.inabi1ity to defend against the anxiety engendered by drive activation. Although the procedures of the present study were not devised to stimulate drive, they were designed to increase anxiety from other sources. Because anxiety is said to be a part of the process.underlying.the production of the SRS symptoms, and because subjects in the present study did report many reactions scorable by the SRS, it can be concluded that the scorable reactions are signs of anxiety. Although there was only minimal support for the experimental hypotheses as initially formulated, the findings suggest that there is indeed a relationship between degree of threat to self-esteem and num- ber of anxiety signs reported. This relationship, however, is much more complex than the hypotheses indicate. First, although the number of subjects who reported self-esteem reduction did not differ between groups, those subjects in the positive evaluation group who did report this sign verbalized it.with significantly lower frequency than subjects in either of the other groups (see Table 11). Thus, the experimental groups did show effects consistent with the predicted ones. More 61 importantly, though, it was found that self-esteem reduction accompanied high SRS totals only for females. Males who reported self-esteem reduc- tion showed no higher totals than males who did not report this sign (Tables 21 and 22, Figure.7). .,Examination of two other variables, expressed anxiety and report of sleepiness, revealed further relationships. Again, females who expressed anxiety reported more sign-types than females who did not report the sign, but males who expressed anxiety reported no more SRS sign-types than males who did not express anxiety (Tables 20 and 22). This relationship supports, for females only, the theory which led to the original hypotheses; that is, self-esteem reduction and verbally expressed anxiety.both predict total number of sign-types. Although no direction of causation can be empirically established here, Sullivanian theory predicts that self-esteem reduction leads to the experience of anxiety. . Interestingly, analysis of sleepiness reports led to the greatest sex differences: while, once more, this sign was significantly positively related to total score for females, it was even more strongly negatively related to total.score for males. Sleepiness in males, then, seems to indicate the absence of anxiety. (See Tables 18 and 22, Figure 8.) Bringing these.results together, it can be seen that sex moderates the relationship between three anxiety signs and total SRS score. That is, if one knows the sex of the subject and whether or not the subject reported self-esteem reduction, anxiety, and sleepiness, prediction of total SRS score is greatly enhanced. 62 Female raw-score variance exceeded male variance for both anxiety sign variables (Tables 1 and 4, Figures 1 and 2). This dif- ference was caused by a greater proportion of females than males in the high range and a lower proportion of females in the mid-range. Thus, while males tended to react to experimental conditions with low or moderate anxiety, females responded with low or high anxiety. This finding indicates a partial explanatory basis for the relationships described directly.above. If,a female reported self-esteem reduction, anxiety, or sleepiness, she was likely to have scored in the high range. The total SRS scores.of males, on the other hand, because the range of their scores.was narrower, were, in general, not as well pre- dicted by responses to single items. The single strongest relationship in the data, though, was that between male reports of sleepiness and total SRS score. This result, plus the lack of any relationship between inability to think and SRS total, indicates that the relationships obtained between individual items and total SRS scores were not merely due to sex differences in SRS variance. Prior studies have established both theoretical and empirical support for sex differences on self-report anxiety measures. ‘Gray and Buffery (1971), in an attempt to integrate behavioral sex differences, neural evidence, and evolutionary theory, predicted that females would 'show more verbally expressed.fear than males. Neufeld and Johnson (1974) found that females reported significantly more subjective anxiety (a checklist method) than males when subjects' responses to slides of mutilated bodies were compared to baseline responses to non-stressor slides. Related studies show the same direction of difference. Women 63 report more fears than men, and girls more than boys (Marks, 1969). The preponderance of people suffering from phobias are women (Marks, 1969), and two major American surveys (Srole, Langer, Michael, 0pler, and Rennie, 1962; Leighton, Hurding, Macklin, Macmillan, and Leighton, 1963) conclude that both "psychological" and psychosomatic symptoms are more common among women than men. The subjects of the four experimenters showed significantly heterogeneous.sign frequency variances (p.< .05; see Table 4). The relevant personality characteristics of the experimenters, of course, are not known, but in view.of.the fact that self-esteem reduction was for women significantly related-to number of sign-types, it seems likely that one experimenter was perceived by some subjects as particularly important or attractive,a person in front of whom the subject wanted to appear adequate.. Another possibility, that the high-variance experi- menter conveyed a more punitive attitude toward subjects, was suggested by the experimenters themselves who, in.an informal discussion of the data collection, each described and demonstrated the tone of voice, postures and gestures they assumed as experimenters and unanimously agreed.that the most punitive of the group was the experimenter later found to be associated.with.the.largest.variance. Sex of experimenter was probably not.a.causal factor, because one of the two highest vari- ances belonged to a male, one to a female. .. The significant heterogeneity of the variances representing the number of words spoken in the.threetreatments (p_< .05; see Table 7) again suggests that the treatments did have a differential effect. Notice in Table 7 and Figure 3 that the variances of both the positive 54 and negative evaluation groups are triple the size of the no evaluation variance.. The effect of potential evaluative feedback from the experi- menter seems.to.have caused high fluency in some subjects. On the other hand, when there was no method of.evaluation available (no evalu- ation.or.uncertainty.gr0up),.fluency was universally low. These find- ings, while they are not direct-support for the theory, accord.with Sullivan's (1953) description of the process of security operations. Because the concept.of self-esteem is defined as an observed reflection of the.other's evaluation of oneself, one must have access to the facial, postural, and verbal.reactions.of the significant other in order to judge whether one's.security operations are having the desired effect of self-esteem maintenance. Facial and postural cues, of course, were unavailable to subjects.in all treatments, since.the experimenter was positioned behind the.subject, outside his field of vision.1 . However, subjects were led to believe that verbal evaluative reactions were available from-the experimenter in the positive and negative evaluation groups. The larger variances in these groups may be interpreted.to mean.that.when clear indications of evaluation were 1It is interesting.in this regard to note how often in the transcripts of the present study and the earlier one by Stern and Reyher (1975), which also placed the experimenter behind the subject, subjects described sounds which they thought were movements of the experimenter in his chair or the scratchings of the experimenter's note-taking. (The experimenter did not take notes.) These movements and scratchings were often then interpreted as indications of experi- menter evaluation. Thus, in the absence of direct evidence, subjects began to pay attention to small cues from the experimenter which were then projectively endowed with evaluative qualities. 65 available, some subjects,.using.the single medium of communication by which they were linked to the experimenter, speech, actively strived for experimenter approval (i.e., to maintain self-esteem). In the no evaluation group, uncertainty-about methods of judging the experimenter's reaction.made such striving useless,.and this group's variance was con- sequently smaller.. According to this viewpoint, then, the type of security operations used to maintain self-esteem differed across treat- ments. More will be said.below about the high-fluency subjects who enlarged the variances.of the two evaluation groups. . Particularly interesting results are effects which were obtained by viewing.the subjects.as one large group. First of all, there was a significantly curvilinear relationship between the total number of sign-types reported by a subject and the number of words he spoke (p_< .001; see Tables 14 and 15, Figure 6). It is not surprising that.sign-type.totals increase from zero to 799 words: it is to be expected that an increase in fluency also increases the probability of anxiety.sign reports.’ The striking finding is the drop in sign-type total at 800 words. .While Table 14 shows that the mean sign-type totals of high-fluency (800 words or more) and low-fluency (000-399 words) subjects.are nearly the same, when each subject's total number of sign- types.was divided by the number of words he spoke (thus partitioning out the.increase in probability of sign reports associated with increases in fluency), the quotients of the high-fluency group are significantly smaller than those of the low-fluency subjects (p_< .0001; see Table 16). .Effectively, then, high-fluency subjects were less anxious than low-fluency subjects. Anecdotal support of this conclusion was provided 66 by two experimenters who spontaneously and individually reported to the writer during the course of the experiment that the subjects who spoke the least seemed the most.frightened. There is no direct evidence which can be used to differentiate the personality characteristics of the high-fluency subjects from those of the other participants, but two speculative hypotheses were generated to account for the difference. The most obvious interpretation, on first viewing the data, is that high-fluency subjects are intellectual- izers, people who use words in an.obsessive-compulsive way to defend against anxiety. This explanation is inadequate, though, because high- fluency subjects often.talked.about quite personal matters (see Appendix E). Their remarks did.aot.show the isolation of affect characteristic of intellectualization. The only other interpretation which fits this relationship rests on the fact that high-fluency subjects were the participants best able to comply with the instructions to free associate. Their trans- cripts did not contain the.long silences which frequently interrupted the verbal flow of most other subjects. While it is possible to inter- pret these.silences from an.intrapsychic viewpoint (i.e., the silences occurred when to speak would.have released anxiety-provoking impulses to the partial motility of speech), it is more likely, given that the subjects were completely unacquainted with the experimenter, that silences were caused by the.subject's apprehension concerning the experimenter's reaction to a particular train of thoughts or feelings. The thoughts or feelings were therefore not verbalized, a silence ensued, and fluency was consequently lowered. If this interpretation 67 is correct,.then high-fluency subjects were evidently either relatively unconcerned with the experimenter's reaction or confident that it would be positive.. The most concise way to summarize these characteristics is, again, under the rubric of.self-esteem: people who maintain a chronically high level of.self-esteem.are able to disclose their thoughts and feelings freely because they feel relatively little anxiety about the significant other's.reaction. This is an important point of contact between Sullivanian and Freudian theory: in intrapsychic terms, the high-fluency subjects can be characterized as people who are not in desperate need of narcissistic supplies, i.e., approval. The exis- tence of a great disparity.between need for such supplies and their perceived availability might constitute a viable operational definition of low self-esteem (Fenichel, 1949). It has been proposed that the increase in number of sign- types from.zero.to 799 words.is.due.to.the increase in fluency. This explanation is unsatisfying,.though, unless there is also an explanation for fluency differences.in this range. One explanation is the follow- ing: Subjects with the lowest fluencies were most threatened by self- disclosure, since their chronically low self-esteems led them to believe that their remarks would be evaluated negatively by the experimenter. Subjects with moderately stable self-esteems were able to speak more, thus describing more of the anxiety they felt and deriving higher sign- type totals, but these subjects' fear of negative reaction still pre- vented them from freely disclosing their thoughts and feelings. Subjects with high self-esteem, discussed above, were able to speak freely and with little anxiety. In essence, then, it is proposed that fluency was 68 a function of.the subjects' chronic self-esteem level, and that the curvilinear.relationship.depicted by Figure 6 represents the statistical dependence of intensity of anxiety and degree of self-esteem in a free association situation...However, data bearing directly on this inter- pretation are not available, and the interpretation must therefore be viewed as an hypothesis suggested by the data rather than a conclusion. More information is needed:. for instance, what is the distribution of verbal ability across fluency.leve1s? A final result is.the high correlation (.91) between number of sign-types and sign frequency. This finding means that subjects' sign frequencies were almost directly proportional to the number of sign-types they reported, and that the two variables were therefore very similar measures. It can be concluded that high anxiety during the verbal association procedures was represented by reports of a wide variety.of.SRS.categories agg,a.relatively high proportion of phrases containing an SRS.category._ A replication of this study with a larger N_would be profit- able. The larger the sample, the more signs would be reported frequently enough for detailed statistical test. Since sex differences in pattern of anxiety manifestations were found in the present study, it is likely that a larger one could show much more intricate and complete relation- ships among sign-types and sign-type totals. SUMMARY The results provided little support for the hypotheses as initially formulated. There.were no differences in number of anxiety signs reported by subjects in the three treatment groups. However, consistent.with the.hypotheses.was the finding that subjects who reported self-esteem.reduction under positive evaluation conditions repeated the.Sign with.significantly lower frequency than did subjects in the no evaluation and.negative evaluation groups. ...,Four signs of anxiety.were reported with high enough frequency for further statistical analysis. For females, verbally expressed anxiety, sleepiness, and self-esteem reduction were significantly positively related to total SRS score, while inability to think was unrelated. .For males,.sleepiness.was significantly negatively related to tota1.SRS.score, while the other three signs were uncorrelated with SRS total. These analyses were carried out separately with analysis of variance and point-biserial correlation. The difference between the male and female correlations was significant for sleepiness in both analyses-and for self-esteem reduction in the analysis of variance. It was concluded that sex was a moderating variable in the relationship of self-esteem reduction, verbally expressed anxiety, and sleepiness to tota1.SRS.score. For women.only, these findings indicate that, con- sistent with Sullivanian theory, loss of self-esteem and expressed anxiety bear a vital relationship to total anxiety. It remains to be 69 70 determined whether, as Sullivan predicted, this relationship is a causal one. For both frequency of report of anxiety signs (regardless of the number of times an SRS category was repeated) and number of SRS categories reported, males and females showed heterogeneous variances. While males tended to react to experimental conditions with low or moderate anxiety, females reacted with low or high anxiety. The greater number of positive part-whole correlations for females than males may have been due to this distribution, since the presence of any SRS cate- gory in.a female's protocol, because of greater female variance, may have led to more predictive power. Additional analyses revealed a significant curvilinear rela- tionship between number of words spoken by subjects and number of anxiety sign-types (i.e., the number of different SRS categories) they reported. Subjects with low and high fluency reported few anxiety signs, while subjects with mid-range fluency reported large numbers of signs. It.was shown that even though their mean anxiety scores were nearly equivalent, high-fluency subjects were effectively less anxious than low-fluency subjects. -The-curvilinear relationship was specula- tively interpreted as a representation of the dependency of chronic self-esteem level and anxiety during verbal association. The variances representing number of words spoken were heter- ogeneous.across the three treatments (p;< .05). The two equally large variances.were obtained in the positive and negative evaluation groups; thus, when evaluative feedback-was available, some subjects' fluencies rose. Speculatively, security operations change when such feedback 71 becomes unavailable, because self-esteem maintenance requires that one be able to judge the effect of one's behavior on others. REFERENCES REFERENCES Alexander, F. An experimental approach to the study of physiological and psychological effects of emotional stress situations. In T. Tourlentes, S. Pollack, and Hi Himwich (Eds.) Research approaches to psychiatric problems. New York: Grume an Stratton, 1962. Aserinsky, E. and Kleitman, N.' Two types of ocular motility occurring in sleep. Journal of Applied Physiology, 1955, 8, 1-10. Barber, T. X. and Calverley, O. S. Multidimensional analysis of "hyp- notic" behavior. Journal of Abnormal Psychology, 1969, 14, 209-220. - Berlyne, D. E. Uncertainty and conflict: A point of contact between information theory and behavior theory. Psychological Review, 1957, 64, 329-339. , Berlyne, D. E. The motivational significance of collative variables. and conflict. .In Abelson, Aronson, McGuire, Newcomb, Rosen- burg, and Tannenbaum (Eds.), Theories of cognitive consistency: A sourcebook. Chicago: Rand c a y an 0., 1968. Bliss, E. L., Clark, L. 0.,.and West, C. 0. Studies of sleep deprivation- relationship to schizophrenia. Archives of Neurology and Psychiatry, 1959,.81, 384-389. Bobbit, R. A. The repression hypothesis studied in a situation of hypnotically induced conflict. Journal of Abnormal and Social Psychology, 1958, 56, 204-212. Breland, K. and Breland, M. The misbehavior of organisms. American Psychologist, 1961, 15, 681-684. Bruner, J. The cognitive consequences.of.early sensory deprivation. In Solomon, Kubzansky, Leiderman, Mendelson, Trumbull, and Wexler (Eds.), Sensory deprivation. Cambridge, Mass.: Har- vard University Press, 1961. Burns, 8. and Reyher,.J.. The activation of posthypnotic conflict via free imagery: A study of repression and psychopathology. Journal of Abnormal Psychology, in press. 72 73 Clemes, S. R. and Dement, W. C. Effects of REM sleep deprivation on psychological functioning. Journal of Nervous and Mental Disease, 1957:.1152 485-491. Cohen, 0. B. On the etiology of neurosis. Journal of Abnormal Psy- chology, 1974, 84,.473-479. Cohen, S. I., Silverman, A. J., Bressler, B., and Schmamovian, B. Problems in isolation studies. In.Solomon, Kubzansky, Leider- man, Mendelson, Trumbull, and Wexler (Eds.), Sensor de riva- tion. Cambridge, Mass.: Harvard University Press, I961. Cohen, S. I., Silverman, A. J.,.and Schmamovian, B. M. Psychophysio- logical studies in altered sensory environments. Journal of Psychosomatic Research, 1962, 6, 259-281. Cook, S. A survey of the methods used to produce experimental neurosis. Journal of Psychiatry, 1939, 95, 1259- 1276. Davis, D. R. An introduction to psychopatholggy_. London: Oxford Univers1ty Press, 1972. Deckert, G. H. and West, L. J. Hypnosis and experimental psychopathol- ogy. American Journal of Clinical Hypnosis, 1963, 5, 256-276. Dement, W. C. The effect of dream deprivation. Science, 1960, 144, 1705-1707. Dement, W. C. Psychophysiology of sleep and dreams. In S. Arieti (Ed.), American handbook of psychiatry, Volume III. New York: Basic BoOks,71966. Dement, W. C., Cohen, J. F. , and Zarcone, V. A sleep researcher' s odyssey: The function and clinical significance of REM sleep. In Madow and Snow (Eds. ), The psychodynamic s1 nificance of the_physiological studies on dreams. Springfie d, Ill. CharWes C} Thomas,71970. Dement, W. C. and Fisher, C. Experimental interference with the sleep cycle. Canadian Psychiatric Association Journal, 1963, 8, 400-405. Dement, W. C., Greenberg, S., and Klein R. The effect of partial REM sleep deprivation and delayed recovery. Journal of Psychiatric Research, 1966, 4, 141-152. Dement, W. C. Henry, P., Cohen, H. ., and Ferguson, J. Studies on the effect of REM deprivation in humans and animals. In S. Kety, E. Evarts, and H. Williams (Eds. ), Sleep_and altered states of consciousness. New York: Grune and Stratton, 1968. 74 Dmitruk, V. M. Conflict and the "experimental neuroses" of cats. Unpublished doctoral dissertation, Michigan State University, 1972. . Dmitruk, V. M. Experimental neurosis in cats: Fact or artifact? Journal of Abnormal Psychology, 1974, 8;, 97-105. Dudley, 0., Holmes, T., Martin, C., and Ripley, H. Changes in respira- tion associated with hypnotically induced emotion, pain, and exercise. Psychosomatic Medicine, 1964, 24, 46. Dworkin, S. Conditioning neurosis in cat and dog. Psychosomatic Medi- cine, 1939, l, 388-396.~ Finger, F. Abnormal animal behavior and conflict. Psychological Review, 1945, pp, 230-233. Fisher, C. Some psychoanalytic implications of the recent research on sleeping and dreaming. .In L. Madow and L. H. Snow (Eds.) The psychodynamic implications of the physiological studies 0 reams. Spring 1e , .: ar es . omas, 1970. Fisher, C. and Dement, W. C. Studies on the psychopathology of sleep and dreams. The American Journal of Psychiatry, 1963, 112, 1160-1168. » Fisher, C., Gross, J., and Zuch, J. Cycle of penile erection synchron- ous with dreaming (REM) sleep. Archives of General Psychiatry, 1965, 12, 29-45. Freedman, S., Grunebaum, H., and Greenblatt, M. Perceptual and cogni- tive changes in.sensory deprivation. In Solomon, Kubzansky, Leiderman, Mendelson, Trumbull, and Wexler (Eds.) Sensor deprivation. Cambridge, Mass.: Harvard University ress, 961. Freud, S. The interpretation of dreams. New York: Avon, 1968. ' Gantt, W. H. Experimental basis for neurotic behavior. New York: Harper, 1944. Gantt, W. H. Experimental basis for neurotic behavior. In H. Kimmel (Ed.), Experimental psychopathology. New York: Academic Press, 1961. Golberger, L. and Holt, R. R. Experimental interference with reality contact. In Solomon, Kubzansky, Leiderman, Mendelson, Trum- bull, and Wexler (Eds.), Sensory deprivation. Combridge, Mass.: Harvard University Press, 1961. 75 Gray, J. A. and Buffery, A. W. H. Sex differences in emotional behavior in mammals inc1uding man: Adaptive and neural bases. Acta Psychologica, 1971, 25, 89-111. Harlow, H. F. and Harlow, M. K. Psychopathology in monkeys. In H. D. Kimmel (Ed.), Experimenta1_psychopathology. New York: Aca- demic Press, 1971. Hays, W. H. Statistics for psychologists. New York: Holt, Rinehart, and Winston, 1966. Hinde, R. A. The relevance of animal studies to human neurotic dis- orders. In J. M. Tanner, Lord Taylor, and 0. L. Zangwill (Eds. ), A_pects ofypsychiatric research. London: Oxford University Press, 1962. Hinde, R. A. and Stevenson- Hinde, J. (Eds. ), Constraints on Learning, London: Academic Press, 1973. Hollister, L. E. Chemical psychoses. Annual Review of Medicine, 1964, .15, 203-214. Hollister, L. E. Chemical 5 choses: LSD and related drugs. Spring- field, III.: Charles C. Thomas, 1968. Holt, R. R. Manual for the scoring of primary process manifestations in Rorschach protocols: Tenth draft of an experimental proce- dure. New York: Research Center for Mental Health, 1970. Huston, P. E., Shakow, 0., and Erickson, M. H. A study of hypnotically induced complexes.by means of the Luria technique. Journal of General Psychology, 1934, 2, 65-97. Jackson, C. W. and Pollard, J. C. Sensory deprivation and suggestion: A theoretical approach. Behavioral Science, 1962, 2, 332-342. Jouvet, M. Neurophysiology of the states of sleep. Physiological Review, 1967, 42, 117-177. Kales, A., Hoedemaker, F., Jacobson, A., and Lichtenstein, E. Dream g§§;i¥ation: An experimental reappraisal. Nature, 1964, 294, - 338. Karn, H. A case of experimentally-induced animal neurosis in the cat. Journal of Comparative Psycholggy, 1943, 22, 589-593. Kehoe, M. and Ironside, W. Studies on the experimental evocation of depressive responses using hypnosis. II. The influence of depressive responses upon the secretion of gastric acid. Psychosomatic Medicine, 1963, 25, 403. 76 Kelly, G. A. The s cholggy ofppersonal constructs. New York: Norton, 1955. 1*— Kimmel, H. 0. Introduction. In H. D. Kimmel (Ed.), Experimental psychopathology. New York: Academic Press, 1971. Kirk, R. Experimental design: Procedupgs for the behavioral sciences. Belmont, Ca.: Brooks/Cole, 1972. Kubzansky, P. and Leiderman, P. .Sensory deprivation: An overview. .In Solomon, Kubzansky,.Leiderman, Mendelson, Trumbull, and Wexler (Eds.), Sensory deprivation. Cambridge, Mass.: Harvard University Press, 1961. Leighton, o. C., Hurding,.J..S.,.Mack1in, o. B., Macmillan, A. M., and Leighton, A. H. The character of danger: psychiatricgypp: toms in selected communities. New York: Basic BoOks, 3 Levitt, E. E. A comment on.tA paradigm for determining the clinical relevance of hypnotically induced psychopathology." Psycho- logical Bulletin,.1963,.§g, 326-329. Levitt, E. E., den Breeijen,.A., and.Rersky, N. The induction of clinical anxiety by means of a.standardized technique. American Journal of Clinical Hypnosis, 1960, 2, 206-214. Liberson, W. T. Withdrawa1.and.fixation reactions in the rat. In J. Zubin and H. F. Hunt (Eds.), Comparative psycho atholo : Animal and human. .New York: Grune and"Stratton, 1967. Liddell, H. S. Experimenta1.neurosis.in animals. In J. M. Tanner (Ed.), Stress and psychiatric disorder. London: Oxford University'Press,TT960. Lilly, J. Mental effects of reduction of ordinary levels of physical stimuli on intact healthy persons. Psychiatric Research Reports, 1956, p, 1-28. Lindsley, D. B. Common factors in sensory deprivation, sensory dis- tortion, and sensory overload. In Solomon, Kubzansky, Leiderman, Mendelson, Trumbull, and Wexler (Eds.), Sensory deprivation. Cambridge, Mass.: Harvard University Press, Luby, E. D. and Gottlieb, J. S. Sleep deprivation. In S. Arieti (Ed.), American handbook of psychiatry, Volume III. New York: . Basic Books,1966. Luria, A. R. The nature of human conflicts. New York: Livewright, 1932. 77 Maier, N. and Feldman, R. Studies of abnormal behavior in the rat. Journal of Comparative Psychology, 1944, 51, 151-158. Malmo, R. B. and Shagass, C. Studies of blood pressure in psychiatric patients under stress. Psychosomatic Medicine, 1952, 11, 82- 93. Marks, I. M. Fears and phobias.. London: Heinemann, 1969. Masserman, J. Behavior and neurosis. Chicago: University of Chicago Press, 1943. Masserman, J. PrincipJes of dynamic psychiatry. Philadelphia: Saunders,’1946. Masserman, J. H. The principle of uncertainty in neurotogenesis. In H. D. Kimmel (Ed.), Experimental psychopathology. New York: Academic Press, 1971. Moore, W. F. The effects of posthypnotic stimulation of hostility upon motivation. American Journal of Clinical Hypnosis, 1964, 1, 130-135. Morris, G. 0. and Singer, M. T. Sleep deprivation: Transactional and subjective observations. Archives of General Psychiatry, 1961, 5, No. 5. ' Moses, I. The relationship.between symptomatology and the degree of repression of sexual conflict. Unpublished masters thesis, Michigan State University, 1974. Moses, I. Personal communication, 1975. Mowrer, 0. H. Learning theory and personaliiy dynamics. New York: Ronald—Press, 1950. Murray, E. J. Slee dreams and arousal. New York: Appleton- Century-Crofts, i965. Neufeld, R. W. J. and Davidson, P. 0.. Sex differences in stress res- ponses: A multivariate analysis. Journal of Abnormal Psychology, 1974, 55, 178-185. Newton, J. E. 0. and Gantt, W. H. History of a catatonic dog. Condi- tional Reflex, 1968,.5, 45-61. 0rne, M. T. The nature of hypnosis: Artifact and essence. Journal of Abnormal and Social Psychology, 1959, 55, 277-299. 78 0rne, M. T. and Scheibe, K. E. .The contribution of nondeprivation factors in the production of sensory deprivation effects: The psychology.of.the "panic button." Journal of Abnormal and Social Psychology,.l964, 55, 3-12. Paulov, I. P. Lectures on conditioned reflexes, Vol. 2. New York: _Internationa1Publishers,1941. Perkins, K. A. and Reyher, J. Repression, psychopathology. and drive representation:i An experimental hypnotic investigation of impulse inhibition. American Journal of Clinical Hypnosis, -1971, 15, 249-258. Persky, H. Adrenocortical function during anxiety. In R. Roessler and N. S. Greenfield (Eds.), Physiological correlates of psycho- lo ical disorder. Madison: Univers1ty oleisconsin Press, Pivik, T. and Foulkes, 0. "Dream deprivation": Effects on dream con- tent. Science,-1966, 155, 1282-1284. Plutchik, R. Foundations of experimental research. New York: Harper and Row, 1968. Pollard, J. C., Uhr, L., and Jackson, C. W. Studies on sensory depriva- tion. Archives of General Psychiatry, 1963, 5, 435-454. Rapaport, D. The theory of ego autonomy: A generalization. Bulletin of the Menninger Clinic, 1958, 22, 13-35. Reyher, J. Hypnotically induced conflict in relation to subception, repression, antisocial behavior, and psychosomatic reactions. Unpublished doctoral dissertation, University of Illinois, 1958.* Reyher, J. Posthypnotic stimulation of hypnotically induced conflict in relation to psychosomatic reactions and psychopathology. Psychosomatic Medicine, 1961, 25, 384-391. Reyher, J. A paradigm for determining the clinical relevance of hypnotically induced psychopathology. Psychological Bulletin, 1962, 52, 344-352. Reyher, J. A reply to Levitt's comments. Psychological Bulletin, 1963, 60, 330-332. Reyher, J. Hy nosis in research on psychopathology. In J. E. Gordon (Ed.§, Handbook of clinical and experimental hypnosis. New York: Macmillan, 1967. 79 Reyher, J. The real-simulator design and laboratory hypnosis: Two methodological culs-de-sac. Paper presented at the meeting of the American Psychological Association, Washington, D.C., . September, 1971. Reyher, J. and Basch, J. A. Degree of repression and frequency of psychosomatic symptoms. Perceptual and Motor Skills, 1970, 30, 559-562. Rokeach, M., Reyher, J., and-Wiseman, R. An experimental analysis of the organization of belief systems. In M. Rokeach (Ed.), Beliefs, Attitudes,_and Values. San Francisco: Jossey- Bass, 1968. Ruff, G. E. Isolation and Sensory deprivation. In S. Arieti (Ed.), American Handbook of Psychiatry, Volume III. New York: Basic Books, 1966. Ruff, G. E. and Levy, E. 2. Psychiatric research in space medicine. The American Journal of Psychiatry, 1959, 115, 793-797. Ruff, G. E., Levy, E. Z., and Thaler, V. H. Factors influencing the reaction to reduced sensory input. In Solomon, Kubzansky, Leiderman, Mendelson, Trumbull, and Wexler (Eds.), Sensor deprivation. Cambridge, Mass.: Harvard University Press, Rychlak, J. F. A Philoso h of science for personality theory, Boston: Houghton-Mifflin, I968. Sachs, E. Dissociation of learning in rats and its similarities to dissociative states.in man. In J. Zubin and H. F. Hunt (Eds.), Comparative_psychopatholo : Animal and human. New York: Grune and Stratton, 19 . Schultz, D. P. Senspry restriction: Effects on behavior. New York: Academic Press, 1965. Shagass, C. Ex erimental neurosis. In A. M. Freeman and H. I. Kaplan (Eds. , Human behavior: Biological,p s cholo ical and sociological. New York: Atheneum, I975. Shettleworth, S. J. Constraints on learning. In D. S. Lehrman, R. A. Hinde, and E. Shaw (Eds.), Advances in the study of behavior, Volume 4. New York: Academic Press, I972. Smart, R. Conflict and conditioned aversive stimuli in the development of experimental neurosis. Canadian Journal of Psychology, 1965, 15, 208-223. 80 Smith, S. and Lewty, W. Perceptual isolation using a silent room. Lancet, l959,2, 342-345. Smythies, J. R. Biochemical factors in schizophrenia. In J. M. Tanner, Lord Taylor, and 0. L. Zangwill (Eds.), Aspects of psychiatric research. London:. Oxford University Press, 1962. Snyder, F. The new biology of dreaming. Archives of General Psychiatry, 1963, 5, 381-391. Sommerschield, H. and Reyher, J. Posthypnotic conflict, repression, and psychopathology. Journal of Abnormal Psycholpgy, 1973, 52, 278-290. Spence, K. W. Behavior theor and conditionipg, New Haven: Yale University Press, 1956. Srole, L., Langher, T. S., Michael, S. T., 0pler, M. K., and Rennie,~ T. A. C. Mental health in the metropolis: The midtown Man- hattan study. New York: McGrawiHill, 1962. Stern, D. and Reyher, J. Psychopathological symptoms during verbal association and visual imagery. Unpublished manuscript, Michigan State University, 1974. Stern, D. and Reyher, J. Anxiety manifestations during verbal associa- tion and visual imagery. Unpublished manuscript, Michigan State University, 1975. ' Sullivan, H. S. The inter ersonal theory of psychiatry. New York: John Wiley and Sons, 1953. Tranel, N. Effects of perceptual isolation on introverts and extro- verts. Psychiatric Research, 1962, 1, 185-192. Tyler, 0. B. Psychological change during experimental sleep depriva- tion. Diseases of the Nervous System, 1955, 15, 293. Weitzenhoffer, A. M. Hypnotism: An objective study in suggestibility. New York: Wiley, 1953. Wittkower, E. Studies of the influence of the emotions on the functions of the organs (including observations in normals and neurotics. British Journal of Psychiatry, 1935, 81, 533- Wolberg, L. R. Hypnotic experiments in psychosomatic medicine. Psycho- somatic Medicine, 1947, 5, 337-342. Wolpe, J. Experimental neurosis as learned behavior. British Journal of Psychology, 1952, 15, 243-268. 81 Woodworth, R. S. Experimentalrpsychology. New York: Holt, Rinehart and Winston, 1938. Ziskind, E. and Augsburg, T. Hallucinations in sensory deprivation: Zubek, J. Zubek, J. Method or madness? ,Science, 1962, 151, 992. P. (Ed.), Sensory deprivation: Fifteen ears of research. New York: lAppleton-Century-Crofts, I969. P., Aftanas, M., Hasek, J., Sansom, W., Schluderman, E., Wilgosh, L., and Winocur, G. Intellectual and perceptual changes during prolonged perceptual isolation: Low illumina- t1pnlggd noise level. Perceptual and Motor Skills, l962,_y5, APPENDICES APPENDIX A ANXIETY MANIFESTATIONS AND PRIMARY PROCESS DURING VERBAL ASSOCIATION AND VISUAL IMAGERY DONNEL STERN AND JOSEPH REYER MICHIGAN STATE UNIVERSITY APPENDIX A ANXIETY MANIFESTATIONS AND PRIMARY PROCESS DURING VISUAL IMAGERY - AND VERBAL ASSOCIATION In response to provocative verbal stimuli, Reyher and Smeltzer (1968) reported that visual imagery (with eyes closed) was accompanied by more anxiety than verbal association (with eyes open), more signs of primary process regulation. more direct expression of drives, and less effective defense. However, the effect of eye closure per se was not assessed. If eye closure were found to be a confounding variable, then visual imagery could not be regarded as having superior uncovering characteristics. In view of the increasing use of imagery in psycho- therapy (Singer, 1974), this possibility has important implications, particularly for psychotherapy in which the uncovering of repressed material is a major objective. Morishige and Reyher (in press) allege that spontaneous visual imagery allows more direct expression of re- pressed material than do other methods, producing anxiety and psycho- pathology in the process. The purpose of the present investigation was to compare the uncovering properties of visual imagery and verbal association while controlling for the possible confounding effect of eye closure. 82 83 Method Subjects Subjects were randomly selected from a large group of under- graduate volunteers who had signed up to participate in an experiment titled "Unconscious Processes." 0f the 36 subjects, 18 were male and 18 were female. All were enrolled in psychology courses, and most were freshmen. Materials and experimental setting The experimental room was small, soundproofed, and windowless. A cot and a six channel Grass #5 polygraph were in storage in the room. The room also contained a large black reclining lounge chair and the experimenter's straightbacked chair. Procedure The experimenter's pre-experimental contact with subjects was minimal. He first contacted the subjects by telephone and made an appointment. At the appointed time, the experimenter met the subject in a waiting area and conducted him into the laboratory, where the experimenter suggested that the subject put his books and/or coat "on the bed" and "sit down in the big black chair." After requesting permission to tape record, the experimenter gave the subject the first instructions. For the duration of the session, the subject reclined in the lounge chair. The experimenter's chair was located behind the subject, out of the subject's field of vision. Once the experiment was 84 under way, the experimenter spoke only to give instructions to the sub- ject. If the subject asked a question about the instructions after the experimenter had given them, the experimenter repeated the instruc- tions once. If the subject asked another question, the experimenter said, "I can't answer your questions." If the subject asked a further question, the experimenter did not respond. Each subject attended a half hour experimental session in which he participated in three conditions: (a) verbal association, eyes open; (b) verbal association, eyes closed; (c) visual imagery, eyes closed. Each condition lasted ten minutes. The order of presen- tation was If a sentence contained primary process content, thein- formation was recorded. The number of primary process content scores in each condition was summed for each subject. Each subject's data was therefore consolidated to three totals. Results Reliability For the scoring of anxiety signs, a disagreement was tallied when one rater scored a sentence and the other did not, and when both raters did not agree on the type of anxiety sign. Using these criteria, the percentage of interjudgment agreement was too low (.60) to indi- cate adequate reliability. In order to compensate for this difficulty, only the data scored the same by both writers were used. The reliability of the primary process content scores was calculated differently. One condition was chosen from the transcript 85 of each subject's verbal responses. The 36 ten-minute segments of data were then ranked according to their total scores, and the Spearman rank-order correlation of the two scorers' rankings was calculated. The degree of interjudgment agreement was .87, which indicates adequate reliability. Primary_process content The number of primary process content scores given to the verbalizations of a subject during each experimental treatment corre- lated number of sentences spoken by the subject in that treatment (r = .51, pg .001). Therefore, the data used in analysis were the ratios of the total number of primary process content scores to number of sentences spoken. There were no significant differences between the three experimental treatments (f_= .64, gr_= 2/48, p? .10). Anxiety_manifestations The data for the analysis presented in Table l were the total number of types of anxiety manifestations reported within each level of each variable. A four way analysis of variance with repeated measures and equal group sizes (Winer, 1962) was used; the single sig- nificant factor was experimental treatments (p< .0005; see Table l). The Newman-Keuls method of post hoc pairwise comparisons (Kirk, 1972) showed that during both verbal association conditions subjects reported significantly (p< .01) more anxiety sign types than they did during the visual imagery condition. A significant difference between the two verbal association conditions was not obtained. 86 Table 1. Analysis of variance of number of anxiety manifestation types during experimental treatments, during different orders of presentation of experimental treatments, and by sex. Source df M5 .5 Order of presentation of experimental treatment (A) 5 6.92 1.41 Sex (B) l 15.56 3.17 A X B 5 5.23 1.07 Error between 24 4.91 Experimental treatments (C) 2 29.06 19.68* A X C 10 2.02 1.37 B X C 2 1.29 .87 A X B X C 10 .55 .37 Error within _fl5_ 1.48 Total 107 * p_<.0005 87 Table 2 shows that almost all categories of anxiety manifes- tation were described more frequently during one or both verbal associ- ation conditions than during visual imagery. Although the frequencies of individual categories were too low for valid statistical comparison, several anxiety reaction types occurred more frequently during eyes- closed verbal association than during eyes-open verbal association: sleepiness, strong urges not carried out in behavior; sensation of floating or rocking; dizziness. Compulsive thoughts and behavior occurred more often in the eyes-open condition. In general, males and females tended to report the same types of anxiety reactions and tended to show low and high frequencies in the same categories. In all, 31 of 36 subjects reported some type of anxiety Sign. Figure 1 shows the frequency of anxiety signs reported by subjects during each experimental treatment. Note that most subjects (24) re- ported no signs during visual imagery, whereas during the verbal asso- ciation conditions, most subjects (23 during the eyes-open condition and 28 during the eyes-closed condition) reported at least one. It can also be seen that high numbers of anxiety sign types (3-8) were reported almost exclusively during the two verbal association condi- tions. In fact, while only six subjects described more than one type of anxiety sign during visual imagery, two or more signs were reported by 19 subjects during eyes open verbal association and by 25 subjects during eyes closed verbal association. 88 Table 2. Distribution of anxiety signs by treatment. Number of subjects who reported the symptom during: eyes-closed eyes-open visual Type Of Symptom verbal verbal imagerya associationa association Sleepiness, tiredness 12 6 l Sweaty, clammy 2 2 Itching 1 Gastric distress l Tremors 1 l Muscular stiffness, aches and pains 3 l Nervous, shaky, jittery, fidgety, . restless, upset, anxious, appre- hensive, unpleasant excitement 8 10 2 Shock 1 Warm sensation 1 1 1 Cold sensation 2 2 2 Dizziness I 4 Sensation of floating, rocking, or motion 4 l Headache I Blushing 1 Blood rushing to head 1 Numbness, anesthesia 2 Limbs feel detached, enlongated, or fatter: sensation of being heavy or light 2 l 89 Table 2. Cont'd. M.- Number of subjects who reported the symptom during: Type of Sympton eyes-closed eyes-open visual verbal verbal imagerya associationa association Scared, frightened, afraid 1 3 Reduction in self-esteem; feeling dumb or stupid 8 9 3 Confusion, inability to think,' bafflement 14 12 5 Strong urges not carried out in behavior 4 1 Feeling confined, restricted, or closed in 1 l Disorientation in direction or place 1 1 Disorientation in time 1 Depersonalization l Apathy, blase l 1 Ego alien affect 1 l 1 Feeling abandoned, guilty, depressed, alone 2 l 1 Obsessive thoughts 2 2 Compulsive thoughts and behavior 1 6 Withdrawal reaction of psychotic proportion 1 Delusional thinking, Paranoid thinking 5 3 l Distortion of visual perception 1 ar=35 Number of subjects |l|||||||||||l||||llllllll||||||||l|||||||||||||||||||||||||l||||||||||l|||||||||li 12-- \V 90 E Visual Imagery V Verbal Association, eyes open E2E§ Verbal Association, eyes closed w— E s- g e- E. 41 E 2~ E _/ / 0 1-2 3-4 5-8 Number of types of anxiety manifestation an=36 Figure 1. Number of anxiety manifestations reported by each subject during verbal association with eyes open, verbal association with eyes closed, and visual imagery.a 91 Additional findings The small difference between the number of sentences spoken during each condition did not reach significance. Also, the correla- tion between the number of anxiety Sign types reported by the subjects and the number of sentences they spoke was negligible (.05). There were no order effects: that is, disregarding type of experimental condition, the number of anxiety sign types reported dur- ing the first, second, and third ten-minute periods differed only slightly. Subjects began reporting anxiety signs almost immediately, and their reports did not diminish in number as the experiment progressed. Discussion The results do not replicate Reyher and Smeltzer's (1968) finding that visual imagery shows more primary regulation than verbal association; neither do they support Reyher and Smeltzer's finding that visual imagery is accompanied by more anxiety than verbal asso- ciation. The failure to replicate the effects of the previous authors may be ascribed to the different procedures used in the present study; namely, the unstructured interpersonal and experimental situation into which the subjects were suddenly thrust. It is not surprising that most of the subjects found the experimental conditions anxiety- provoking. The provocative result is that minimal instructional dif- ferences led to large differences in number of anxiety reactions and smaller differences in type of reaction. We have devised tentative explanations of these differences. 92 Although many of the responses labeled here as anxiety mani- festations have been described in the sensory deprivation literature (see, for example, Freedman, Grunebaum, & Greenb latt, 1961), subjects in the present experiment experienced neither the nonpatterned nor stimulus-poor condition which typify sensory deprivation situations. Subjects were not alone in the experimental room, were free to move about in the reclining chair, and had full access to tactile and kines- thetic stimulation. Visual restriction (i.e., eye closure) when it was used (verbal association with eyes closed and visual imagery) did not consistently lead to higher numbers of anxiety Sign reports. Furthermore, reactions were reported immediately, whereas typical senwory deprivation effects have usually required several hours ex- posure to experimental conditions. Subjects in the present experiment did experience a deprivation, though, a type of social deprivation seemingly more anxiety-provoking than the simple absence of another person. Subjects found themselves in the highly unusual circumstance of speaking to a person who would not respond, verbally or nonverbally. We call this situation an extremely asymmetrical dyadic relationship with zero interaction. (A fifty-fifty distribution of initiated and received messages would be a symmetrical relationship, characterized by maximal interaction.) The high degree of asymmetry in the present study's experimental dyad was for the subject a kind of sensory depri- vation: expected incoming stimuli from another person were absent. Bruner's (1961) view of sensory deprivation is germane to the argument. He contends that sensory deprivation interrupts a pro- cess by which the organism evaluates incoming stimuli and accordingly 93 corrects the strategies it uses in perceiving, manipulating, and cog- nizing the environment; "the unhampered operation of the evaluation process is critical in the continuing adaptation of the organism" (p. 205). When the process is disrupted, as it is during sensory dep- rivation, the organism becomes preoccupied and sets off on a battle for adequate feedback that makes all other activities seem trivial. Bruner suggests that adjustment may be threatened and anxiety caused by any kind of prevention of an organism's continuous comparison of its percepts and cognitive structures with the actual state of the environ- ment. Prior to Bruner, H. S. Sullivan (1953) described in clinical terms an interpersonal function similar to Bruner's ideas in its em- phasis on maintenance processes vital to moment-to-moment adaptation. Sullivan termed the concept "security operations" and defined it as those activities (or lack of activity, e.g., selective inattention) undertaken by the self to protect and maintain self-esteem in an inter- action with another person. The anxiety concerning loss of self-esteem and the activities required to allay it are so important to the self that they supersede all other feelings and psychological functions. The similarity of the views of Bruner and Sullivan is not merely analogy. Rather, Sullivan defines a particular case of Bruner's more general hypothesis. Other people are a part, albeit an important part, of an individual's external environment. Stable, symmetrical, or slightly asymmetrical relationships Often have familiar security operations built into them. A student, for instance, derives part of his self-esteem from the assurance that 94 he (the student) has satisfactorily performed the tasks set by the in- structor. However, since the situation in the present study was so unusual and unstructured, subjects found that security operations appropriate and effective in other interactions simply did not work in this one. The experimenter did not interact with subjects, so sub- jects could not directly appease him or accede to his wishes: and how to complete successfully the task set by him was so unclear that sub- jects had no idea when they had done so. Thus, self-esteem was threatened. Almost all subjects then tried, by asking questions, to garner information about the purposes of the experiment, information which would have resolved the uncertainty in the situation and made it possible to perceive, and thus to satisfy, task demands. When this information was withheld, anxiety seemed to be aroUnsed in most subjects. (One subject, whose data were not used, bolted from the room in the middle of the experiment, and two others needed what amounted to crisis intervention after the completion of the experiment.) This was the point of failure of interpersonal maintenance functions, or security operations, and most subjects then reported anxiety manifestations. Obviously, though, such disruptions do not happen frequently. Under most conditions, individuals can interact with others in a variety of ways. The relationship of a subject and an experimenter previously unknown to one another, however, is severely curtailed. The subject has only one way to maintain self-esteem: satisfy task demands. If this single method cannot be used, the maintenance process is disrupted. This task ambiguity hypothesis accounts for the significantly lower number of anxiety reaction types reported in the visual imagery 95 condition than in the verbal association conditions by emphasizing that visual imagery is a more unusual task than verbal association. It is common to monitor one's thoughts; it is also common to communi- cate them. On the other hand, it is a relatively novel activity to monitor and describe one's visual images. If visual imagery was per- ceived as a more well defined and novel task than verbal association, subjects could discriminate when they had completed it (i.e., by coming up with an image). If it is agreed that perception and satisfaction of experimenter demands is a way for subjects to maintain self-esteem in an extremely asymmetrical relationship, then visual imagery was less threatening than verbal association. Anxiety, consequently, was not as frequently or intensely experienced. Another possible explanation of the reported anxiety signs is that, contrary to Reyher and Smeltzer's (1968) findings, verbal association is a more effective uncovering technique than visual imagery. According to this view, verbal association allows more derivatives of unconscious conflict to elude the censoring defenses; the derivatives, now conscious, produce anxiety. A greater number or intensity of anxiety-producing derivatives in verbal association than in visual imagery, though, should have produced a greater amount of primary process in verbal association. Since the amount of primary process content did not differ between experimental conditions, the intra- psychic hypothesis can be discarded. In fact, the anxiety-provoking interpersonal situation seemed to overshadow any possible differences between the uncovering effectiveness of the three experimental condi- tions. 96 Although the incidence of most types of anxiety manifesta- tion was too infrequent to permit statistical comparison across treat- ments, we nevertheless generated some hypotheses to account for the observed "differences." In fact, the qualitatively different nature of some of the anxiety manifestations reported by subjects in the two verbal association conditions suggests that anxiety may not be associ- ated with a unitary and consistent pattern of neurophysiological activity. Instead, the type of innervation and manifestation of anxiety may covary with certain organismic variables (e.g., open vs. closed eyes). Jr For instance, dizziness and sensations of floating or rocking occurred almost entirely within the eyes-closed verbal association 1 condition. Eye closure and anxiety perhaps interaCt in a way that affects the vestibular system. It is interesting to note that neither of these anxiety signs were ever reported during visual imagery, even though during visual imagery subjects' eyes were closed. In the same vein, one would expect reports of sleepiness to occur most frequently when eyes are closed. This expectation is born out by the fact that 12 subjects reported sleepiness during eyes-closed verbal association. Yet only one subject reported sleepiness during visual imagery. Again, eye closure was not sufficient to produce this anxiety sign, but eye closure plus anxiety (eyes-closed verbal association) produced more reports of this sign than anxiety alone (eyes-open verbal association). It is intriguing that numbness occurred only in eyes closed verbal association, but the low number of reports of this anxiety sign do not allow even speculative hypotheses. Strong urges not carried out in 97 behavior (e.g., "I'm just going to get up and run right out of here.") were reported exclusively during eyes-closed conditions (four during eyes-closed verbal association, one during visual imagery). This pattern suggests that closed eyes are more conducive than open eyes to fantasied escape from an anxiety-provoking stimulus situation. That these responses did not occur during eyes-open verbal association further suggests that, with open eyes, subjects could more effectively defend against anxiety and thus could tolerate their continued presence in the experimental room. The compulsive verbalizations may have had a partially situ- ational cause. Because of the small size of the experimental room, much of subjects' field of vision was filled by the wall and ceiling. Since these surfaces were covered with accoustical tiles, there was probably a stimulus pull for subjects to deal with anxiety generated by the experiment by compulsively counting and ordering the holes in the tiles. Interestingly, four subjects mentioned during the experi- ment that they usually dealt with anxiety at a physician's or dentist's office by counting and ordering the holes in acoustical tiling on the waiting room walls. This particular defense was not available to sub- jects when their eyes were closed, which explains why more compulsive verbalizations were observed during eyes-open verbal association than during eyes-closed verbal association. The frequent confusions and inability to think are probably directly related to subjects' uncertainty about task demands. That is, out of all the things the subject had in mind, what did the experimenter 98 want to hear? The paranoid thinking seems to have been due to subjects' perception of the experimenter as a threatening figure, one who with- held the means of reducing their discomfort. Several other manifestations of anxiety, such as sensations of heat and cold, are anxiety concomitants stemming from activation of the autonomic nervous system. Others, like complaints of stiffness, are also anxiety concomitants, innervated by the somatic nervous sys- tem. The remainder of the anxiety signs probably reflect the charac- teristic styles of defense of those subjects who exhibited them. Our observation that eye closure during stress adversely affects some neurophysiological systems and increases the incidence of non-coping defenses (desire to escape), if confirmed, has implications for the psychophysiology of anxiety as well as in the practice of dentistry and medicine. Requesting the patient to keep his eyes open during injections and suturing may reduce the incidence of fainting, nausea, and bolting, particularly if the patient is given the oppor- tunity to engage in an obsessive-compulsive defense. In any event, if patient discomfiture can be reduced, fewer individuals would be motivated to avoid essential diagnostic and treatment procedures. As Hilgard (1969) observed in comparing verbalized and physio- logical responses to experimentally induced pain, the former is the more orderly and meaningful. With this in mind, the spontaneous verbal manifestations of anxiety under different conditions may elucidate the functioning of a variety of neurophysiological systems in more differ- entiated and informative ways than is possible purely on the basis of electrophysiological responses. 99 References Bruner, J. S. The cognitive consequences of early sensory deprivation. Burns, 8. Freedman, In P. Solomon, P. E. Kubzansky, P. H. Leiderman, J. H. Mendelson, R. Trumbull, & D. Wexler (Eds.) Sensor de riva- tion. Cambridge, Mass.: Harvard University Press, I961. P. The activation of posthypnotic conflict via free imagery: A study of repression and psychopathology. (Doctoral disser- tation, Michigan State University, 1972). Dissertation Abstracts Internation, 1972, 55, 911B. (University Micro- films No. 72-22, 198). S. J. Grunebaum, H. U., & Greenblatt, M. Perceptual and cognitive changes in sensory deprivation. In P. Solomos, P. E. Kubzansky, P. H. Leiderman, J. H. Mendelson, R. Trumbull, & D. Wexler (Eds.), Sensory deprivation. Cambridge, Mass.: Harvard University Press, 1961. Hilgard, E. R. Pain as a puzzle for psychology and physiology. Ameri- can Psychologist, l969, 25, 103-113. Holt, R. R. Manual for the scoring of primary process manifestations in Rorschach protocols: Tenth draft of an experimental procedure. New York: Research Center for Mental Health, 1970. Kirk, R. E. Experimental design: Procedures for the behavioral sciences. Belmont, Ca.: Brooks/Cole, 1972. Morishige, H. & Reyher, J. The alpha rhythm during three conditions of visual imagery and emergent uncovering psychotherapy: The critical role anxiety. Journal of Abnormal Psychology, in press. Reyher, J. & Smeltzer, W. Uncovering properties of visual imagery and verbal association: A comparative study. Journal of Abnormal' Psychology, 1968, 15, 218-222. Singer, J. L. Imagery and daydream methods in psychotherapy and be- havior modification. New York: Academic Press, 1974. Sommerschield, H., & Reyher, J. Posthypnotic conflict, repression, and psychopathology. Journal of Abnormal Psychology, 1973, 52, 278-290. Sullivan, H. S. The interpersonal theory of_psychiatry. New York: Winer, B. John Wiley & Sons, 1953. J. StatisticalrprincipJes in experimental design, New York: McGraw-Hill, 1962. APPENDIX B APPENDIX B Subjects: Please do not write on this sheet. The experimenter will provide paper for you. Also, please be as honest as possible--this is very important. Questions: 1. Did you have any thoughts about the instructions you received from the experimenter? How did the instructions affect you? 2. Have you ever been in this laboratory before? If so, what happened in the experiment? 3. Have you ever heard of this lab before? If so, what have you heard? 4. Do you know what professor runs this lab? If so, what have you heard about his work? 5. What did you expect the experiment to be about before you came? Be1as specific as you can--even any vague thoughts you had would be he pful. 6. What do you think the experiment was about? In other words, what was its purpose? Again, be as specific as you can. 7. Other than repeating the instructions for you (if you asked), did the experimenter speak to you during the experiment? 8. If the experimenter spoke to you during the experiment, how did you feel about what the experimenter said to you? How do you think the experimenter felt ab out what ypp_said? 9. If the experimenter did not speak to you during the experiment, how did you feel about the silence? How do you think the experimenter felt about what ypp_said? 10. Did you have any other feelings or reactions about the experiment that you think it might be useful for us to know? 11. Do you think we will find the data you have provided useful? 12. How do you think you did in the experiment? 100 APPENDIX C APPENDIX C Symptom Descriptions for the Scoring of Free Verbal Material The following descriptions of symptoms are from the Sympto- matic Reaction Scale (Burns, 1972; for general descriptions see also Perkins & Reyher, 1971; Sommerschield & Reyher, 1974). Some of the Scale's original categories have been eliminated, and others have been condensed for the purpose of scoring free verbal material. Also, the original Scale groups symptoms by hypothesized sources of innervation. The modification presented here has eliminated these groupings since they are not relevant to the aims of this study. It is important to bear in mind that no description of a symptom is scorable unless the behavior or experience in question re- fers directly to experimental conditions. Thus, for example, a phobia of open places is not scorable: a phobia of the polygraph being stored in the room is scorable. Another example: anxiety about a friend's reaction to the subject's remark of the previous evening is not scor- able; anxiety about what the experimenter is thinking is scorable. Subject reports any of the following: 1. sleepiness, tiredness. 2. tingling sensation. 3. itching or weals. 4. feeling sweaty or clammy. 5. abdominal pain or gastric distress. 6. chest pain. 7. cold sensation. 8. goose flesh, shiver. 9. dryness of mouth, hard to swallow. 10. tachycardia, heart pounding, heavy pulse. 11. breathing faster or heavier. 12. sensation of warmth. 13. dizziness or lightheadedness. l4. headache, throbbing in head, pain behind eyes. 101 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 28. 29. 30. 30. 31. 31. 32. 33. 34. 102 watery or irritated eyes. blushing. blood rushing to head. tic, muscle quiver or twitching. tremors, shaking. stiffness or tightness or muscles, cramp or tension of muscles. aches or pains (score a painful cramp here.) temporary blindness. temporary deafness. temporary paralysis. numbness or anesthesia. limb or other body part feels detached. limb or other body part feels elongated. limb or other body part feels as if it had disappeared. limb or other body part feels heavy. sensation of bodily heaviness. sensation of bodily lightness. sensation of rising. sensation of being squeezed. sensation of sinking. sensation of being stuffed or bloated. sensation of floating. sensation of rocking. 34.25. sensation of tipping. 34.5 sensation of shaking or vibration. 34.75 sensation of spinning or revolving. 35. 35. 36. 37. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 103 sensation of being outside of one's body. sensation of falling. disorientation in direction or place. disorientation in time, e.g., "I can't tell if I've been in here ten minutes or an hour." dislike of closed eyes, desire to open eyes. feeling like someone else or experiencing a loss of personal identity. Actions may seem nonvolitional, automatic, mechanized. amnesic or fugue state: a specific loss of memory for some as- pect of the experimental situation. inability to think, expressed difficulty in thinking, blank mind. bafflement or puzzlement, confusion. Score 5fs expression of lack of understanding of why a particular thought or feeling comes to mind. strange movements in the body, such as "something racing up and down." ‘ alteration in state of consciousness: feeling "stoned“ (unless, of course, the subject is intoxicated), "out of it," "feel like I'm talking in my sleep," “even though I'm talking, it's like I'm daydreaming." verbally expressed anxiety: fidgety, jittery, restless, anxious, upset, troubled, worried, on edge (unless this indicates irrita- tion), shaky (other than muscular tremors), funny (unless this is scorable as ego alien affect), uneasy, uncomfortable (other than physical discomfort), apprehensive, tense (other than muscu- lar tension), nervous. Again, a statement like, "I'm worried about my chem test" is ppi_scorable. fear, fright, scared. apathy or blase. feeling weird, strange, odd, unreal, crazy, unnatural, foreign, spooky, eerie, or other ego alien affect. There must be a dis- tinct reference to a negative feeling, surprise or shock, since some of these words are used colloquially for quite different purposes. feeling depressed, sad, down, or unhappy. feeling abandoned or lonely: this may also be scored if the sub- ject is afraid the experimenter has left the room. 50. 51. 52. 53. 54. 55. 56. 57. 58. 104 feeling guilty. feeling disgusted. feeling bothered, feelings of dislike, impatience, anger, irrita- tion. obsession: any perseverative thought, tune, image, verse, etc. Obsession must be scored carefully, since repetition can denote concern rather than obsession. For instance, unless the subject is unaware of why he continues to dwell on his inability to think, repetition of the inability to think is not scorable--the inability to think may be of real concern to the subject, and may even be somewhat frightening. In general, it is only the exceptional case which can be scored without the subject's awareness of its intru- sive nature. The sentence scored for obsession must usually 55. intrusive, i.e., it must not follow in the subject's remarks. For instance, a subject might say, "I can't figure out why cathode tubes keep occurring to me, but they do." This sentence would be scorable (and is, in fact, an example from the earlier data). An example of a score without the subject's awareness of the thought's intrusive nature is one subject's repetition (15 times in ten Einutes) of his perception that the sound of a fan sounded like ees. compulsive behavior: any behavior or intention designed, in the judgment of the scorer, to increase the symmetry, neatness, or order of the experimental room, materials, or personal clothing. This includes such "mental games" as mentally drawing lines be- tween things to make geometric forms. Also score descriptions of the room. phbbias: fear of a specific object in the experimental room, or of the room itself. conscious denial or suppression of thoughts or affect: e.g., "I'm just not going to think about that." reduction in self-esteem: experience of failure, feeling inade- quate, feeling as if "I really goofed up here," feeling dumb, stupid, idiotic, self-conscious, embarrassed, humiliated. Include self-deprecating remarks, if relevant only to the experiment. Include verbalized concern with the quality of performance, but only if subject expresses worries in this area, e.g., wondering how "well" others do in the experiment, feeling that he (5) is not doing what "should be done." projections: subject imputes to the experimenter desires or thoughts that the experimenter has not expressed and that the subject has no objective reason to suspect. 59. 60. 61. 62. 62.5 63. 64. 65. *66. *67. *68. 69. 105 delusions: grossly unrealistic thoughts. Do not score if also scorable as projection. positive hallucination: seeing or hearing something that is not there. negative hallucination: subject cannot see something (or hear something) which is there. perceptual distortion of room elements, e.g., subject notes that sounds change quality or location when they do not, things in the room change size or Shape, sounds become louder or softer when 5 can detect no change. If there is any doubt--if the sound, etc., could have changed in the way 5 perceived it--do not score. explosive feelings, no specific location. explosive feelings in stomach. explosive feelings in chest. panic or terror: this must be qualitatively and quantitatively different from fear or fright. Do not score this category unless you are quite sure that another scorer would agree that the phrase in question actually expresses panic or terror. inappropriate humor. Although this category has been used in the past, it will not be scored for the present study. feeling confined, restricted. head feels pressed, tight. desire to leave the experimental room. * These symptoms were assigned lower numbers in the original Scale. Because the present version is only a nominal scale, though (i.e., the size of the number assigned to each sympton has no theoretical importance), it makes no difference that these symptoms have been assigned a different position relative to the other symptoms. APPENDIX D APPENDIX D REPRESENTATIVE TRANSCRIPTS Transcripts of varying length are presented here. Introduc- ing each transcript is the number assigned to that subject (to make cross-reference to Appendix E possible), the number of words the sub- ject spoke, sex of the subject, the experimental treatment received by the subject, and the SRS scores the transcript received. The numbers in parentheses after the SRS categories indicate the number of such scores the transcript contains. Subject #65. 2080 words. Negative evaluation group. Male. Inability to think (4), sleepiness (2). OK - a - I'm really glad to be here cause I really need these points for psychology. This is the first experiment I've been to and I kind of think I waited too far to the end. I don't know if I'll be able to get them all in now. But a - like I said, I really need the points. I didn't do too well in the first test and a - . I can't wait to go home this weekend because I have been doing a lot of work. And - a - I can't wait to get - go home and see some of my friends who are coming home from Western. They get out this week for a Vacation. They have a ten day vacation. So a lot of them will be home and a lot of my friends from other schools are coming home too. So it will be sort of like a reunion. I don't know, we will probably go out on Friday or Saturday night and go to the bar ow something like that. But I'm not really sure, my roommate's going home with me - a - areally kind of tired now too. Because a - I've been working hard and I stayed up late last night to get some homework done and I really like laying in this chair. I feel like I'm going to fall asleep, A - wish I had one of these in my room. A - I can't wait 'til the semester is over 106 107 already cause - a — I've been doing a lot - . I've really been working hard this term trying to get my grades up but - haha - it's not working too much - a - I'm still up with the work but I'm just not doing too well on the test. So I'll really just have to keep working. Ane - I - a - just got done with dinner I went to a - to Sears with a few of my friends to get a few things for our room and a few things for myself. One of the friends I went with has a car and he drove us so that was really good. Cause I don't like walking in this kind of weather. In fact, I hate this kind of weather. And this is a - and that is another reason why I'm really down cause of all the work and the weather. Really bugs me but I'm sure it will start getting better pretty soon. A - today wasn't such a bad day, I only had two classes and a - one of them I got out of early so I didn't really have tOo much work to do. And I've been thinking about the experiment all day. Didn't know what it was going to be about or what I was supposed to do but I was, you know, just willing to do it for the points. Cause I really think, I really need them. Am - a - got a - not going to be able to go home 'til late this Friday cause my fri- - the way I'm getting home is my friend, one of my friends' from Western is going to come and pick me up and a - we have to - have to wait for him to get in so I won't be able to leave 'til late. And a - two weeks ago I had to go - or last week I had to go home to a funeral. One of my - a girlfriend I know, her mother died and so - a - it was really sort of depressing. It was really sort of depressing. It was the first funeral I had ever been to and I really - really felt sorry for her cause she was really close to her mother. And “I'nnsure it really hurt her but she seems to be 108 doing alright. I talked to her a few times and she has got to go home this weekend too because a - her father wants her to come home now a little more, since her little brother - cause she has a smaller brother and since - she doesn't have a mother anymore - her father thinks that she should come home a little more now. So she will be going with us and probably my roommate will be going too. And a - Hm - am - I don't know how my classes are going to be next semester. I'm going to try to schedule a little easy - a little easier classes because I don't want to work too hard in the spring. You get kind of lazy as you get towards the end of the year. And with the nice weather, I don't want to spend all my time inside doing homework so -. I might a - I'm probably going to take one or two gym classes so I can get outside and do more of the things I want to do. Am ' I switched my major in the first a - after the first term, from Building Construction to Business. And a - I've got to take a lot of different cla-ses now and it is a lot harder. First time, I had math and chemistry, which I am really good at. But now I have more English classes and a lot of reading classes. Geography and Psychology, which a - I'm not doing too well in, cause I'm not really that good of a reader. So I'm having a hard time now. But I think I can do alright as long as I keep up with the work. And - a - I think I'm going to be an accountant, — if I — ,if I can get through my four years here but a - I'm not really too sure yet. I might change my major again. Not really sure - the first - the reason I was going into Building Construction first was because my father was a builder. And a - I worked for him in the summer and I really - I really enjoyed, you know, I didn't enjoy so much the work 109 I did but a - I liked the kind of work he does. And I really thought that's what I'd like to do but I'm not so sure anymore. And - a - my mother just - a my mother just went back to school too, so now we have three people in our family in college. My sister goes to college in California, UCLA. And she's living out there. I think she's living out there now and I think she's going to stay out there for good. She comes home once in a while though and I haven't seen her in a long time. But I think she will be coming in Easter and I get to see her. And - a - like I said, my mother went back to school and she's taking account - she's taking an accounting course too. And according to her, it's really hard. She's having a hard time but my father gives her help, so I guess she is going to get through. Am - she - she already gradu- ated from U of M. K don't know how many years ago but she - she didn't have much to do now that most of the kids are away. Now that all of them but one are away. So she a - she went back to school just to have something to do. And a - I guess she really likes it. She - a she used to be an English teacher and she a - so she is doing pretty well in all those courses. But stuff like math, she was never too good in so she is having a hard time in her accounting and data process- ing course. Am - my younger sister is still in a - still in high school and a - she is going to the same high school I did. And not too long ago, we just a - we moved to where we live now and a - so it was a big change for her and for me. Cause we were both - she was in junior high and I was in high school. But it didn't take too long to get adjusted. And then I came up to school here and a - it was a lot harder than I thought it would be. I, you know, I didn't do a - I - 110 got pretty good grades in high school but - I didn't do that much work. And when I got up here I really had to change my ways, to get by. But I'm doing alright - I'm doing pretty good now so it's not that bad. Am - haha - well - I'm running out of things to say. Am — after - probably when I get done with it, I'll have to go back and do some more homework. Got a lot of reading to do still so. I'm not looking forward to the rest of the night. I'll probably be up late tonight. But - haha - I guess it will all pay off in the end, I hope. Probably the class I don't - I don't like the most is - a psychology. I don't know why, I just - I guess I'm not - psy - I'll never be a psychologis- psychiatrist, or psychologist. I just - I can't get the hang of a lot of these ideas and stuff that I read. And it's really given me the most trouble of all my classes. So it's just not the thing for me, really. (Pause [9 sec.]) I don't even know - I'm not sure if I'm going to come back here next year. I might switch to another college. But I'm not really sure about that yet. I wouldn't - I'm not sure where I'd go or if I'd switch at all. But I'm going to give it some thought over the summer. I might even take a one or two classes over the summer at LCC or some other - some other small college. If I don't get a - if I don't get a job. And I really hope to get a job this summer, cause I didn't do anything last summer. And I really got bored about the middle a, you know, about the middle of - the middle of Jun- - July. And a - I had nothing to do in the day. All my friends were working and most of the time I would just spend at home either sleeping or eating. And then I'd go out and stay out all night. And do the same thing every day so it got kind of boring and a - you 111 really can't do much when you don't have any money either. So I really hope to get a job this year, this summer. I'm going to try and get one at a furniture store that my friend works at, delivering furniture. Because he is moving. He's going to give up his job so Ill try to get a job there. I worked there for a couple days over Christmas vacation and the guy has got my number and everything so. And I've got his number so I'm going to give him a call and see if he needs anybody to work when this guy leaves. Am - hm - amm - whin I get back to my room, I've got to call that girl. And tell her what time we are going to pick her up. Because a - so - because she won't be able to leave until four, four-thirty. So I'll tell her that a we'll be there and if she is ready and if she's not she is going to miss her ride. She's a pretty good friend of mine. I went out with her a few times during a -high school but she-s got another boy friend now so I don't see that much of her. But we're still really good friends. I'd like to see more of her really but well she a - she's working hard and she's not - she lives quite a ways from where I do so I don't get to see her that often. Am - getting harder to think of things to say. Running out of all my thoughts. Am - it's not - the room is not - like where we are doing this is not what I expected. I expected it to be a bigger room and I thought I might be filling out some survey or something. But a - I guess I was wrong. I don't know, I gotta - I think I'm going to do another psychology experiment tomorrow. I signed up for one but I don't even know what it's about so a - I don't know what it's going to be like. This is kind of nice though, just laying here and a - just letting every thing go out. Just talk about what I feel. What I 112 think and about - a - I'm running out of things to say. Am - I don't know what your - what this is all about or what's, you know, what's the idea of it but a I'm sure you know what you're doing. I'm sure, I'm sure you have a good reason for it. Am - when I first came in I saw that thing on the wall over there. And it didn't frighten me but I just wondered what it was. It looked kind of, kind of funny to me and kind of strange so. I didn't know whether you were going to use that or not. I don't know what the deal is but - 113 Subject #60. 1190 words. ,Negative evaluation group. Female. Inability to think (1), verbally expressed anxiety (1), reduction in self-esteem (2). You want me to just start talking (pause) well (laugh) this is kinda weird, I don't know uh uh. I don't know we had well my roommate's in a fight with her boyfriend which is totally burning me out cuz she's been walking around for a week and everything that comes out of her mouth is about him and its just kinda getting to the point where it's driving me crazy cuz I can't even go to sleep at night cuz she talks to me til 3 in the morning about it and she kept me up all night last night. So today it was really shitty cuz I couldn't hardly make it to my classes. I was so tired and you know it's weird uh (pause) um ah this is just really strange we went to dinner and 3 girls sat with us that (laugh) haven't talked to us for like a year they just sat there talking about how they were gonna get a house and on and on and on and on and it was kinda weird I don't know and my brother sat with me for like the first time in ages and started lecturing me cuz I don't have a job yet and we had the grossest meal the weirdest hot-dogs (laugh) they were about 2 feet long and a quarter of an inch thick and (pause) and I was sitting at the meal and there was this guy sitting next to me just talking to this girl I didn't know and all of a sudden she just took a coke and poured it down his shirt and he just he was just sittin' there smokin' a cigarette and he was just totally calm didn't even flinch didnt do anything just sat there with a whole glass of coke in his shirt and she just got up and walked away couldn't figure out that one (laugh) I don't know what she was doing that was kinda weird today was weird (pause) um before I wasn't really livin' in the 114 dorm like the first 2 years and so I've really only been there for like a COUple of weeks it's just the strangest place oh god (pause) I don't really know what I'm supposed to say (laugh) and since I moved back into the dorm like usually I can study really easily and stuff but since I've been back in there I haven't been able to do any of my work and I did the dumbest thing last week like the first time ever I just went in and took an exam without even studying for it it was like so stupid but I must have been just like really scared or something cuz I lucked out and got a 4 point on it I couldn't believe it I got the test back today and went out bought an album (laugh) to celebrate cuz I haven't bought an album in 2 years (laugh) and I don't know (laugh) um (pause) and uh (pause) went uh (pause) my classes are weird and uh I am just kinda fed up at this point today was kind of just totally blah all over was thinking of going home this weekend which I haven't done for a long time and um um I'm just getting totally nervous sitting here (laugh) babbling (laugh) and but (pause) I really don't know what to say (long pause) I'm glad its gettin' nicer outside didn't even have to wear gloves walking over here (long pause) I just broke up with my boyfriend about 2 weeks ago and I don't know started going out with this other guy just totally uh I don't know what I want at this point but (pause) my old boyfriend keeps popping up everywhere haunting me (laugh) (long pause) really can't think of anything to say after dinner I usually don't function for a couple hours (laugh) I just Sit around watching the news or something (long pause) You mean I have to do this for an hour (laugh) (pause) I was reading this book for my one class and its Judicial Policy making so I decided I'm going 115 to be a judge (laugh) cuz half these judges don't know what they're doing (pause) its weird cuz none of my friends want to do anything all my girlfriends are getting married and I'm gonna be like the only one left (pause) and we've been out lookin' for a house and already one of the people is getting married and possibly 2 others are gonna get married god by the time we get the house I'm gonna have it all to myself (laugh) which won't be too good but it's such a drag we went out just walkin' around like door to door to every house and like the first time we went out everybody was really bitchy to us or else nobody was home and then and we went out Saturday and everybody was like really nice they'd come out and tell us what houses to go to but it's like we're having no luck at all they're all like really big houses and we need a three room house and we just kept lookin' and lookin' and like we looked all afternoon Saturday we started like behind Dooley's and we worked every house from like 4 blocks back from Grand River all the way down to Gunson and the very, very last house we went to was one and it was just like perfect you know little house with three bedrooms it wasn't very much it was like really close to campus and the guy said that they might sell it to a professor but if they didn't then we could have it but it's just totally burning me out looking for a house so we're really excited and we hope that um we're going to get that one and it's, it's so strange cuz in the dorm like everybody's just like cuttin' each other's throat to get a house nobody'll give out any information and like people that are your really good friends won't talk to you cuz you know they don't want to tell you where they might possibly have a house cuz they're afraid you might 116 run out and get it the girl down the hall who you know we've been really good friends for 2 years won't talk to us cuz she's looking for a 3 room house too she just keeps coming up making little remarks about you know how she has one and you know on and on haven't we found any- thing yet you know and its just its really sickening like everybody's just totally on each other's backs about it its just kind of stupid but I couldn't see living in the dorm again I don't know 2 years is plenty but (long pause) I don't know 117 Subject #31. 100 words. Positive evaluation group. Female. Inability to think (1), reduction in self- esteem (l). I can't really think naturally if you talk when I'm thinking when I feel I'm being taped. (long pause) I don't really understand what you want me to do you just want me, any thoughts that come into my head. (repeat instructions) (long pause). It's not really what I expected of this experiment. I thought it would be a lot different like a questionnaire like the rest of them. .It's kind of strange. (long pause) I don't think I make a very good subject for you because I can't do this. (long pause) It's hard trying to say what I'm think- ing. I think I'm trying to think of things to say, (long pause) 118 Subject #7. 645 words. No evaluation group. Female. Sleepiness (2), stiffness or tightness of muscles (9). pains (l), sensation of floating 1;, sensation of rocking 7), deSIre to open eyes 2 , verbally expressed anxiety 2), self-esteem reduction (5), explosive feel- ings (l). (pause) um, I want to see if I can transfer to a different school, um, it's a nice day outside, there's sunshine, uh, now the chair feels like it's moving. Um, I'm very tired and I can feel the pain in my back, it's tired. (pause) I get, it feels like the chair is moving back and forth, my feet are going down and my head going up and it goes back again. (long pause) This feels like a continual movement in my body, the chair. (long pause) It just, it just keeps moving. (pause) I think I'm all tense and uptight inside and I can just feel different parts of my body trying to relax. (long pause) It sort of feels like floating in space, you just, it feels like I'm moving but nothing is inhibiting me. (long pause) The chair just keeps it feels like it's moving, that's all (long pause) I'm thinking about what I'm supposed to be saying or what I'm supposed to be thinking about, I really don't know what I'm supposed to be doing. (long pause) It seems like I got so much to do today, after this I've got a meeting, a class, and another meeting, it's just that Tuesday's are awfully busy for me, just to have to get up tight day. (pause) I'm awfully tired. I wish I could find a place off campus, I'm going through a hard time looking for a place off campus to live next year. (pause) So many things you have to look for when you look for a house. It's just craxy, some of these places I've seen, just crazy. I couldn't even live there. (long pause) I would really like to find a place to live outside of 119 the dorm without having to go home or having to transfer to a different college just because I can't sleep at night and people all they want to do is smoke and drink. I wish it was quiet at night in the dorm, but it just isn't, it's too bad. (long pause) I don't know, I have a hard time keeping my eyes closed, I want to look around the room, when I first came into the room I saw something against the, the wall, looked like a box with different gadgets in it, ( ? ) mechanical. (pause) I'm awfully worried about what my voice sounds like on tape. (long pause) When I signed up for, signed for this I didn't know what cognitive thinking was, so I don't know what I'm doing really. (long pause). I don't know, on Ms. Katz lecture she sure don't know what one's talking about half the time, she either goes too fast or she talks about stuff that really has no meaning behind it, or the examples sometimes, well they'll, they don't have any relevance to the topic. (mumble) I feel all tense and my legs are tight, my muscles are just tight, I feel stiff as a board now, not moving at all. I'm very tense and, just locked up, all the energy locked up inside my body. I feel like a time bomb, ready to go off, just pressured, just pushed. (pause) I'm not understanding my classes, I'm having a hard time in my classes, I'm worried about my ec--economics class. I don't really understand what he's talking about at all but I don't have, I just don't know if I should go up and talk to him or if the whole class is lost like I am, or just what I should do, I just, don't know what to do in that class. That and my psych. class, I just don't know what to do. (pause) I'm all tense and tied up right now talking about it. (long pause) I wish I was back home. (long pause) Well, I'm just all 120 uptight, just ( ). (long pause) I don't know, what am I supposed to be saying, am I saying anything right, or, nothing's right and nothing's wrong, I don't know. (long pause). I feel like I'm suppoSed to be APPENDIX E APPENDIX E RAW DATA The following table represents each subject's raw data. The column headings are, for the most part, self-explanatory. Experimenters JP and BK are male, LJ and LW are female. The column at the far right, labelled SRS Category (freguency) is the summary, by individual subject, of the kinds 0 anz ety s gns escribed and the frequency with which each category was mentioned. Each number in this column outside paren- theses represents a particular type of anxiety sign (to match a number to a sign, see Appendix C), while numbers inside the parentheses rep- resent the frequency of that sign's description. The column "2 fre- quency" is the sum, by subject, of the numbers found in parentheses in the column to the far right. The numbers assigned to subjects (Subject #) are arbitrary: they do not represent the order in which the subjects participated in the experiment. NO EVALUATION (UNCERTAINTY) GROUP FEMALES . No. words 2 2 Subject # E Spoken Types frequency SRS category (frequency) 1 JP 275 10 17 1(1), 14(1), 37.5(1). 40(3). 44(3). 45(2). 48(1). 52(3). 67 1). 2 JP 1145 1 1 44(1). 3 JP 220 o o --- 4 BK 450 2 2 40(1). 44(1). 5 BK 710 o o --- 6 BK 925 7 14 1(2). 5(3). 40(2). 41(1), 44(2), 56(1), 57(3). 7 LJ 645 9 30 1(2), 20(9). 21(1). 33(1), 34(7). 37.5(2). 44(2). 57(5). 62.5(1). LJ 30 o o --- 9 LJ 770 4 9 1(3). 41(2). 44(3), 49(1). 10 LW 575 3 4 40(1). 44(2). 57(1). 11 LW 555 3 4 12(1), 40(1). 57(2). 12 LW 325 4 6 1(1), 12(1), 34.75(1). 40(3). 121 122 MALES 13 JP 935 1 1 57(1). 14 JP 320 3 5 40(1). 54(2), 57(2). 15 JP 280 4 5 41(2), 44(1), 54(1), 67(1). 15 BK 525‘ 7 21 23(3). 40(1). 44(4). 47(2). 57 9 , 62 1 , 67 1 . 17 BK 415 8 40(3), 41(1), 44(1), 57(3). 18 BK 205 6 7(1), 40(1), 54(1), 57(2), 69(1). 19 LJ 180 O 0 --- 20 LJ 515 2 3 10(1), 44(2). 21 LJ 365 6 9 30(2), 31.5(1), 35(1), 40(2), 42(1 , 47(2). 22 LW 150 3 3 40(5). 44(1). 57(1). 23 LW 530 0 0 --- 24 LW 400 6 l4 37.5(6), 40(1), 44(3), 52(1), 52(1), 57(2), 69*1). POSITIVE EVALUATION GROUP FEMALES Subject E No. words 2 ('2‘, # Spoken Types frequency SRS category (frequency) 25 JP 880 O 0 --- 25 JP 135 5 9 12(1), 40(3), 44(3), 57(1), 62 1 . 27 JP 350 2 2 1(1), 40(1). 28 BK 645 12 25 1(3), 7(1), 11(2), 12(2), 19(1), 37.5(1), 40(5), 44(3), 57(2), 58(2), 67(2), 69(1). 29 BK 185 3 3 1(1), 35.5(1), 40(1). 30 BK 50 1 1 45(1). 31 LJ 100 2 2 40(1), 57(1). 32 LJ 215 3 3 21(1), 40(1), 44(1). 33 LJ 250 3 3 40(1), 44(1), 57(1). 34 LW 1485 5 7 1(1). 34.75(3). 40(1). 47(1). 58(1). 123 POSITIVE EVALUATION GROUP FEMALES Subject E No. words 2 Z # Spoken Types frequency SRS category (frequency) 35 LW 910 1 1 40(1). 36 LW 955 3 3 1(1), 12(1), 60(1). MALES 37 JP 230 l 1 44(1). 38 JP 1100 0 O --- 39 JP 0 0 0 --- 40 BK 1630 4 4 25(1), 34.5(1), 40(1), 47(1). 41 BK 445 5 5 31.5(1) 33(1), 34.5(1), 41(1), 45(1). 42 BK 1345 3 4 40(1), 44(1), 57(2). 43 LJ 80 l 1 1(1). 44 LJ 245 2 3 1(2), 57(1). 45 LJ 655 8 13 19(1), 28.5(1), 34.25(l), 34.75(6), 37.5(1), 38(1), 43 (1). 47(1). 46 LW 585 6 10 18(1), 37.5(1), 40(4), 41(1), 47 LW 1285 0 0 --- 48 LW 800 O O --- NEGATIVE EVALUATION (CERTAINTY) GROUP FEMALES Subject E No. words 21 2 # Spoken Types frequency SRS category (frequency) 49 JP 370 6 10 7(1), 40(1), 44(3), 52(3), 56 (3). 56(1). 58(1). 50 JP 1085 2 4 40(2), 44(2). 51 JP 365 l 1 40(1). 52 BK 440 5 21 1(1), 40(5), 44(4), 57(10), 58(1). 53 BK 460 8 11 36(1), 40(2), 43(1), 44(1), 47 (1). 49(1). 57(1). 62(3)- 124 NEGATIVE EVALUATION (CERTAINTY) GROUP FEMALES Subject E No. words 2 2 # Spoken Types frequency SRS category (frequency) 54 BK 580 0 0 --- 55 LJ 605 12 28 1(1), 6(1), 7(1), 10(1), 12(1), 15(1), 20(1), 21(1), 40(3), 44(4), 45(2), 57(11). 56 LJ 95 1 1 40(1). 57 LJ 1305 2 2 1(1), 58(1). 58 LW 325 6 8 1(1), 12(2), 35(1), 37.5(2), 57(1), 62(1). 59 LW 350 5 12 1(1), 14(1), 37.5(1), 40(6), 57(3). - 6O LW 1190 3 4 40(1), 44(1), 57(2). MALES 61 JP 230 3 4 33(2), 34(1), 40(1). 62 JP 375 4 5 1(1), 14(1), 44(2), 67(1). 63 JP 95 3 4 12(1), 40(1), 57(2). 64 BK 80 1 2 40(2). 65 BK 2080 2 6 1(2), 40(4). 66 BK 935 4 4 5(1), 40(1), 52(1), 54(1). 67 LJ 540 7 l7 37.5(2), 40(6), 41(2), 44(1), 53(2), 57(3), 68(1). 68 LJ 480 3 3 40(1), 44(1), 57(1). 69 LJ 1445 2 2 1(1), 56(1). 70 LW 220 3 4 1(1), 40(2), 41(1). 71 LW 290 l 1 1(1). 72 LW 225 4 9 1(1), 40(4), 57)3), 58(1). APPENDIX F APPENDIX F LITERATURE REVIEW It has recently been asserted that "we seem to be marking time with respect to Specifying the etiology of neurotic phenomena: (Cohen, 1974, p. 476). The difficulty, Cohen suggests, is due to the choice of group comparison as the method in most past studies as psychopathology. So, for example, the fact that adult neurotics Show more autonomic nervous system reactivity than normals (Malmo & Shagass, 1952) does not give conclusive evidence of physiological factors in etiology. From corre- lational studies (group comparison) we cannot tell whether significant differences are etiological or descriptive characteristics. "Testing an essentially outcome process (deviant behavior) by treating it as an independent variable (as in group comparison studies) usually limits potential conclusions about etiology" (Cohen, 1974, p. 477). One solution to the problem is to conceptualize abnormal behavior as the dependent variable. We need, then, more studies which "isolate theo- retically or empirically derived variables which are then used to elicit or eliminate the target behavior" (Cohen, 1974, p. 477). Because the present study hypothesized that certain empirically and theoretically derived variables (positive experimenter evaluation, negative experi- menter evaluation, and no evaluation) would lead to differing amounts of anxiety manifestations, this literature review will focus on the methodology of experimental psychopathology and on those fields of experimental study which have conceptualized psychopathology as a dependent variable. 125 126 Methodology and issues of validity Shagass (1972) has made the point that experimentally in- duced psychopathology in humans must necessarily (or at least, hope- fully) be of shorter duration than naturally occurring symptoms are less intense, though not all researchers would agree with this obser- vation. Shagass cautions that since psychopathology caused by experimental manipulation may be qualitatively different than the psychopathology seen by clinicians, interpretation of such studies must be conservative. (For an opposing point of view, see Reyher, 1962). Shagass himself, however, investigates psychopathology experimentally: what is to stop a less sympathetic reviewer from dismissing the whole field on the grounds of lack of validity? And what characteristics in a research design in this area will do most to allay such criticism? The advantages of manipulative (experimental) methods over correlative approaches have been stressed by many writers (Plutchik, 1968; Spence 1956; Woodworth, 1938). The experimental approach is often considered the more useful of the two because conditions impor- tant to the outcome can be controlled and, therefore, replicated, and because, "in a good experiment, it is possible to say that the condi- tions manipulated by the experimenter caused the reactions obtained" (Plutchik, 1968, p. 25). The correlational method is most appropriate when the experimenter cannot control the relevant conditions and in the past has been described in introductory texts as the method of choice in such "applied" research as clinical, industrial, and educational studies (Plutchik, 1968). Spence (1956) has characterized correlative studies as "only one small segment of the total framework of a science 127 of behavior, and unfortunately not a very basic one at that (p.9)." Rychlak (1968) feels that the attitude of Spence and others has been destructive, and goes on to show that, at least theoretically, one cannot separate stimulus from response. Response is a concept which derives its meaning from the concept of stimulus, and vice versa. Therefore, the distinction between manipulative and correlational studies is an artificial one. In the search for empirical evidence for the etiology of psychopathology, however, the writer agrees with Cohen (1974) that the experimental method is a discriminable technique and is the more use- ful of the two. We must begin to separate stimulus from response in this field; otherwise, we concede that the concept of etiology has no meaning. On the other hand, as Rychlak has said, Spence's dismissal of correlative research is too cursory. Even in Spence's own field, learning, this point is beginning to hit home: learning theorists are now being faced with an accumulating body of evidence that different species exhibit various biological constraints on type of learning and capacity to learn, constraints which must be taken into account in any attempt to explain the process by which a particular species learns (Hinde & Stevenson-Hine, 1973; Shettleworth, 1972; Breland & Breland, 1961). The issue, for the present purpose, might be termed one of construct validity: just because an animal appears to learn in one way in the laboratory does not mean that it learns in the same way or for the same reasons in the field. Descriptions of field learning are badly needed, and these observational studies will be essentially correlative in nature. Analogously, laboratory studies in which abnormal 128 behavior or experience is used as the dependent variable are always open to the criticism that the behavior and experience in question are not caused that wgy under naturally occurring conditions. Stated in another way, this is the problem of mistaking the phenotype of behavior for its genotype. Group comparisons, at least, can be carried out with the assurance that the variables being measured have been caused by natural factors. (Furthermore, the more good descriptive material is available, the kind of material yielded by group comparison studies, the more likely it is that informed hypotheses about etiology will be made.) This validity problem poses a paradox: we cannot be sure our experimental manipulations are valid ones until we know the causes of abnormal behavior, and we cannot be certain what the causes of such behavior are until we demonstrate them experimentally. There seems to be no complete solution to this problem, but Cohen (1974), quoted above, has given a partial one. At this point in the discussion, his remarks gain meaning: We need studies which "isolate theoretically or empirically derived variables which are used to elicit or eliminate the target behavior" (p.477). That is, the best defense of a study in experimental psychopathology is that the experimenter, previous to the experimental manipulations, formed clearly articulated hypotheses based on thoughtful consideration of theory and data, and that this consideration took into account the problems of generalizing to the experience of individuals who exhibit the same behaviors in natural settings. It will be shown below that many studies in experimental psychopathology have been carried out without sufficient regard of these guidelines. 129 If abnormal behavior is to be the dependent variable, the ideal experimental design would have two attributes: l) Within-subject controls. When clinical work is successful, it whows how complex are the determinants of a symptom, and how different may be the dynamics (or learning histories) of two people whose symptoms appear similar. Therefore, if psychopathology is to be the outcome measure, it is most prudent to make the underlying dynamics and learning histories in the control and experimental groups as similar as possible. This argument can be made, of course, for any research, but it is particularly im- portant in studies which purport to explore defensive reactions to threatening external stimuli. Unforthunately, reasonable departures from this suggestion are more the rule than the exception. Unless the experimenter is interested in only one lev el of the independent vari- able, for instance, a within-subject design is impossible for studies in which the variable to be manipulated is a subject characteristic. Another potential problem is that the counterbalancing procedures required of a within-subject design may cloud the results by the addi- tion of another variable, the effect of orders of presentation of the independent variable. Counterbalancing may actually make the results more difficult to interpret, particularly if the independent variables to be counterbalanced are closely related to one another and have intense affective meaning to the subjects. By presenting a strongly affect-provoking stimulus before a less strong one, for instance, the affect elicited by the weaker might be artificially inflated or de- flated. Even though this effect could be partialled out, the mean differences between treatments might be decreaSed. 2) Several levels 130 of the independent variable. The independent variable may turn out to have the hypothesized effect, but with only the use of only one level, the possibility cannot be excluded that the variable had this effect for non-hypothesized reasons. This confusion, for instance, is charac- teristic of most of the sensory deprivation literature (Kubzansky & Leiderman, 1961); researchers in this area have not known enough about the effective factors contained within their rather global independent. variable to vary it systematically. Again, this recommendation applies to any research, but is a most important design attribute for studies in which there exists doubt about the validity of the experimental manipulations. Review of research in experimental psychopgthology Studies were selected for inclusion in this section on the basis of the following criteria: experimental psychopathology refers to those investigations in which the pathological behavior being studied is induced experimentally in normal subjects, and in which the focus of attention is not a particular symptom, but a broad array. Effec- tively, this means that the studies to be reviewed were concerned with the symptom-producing properties of certain stressors rather than with the discovery of the stressors which are necessary and sufficient to produce a particular class of abnormal behavior. The range of studies was further limited by restricting the review to areas of study which have generated a relatively cohesive line of research. Although the animal literature is touched on, it is too massive to be completely reviewed here. Emphasis is placed on human experimental psychopathology. 131 Animal neurosis Excitement about the possibility of observing the process of neurotogenesis in animals began in Pavlov's classical conditioning laboratory. Quite accidentally, one worker (Yerofeeva), in attempting to show conditioning under novel conditions, produced what Pavlov felt were neurotic responses (pathological responses to subsequent spatial generalization testing) in the canine subjects. The paradigm paired food presentation with electric shock, and Pavlov interpreted the dogs' behavior as the result of simultaneous arousal of the mutually inhibi- tory sympathetic and parasympathetic nervous systems (Kimmel, 1973; Davis, 1972). The second and more well known study, also carried out with dogs, was a discrimination task in which food preSentation was repeatedly associated with a circular stimulus and never associated with an elliptical one. When the salivary response to the circle was well established, the ellipse was slowly changed, in small increments over a large number of trials, into the shape of a circle. The dogs' responses were extreme motor excitation, aggression and destructive behavior, and loss of the previously learned circle-ellipse discrimina- tion. When brought into the experimental room, contrary to their previous behavior, the animals barked incessantly. When the restrain- ing harness was put on them, they howled, tore at the harness, and struggled. Most of these responses became chronic, and Pavlov concludes that they represented all the symptoms of an acute neurosis (Pavlov, 1941). The conclusion seems questionable, although there appears to be little doubt that the animals found the experimental conditions aversive. Later researches in Pavlov's laboratory produced other long- 132 standing symptoms in dogs: disturbances in cardiovascular and respira- tory functions, apathy, dullness, stereotyped movements, disturbance of social behavior, and irregularities in excretory and genital func- tions (Davis, 1972). Gantt (1944, 1973), a follower of Pavlov, has performed a long series of experiments demonstrating the development of "neurotic" reactions in the dog. The doubtful validity of these studies is best expressed by Gantt's own description of his measures: "If we have known the particular dog previously, a deviation from its usual be- havior would signify a disturbance" (1973, p. 37-38). The canine equivalent of depression is signified by the dog's retreating to a corner, tucking its tail, unresponsive- ness, or refusal of food in the presence of a human being. What corresponds to anxiety in the human being is shown in the dog by exaggerated muscular activity (daily running). inability to remain quiet, increased respiratory rate, and tachycardia. This over- activity leads to what is comparable to a manic state; extreme restlessness and markedly exaggerated autonomic symptoms (cardiac, respiratory, genito-urinary) (Gantt, 1973, p. 38). Gantt has devised two principles to account for neurotogenesis. Schizokinesis describes the fact that autonomic reactions to stress, particularly cardiovascular reactions, condition previous to secretory and motor reactions and extinguish much more slowly. The original cardiovascular disturbance can persevere long after the dog appears to react again normally to the conditioned stimulus. Autokinesis is a term coined to refer to the finding that a dog may develop other symptoms than those with which it first reacted to the original stress. For instance, one dog, Nick, who was studied intensively from the age of two until his death at 16, developed new symptoms in the genital, cardiovascular, and respiratory systems. Gantt does not provide any 133 evidence, though, to contradict the objection that the dog might have developed these reactions in the absence of unusual stress. Gantt's two principles are interesting observations, but the remainder of his work suffers from the absence of control subjects and the anthropo- morphic interpretations he makes of canine behavior. Liddell (1960), also a follower of Pavlov, showed that, in goats, the effects of stressful conditioning procedures were accentu- ated if the kids had been separated from their mothers for a short period immediately after birth. Experimental animals were also later found to be poor mothers. Rats trained to jump to one of two patterns, behind one of which is food, often develop seizures and "psychopatho- logical" behavior when one pattern is removed, giving the animal no choice, and an air blast is used to force the rat from the jumping stand (Maier & Feldman, 1944). Harlow (1964; Harlow & Harlow, 1971) has demonstrated in rhesus monkeys the severe effects of early social deprivation on later social behavior, particularly on later sexual and maternal behavior. Isolates also showed self-destructive and "autistic” behavior. Well known "animal neuroses" have been produced in cats by Masserman (1946, 1973), and supposedly encompass such diverse symptoms as compulsive behavior, claustrophobia, and anorexia nervosa. Masser- man used various conditioning techniques (see below) all of which employed experimental conflict as their essential ingredient, and added successful "therapies" for the disorders he caused. Others (Karn, 1943; Dworkin, 1939) have also claimed to have shown experimental neurosis in cats. Dmitruk (1972) cites studies which purport to have produced neurotic reactions in sheep, pigs, and--to bring home the absurdity of the situation-~ants. 134 The methods used in the production of animal neurosis have been, for the most part, techniques derived from learning theory: the learning of new responses incompatible with innate behaviors, rapid transition from one conditioned stimulus to another (particularly if the two stimuli elicit mutually antagonistic conditioned responses), delay of reinforcement, presentation of highly aversive stimuli, and presentation of novel stimuli (Cook, 1939). The area has not lacked critics. Dmitruk (1972) observed that most studies of animal neurosis have defined neurotic behavior as the behavior shown by an animal which has been through conditions hypothesized to produce neurosis. Finger (1945) criticized the re- search on the grounds that observations of psychopathology based on human criteria were not appropriate to the study of animal behavior. Nowrer (1950) went as far as accusing the investigators in this field of "self-mysticism" (p. 510). The most thoroughly investigated paradigm in animal neurosis has been Masserman's (1943, 1946). He presented aversive stimulation (shock) to one group of cats while they were feeding (experimental group) and to another group while they were not feeding (control group). The experimental animals were said to develop neurotic symptoms, while control animals did not, and Masserman maintained that the paradigm supported the argument that conflict was essential to the etiology of neurosis. Wolpe (1952), however, was not satisfied that Masserman had presented adequate data from his control animals to support the con- clusion that the difference between the two groups was significant, and he attempted a disproof of the hypothesis that conflict was necessary 135 to neurotogenesis. Wolpe's version of the study was much like Masser- man's, but, as the next investigator to use the paradigm (Smart, 1965) pointed out, Wolpe failed to equate the number of shocks given to the control and experimental animals. The results obtained were indeed in disagreement with Masserman's, but the experiment was contaminated. The next replication of the study (Smart, 1965) was also methodologi— cally inadequate. It remained for Dmitruk (1972, 1974) to settle the question. In his carefully controlled study, cats which received shock while eating did not display any more of the behaviors Masserman called neurotic than did animals receiving the same number of shocks not con- tingent on the presence of food. Dmitruk concludes that Masserman's "neurotic behavior" was an artifact, a consequence of being confined in the experimental apparatus. The animal neuroisis studies are perhaps the best examples of investigations in experimental psychopathology which have not ade- quately considered the problem of generalization. It is not neces- sarily true, as one commentator on the field has asserted, that "once this difference is complexity between animals and men is recognized, we can turn it to advantage, for the greater Simplicity of the animal case makes it more accessible to analysis" (Hinde, 1962, p. 241). It seems more probable that differences between the abnormal behavior of animals and men are qualitative as well as quantitative. At the very least, the burden of proof of validity, particularly after Dmitruk's study, rests on those who have conducted animal neurosis research. 136 Psychopathology induced by drugs Scientific interest in the effects of drugs on various be- haviors and functions began with Kraepelin's work in Wundt's laboratory on the effects of caffeine and alcohol ingestion on mental functioning. Impetus was provided by the discovery that phenylketonuria, a disease which may lead to severe mental retardation, is caused by a matabolic dysfunction. Kraepelin, impressed by the similarities he saw between the psychoses and the states which can be produced by some drugs, hypothesized that psycholis was a result of biochemical changes in the body. These changes, he thought, must be matabolic, and the disposi- tion for dysfunction must be inherited. Psychosis, therefore, must be caused by mechanisms similar to those seen in drug-induced psycho- pathology, the only difference being that the toxic agent in psychosis had not yet been discovered (Davis, 1972). Much the same point of view was espoused by those researchers who viewed the effects of psychotomimetic drugs as "model psychoses" (Levine, 1969). It was argued that the psychotomimetics, particularly LSD (lysergic acid dyethylamide), produced reactions so akin to schizo- phrenia that such substances should be sought in the bloodstreams of psychotics (e.g., Smythies, 1962). A spate of studies on the "psycho- tomimetic" effects of these drugs appeared in the next few years. These studies will not be reviewed, though, since, apart from the logical inconsistency inherent in assuming that psychosis is caused by an agent which produces psychotic-like symptoms, closer clinical examination has established that LSD symptoms do not have direct rele- vance to schizophrenic psychopathology. LSD-induced hallucinations 137 are primarily visual, while schizophrenic hallucinations are typically auditory; many of the symptoms which early investigators claimed were shared by schizophrenics and people who had ingested LSD were actually not at all similar, and were spawned by imprecision in the language used to describe LSD effects; thought disorder is more severe in schizophrenia and is not recognized by the schizophrenic, whereas most people who take LSD recognize that the effects are due to the drug; and perceptual distortions are much more common in LSD subjects than in schizophrenics (Hollister, 1964, 1973; Levine, 1969). Sensory deprivation Many of the studies which have produced psychopathology have done so incidentally. The literature on sensory and perceptual iso- lation, for instance, is full of anecdotal accounts and descriptions of behavior and experience which appears to be symptomatic, but well organized data is available only for "hallucinations" (e.g., Freedman, Grunebaum, & Greenblatt, 1961), and there is even some doubt about whether sensory deprivation hallucinations are qualitatively the same as those experienced by schizophrenics (Ziskind & Augsburg, 1962). Other data have shown that experimenter and situational demand charac- teristics probably play a large role in the production of many typical sensory deprivation phenomena (0rne & Scheibe, 1964; Jackson & Pollard, 1962). Nevertheless, almost every sensory deprivation experiment has reported perceptual and motor changes, including bizarre changes in body image and loss of spatial orientation; cognitive changes, includ- ing difficulty in thinking, concentrating, or talking, and paranoid 138 fears with no basis in reality; and affective changes, including irri- tability, intense anxiety, and lability (Zubek, 1969; Ruff, 1966; Kubzansky & Leiderman, 1961; Freedman, §1_51,, 1961). Typical sensory deprivation procedures require the isolation of the subject in a soundproofed and unlit room, and often tactile and kinesthetic stimulation is diminished by immersion in water or restraint, and by tubular cardboard cuffs placed over the subject's hands and lower arms. "Perceptual monotony,“ achieved by placing over the sub- ject's eyes halved ping-pong balls so that the visual field is a Ganzfeld, has also been effective in producing sensory deprivation effects. Experiments last anywhere from a few hours to ten days or more, and subjects are free to terminate at any time. During the experiment, the subject is usually asked to report his experiences, and the remarks are noted or tape recorded. Prior and subsequent to the experiment, perceptual, motor, and cognitive measures are usually collected. Although most of the work in this field is not directly con- cerned with subject variables, there are some notable exceptions. Goldberger and Holt (1961) showed that maturity in the handling of the primary process on the Rorshach predicts how well a subject will toler- ate later sensory deprivation. Cohen. Silverman, and Schmamovian (1962) used Witkin's rod and frame test to differentiate subjects who depended most heavily on external visual cues from those who depended most heavily on internal proprioceptive ones. Subjects who depended most heavily on external stimulation showed during sensory deprivation more intense psychological discomfort, a higher incidence of visual IN 139 and auditory "imagery," more disorganization of thought, greater dis- comfort with the bodily sensations they felt, higher levels of central nervous system activation, and decreased ability to discriminate soma- tosensory cues. Tranel (1962) found that extraverts tolerated isolation better than introverts. One study found that two schizoid personalities appeared comfortable during isolation (Cohen, Silverman, Bressler, & Schmamovian, 1961), while another found that more anxious neurotic persons tended to terminate isolation earlier than others (Smith & Lewty, 1959). Freedman, 55551, (1961) also found that schizoid per- sonalities tolerated isolation well, but added that when reality testing is impaired and acting out is the basic defense, tolerance is lessened. Ruff and Levy (1959), using a longer procedure, reported that passive dependent people or those with the fewest signs of conflict tolerated isolation best, while schizoid personalities had difficulty. Explanations of sensory deprivation effects have come from a variety of viewpoints. Lindsley (1961) hypothesizes a physiological basis. The ascending reticular activating system, he says, is a barometer of sorts, regulating incoming sensory stimulation and adapt- ing to it. This adaptation level is projected to the cortex, where it influences learning, emotion, perception, and cognition. The regulat- ing system is disturbed by sudden and wide variation in sensory input (sensory deprivation and overload), and behavioral results. Schultz (1965) takes this idea a step further and posits a drive for sensory variation. The drive has the goal of an optimal level of sensory stimulation, or sensoristasis. Schultz deduces seven hypotheses from 140 his theory, all of which are born out by the sensory deprivation data. Unfortunately, the hypotheses were made after the data used to validate them had been collected. A difficulty for explanations based on physiology or some other constant factor (like drive) is that sensory deprivation does not have equally strong effects if subjects are exposed to it a second or third time (Zubek, Aftanas, Hasek, Sansom, Schludermann, Wilgosh, and Winocur, 1962; Ruff, Levy, and Thaler, 1961; Lilly, 1956). The only study to use objective measures of performance found that a second exposure resulted in only one third the effects that the first one had (Pollard, Uhr, and Jackson, 1963). In trying to incorporate these findings into his theory of sensoristasis, Schultz (1965) suggests that perhaps a "psychological" adaptation occurs without a "physiologi- cal" one. Another group of theories comes from those investigators interested in the cognitive and perceptual effects of sensory depriva- tion. Bruner (1961) explains the effects of sensory deprivation on adult humans as the breakdown in maintenance of models of the environ- ment used by the organism to carry on its essential environmental transactions. "Sensory deprivation in normal adults disrupts the vital evaluation processes by which one constantly monitors and cor- rects the models and strategies one has learned to employ in dealing with the environment" (Bruner, 1961, p. 207). Freedman, Grunebaum and Greenblatt (1961) explain the phenomena similarly. They suggest that the properly functioning organism constantly seeks to impose structure on the environment according to a learned set of relationships which 141 have proved useful in the past. It is the person's attempt to find recognizable order in the non-order of sensory deprivation conditions that is said to be the cause of perceptual changes and distortions. A final group of theories has been offered by psychoanalyti— cally oriented writers. The best known and most influential of these is Rapaport's (1958) extension of the concept of ego autonomy. Rapaport contends that ego is suspended between the environment and the id, held in balance like a bit of metal set between two equally strong magnets. If either force becomes stronger, the ego loses some of its autonomy. Sensory deprivation, by decreasing the power of the external environ- ment, increases the power of the id, and primary process begins to overtake the ego. Thus anxiety is mobilized and psychopathology caused. The secondary process requires for its maintenance continual commerce with the external environment. Goldberger and Holt (1961) provided empirical support for this idea with their demonstration that individual differences in reaction to sensory deprivation can be pre- dicted by the degree to which an individual handles maturely the primary process during a previous Rorshach administration. Analysts of the methodologies of sensory deprivation experi- ments (Ruff, Levy, & Thaler, 1961; Cohen, Silverman, Bressler, & Schmamovian, 1961) have suggested that future research be designed to differentially assess the effects of the experimenter, the subject's personality structure, and variations in the experimental situation. Implicit in these critiques are suggestions that the perceived inter- personal world of the subject is an important variable. For instance, beyond the deprivation of the senses, what is the effect of the absence 142 of other people in experimental chamber? Is psychophysiological record- ing during the experiment construed by the subject as "communication" between he and the experimenter? What is the effect of being constantly observed without having the opportunity to see the observers? Do extraneous noises from outside the experimental room, an inevitable occurrence, help the subject maintain a feeling that he is not isolated from the external world? Unfortunately, the questions and theories generated to explore and explain sensory deprivation effects still await empirical test: interest in sensory deprivation has waned to the point that a reference to this body of literature is only seldom seen in the literature today. Sleep deprivation The Korean War provided the impetus for the study of both sensory and sleep deprivation, since the two phenomena appeared to be important components of the "brainwashing" technique used at that time by the Koreans and Chinese. Additionally, loss of sleep has traditionally been implicated in the onset of acute combat psychosis (Luby & Gottlieb, 1966). However, while interest in the functions and mechanisms of sleep has increased since the Korean conflict, the flurry of research concerning the psychopathology of Sleep deprivation has diminished. Sleep deprivation allows the observation of the development of a psychosis telescoped in time. The gross psychotic manifestations usually require 100 or more hours to make their first appearance (Murray, 1965). Up to that point, subjects show increasing irritability, 143 cognitive disorganization ( a general slowing of thought processes, a large number of spoonerisms, incoherent speech, and "dream thoughts interspersed with secondary process thinking as though there were a failure in repression" (Luby & Gottlieb, 1966, p. 409), and difficulty with new learning (Luby & Gottlieb, 1966). After about 100 hours, many subjects lose the ability to recognize the discrepancy between intrusive and internal stimuli and the actual external world; delusions and visual hallucinations then appear. Auditory hallucinations are more rare. For example, one subject (Luby & Gottlieb, 1966) saw smoke filtering from under a door and found the sight humorous, because he knew it was unreal-~the next day, upon seeing the smoke again, he felt compelled to make sure there was not a fire in the next room. Deper- sonalization is common, as are "Skullcap" headaches in the occipital region. The headaches often turn into “hat delusions" beyond 100 hours. Full-blown paranoia has been frequently observed. Tyler (1955), in one of the early studies, worked with sleep--deprived Marines. Just before dawn on the third day, one of the men left the line of march (the men were forced to march to stay awake), began to talk to himself, started a fight with the man in front of him, insisted he was carrying out a secret spy mission for the president, and finally broke ranks and tried to escape. Time distortion is also common, with time esti- mates being of consistent shorter duration than reality (Morris & Singer, 1961). Regulation of affect is diminished, and loud, inap- propriate laughter, as well as hostility, is observed (Luby & Gottlieb, 1966). 144 Overlapping speculative explanations of these "psychoses" have been offered by biologically and psychoanalytically oriented in- vestigators. Bliss, Clark, and West (1959) suggest that some acute schizophrenic episodes develop because of the social isolation and insomnia experienced by some predisposed individuals. The implication is that metabolic changes occurring during prolonged sleeplessness are conducive to a schizophrenic reaction. This hypothesis, though, seems not to have been explored further. A related and more detailed psycho- analytic synthesis of the sleep and REM sleep (the stage of sleep characterized by rapid bursts of lateral eye movements) has been offered by Dement and Fisher (Dement, 1960; Fisher & Dement, 1963; Dement, 1966; Dement, Greenberg, & Klein, 1966; Fisher, 1970; Dement e§_51,, 1970). To begin with, Freud, in The Interpretation of Dreams (1968, p. 617-618, 110-111) made the hypothesis that persons deprived of dreaming would become mentally ill. This "derangement" would result because the dream, no longer allowed to occur, could not fulfill its function of drive discharge; the dream-deprived individual would then be overwhelmed by instinctual urges during the waking state. When it was discovered that dreams were confined for the most part to the REM stage of sleep (Aserinsky & Kleitman, 1955), (insert below) it could be said that dreaming had its own physiological substrate. Insert: and that REM sleep is a different form of sleep, set off and regulated by a specific neurophysiological mechanism (Jouvet, 1967). Since many observers, including Freud, Jung, and Hughlings Jackson had noted the similarity of dreams and psychotic experience, the discovery of REM sleep paved the way for a theory linking the two phenomena on 145 neurophysiological and psychoanalytic grounds. Further, and most ger- mane to the present discussion, normal subjects specifically deprived of REM sleep (by awakenings whenever the REM EEG pattern appeared) showed behavioral disturbance similar (though less intense) to the symptoms manifested by totally sleep-deprived subject, while control subjects awakened the same number of times during non-REM sleep did not (Dement, 1960; Clemes & Dement, 1967). The most dramatic finding, one which has withstood dozens of replications, is that both total sleep—deprived and REM sleep-deprived subjects show a sharp rise in the percentage of REM sleep (dreaming) when the deprivation period ends(Dement, 1960). It appeared, then, that there existed a biological need for the type of sleep during which dreaming occurs, and that lack of such sleep produced psychopathology. Early attempts to integrate this material (Fisher & Dement, 1963; Dement, 1966; Dement, Greenberg & Klein, 1966) suggested that the prolonged sleeplessness character- istic of pre-psychotic periods in some individuals, by preventing REM sleep ahd its necessary biological function of drive discharge, forced the processes underlying REM sleep to operate during the waking state. A psychotic person, then, was one who literally dreamed while awake. Dement (1966) went further and postulated that REM sleep deprivation caused the build-up of a chemical compound in the nervous system which was the mechanism by which dreaming carried over into waking hours. Three developments have tempered these speculations, however. First, not all normal subjects develop a psychosis during sleep de- privation, not even after 264 hours (Luby & Gottlieb, 1966) or 15 to 16 days (Dement & Fisher, 1963). Also, this theory predicts that 146 actively psychotic individuals without sleep disturbances have abnor- mally high percentages of REM sleep time, a finding which is consistent neither across individual psychotic persons nor across studies (Dement, Cohen, Ferguson,& Zarcone, 1970). Furthermore, subsequent investiga- tions have not confirmed Dement's (1960) finding that REM deprivation causes more behavioral disturbance than non-REM deprivation (Snyder, 1963; Kales, Hoedemaker, & Jacobson, 1964). Dement has therefore renounced his "chemical build-up" hypothesis (Dement, eg_51., 1970). However, the poSsibility that psychotic people have a predisposition to REM Sleep deprivation is still possible. Indeed, Dement (Dement, ‘eg_51,, 1970) found that nine of nine actively psychotic individuals did not Show the normal increase in REM sleep after a period of REM deprivation, whereas psychotic people now "in remission" showed the characteristic REM rebound. Since Dement had almost never, in several hundred previous normal human and animal subjects, failed to observe the REM rebound following REM deprivation, he considered this finding of special significance. The linking of REM sleep with Freud's formulation that dreaming is a drive-discharge mechanism has been supported in various ways. Fisher ei_51,, (1965) found that 80-90.percent of REM periods are accompanied by full or partial erection, which suggests activation fo sexual impulses. Furthermore, while dreams with erotic manifest or latent content are associated with erection, anxiety dreams cause detumescence of ongoing erections. Pivik and Foulkes (1966) demon- strated that dream content is intensified following REM deprivation, while subjects awakened during periods other than REM sleep did not 147 show this phenomenon. The authorx measured intensity by means of an eight point scale which ranges from a rating of mental activity upon awakening as "mind blank" to "perceptual, hallucinatory, bizarre." This study is particularly important, since it reaffirms Dement's original contention that REM deprivation does have an effect. Its methodology is also a significant advance, because the experiment focused on dreams themselves, using REMs only as a gauge of their oc- currence, rather than on the biological process which accompanies dreams. The latter methodology, which was used by Dement (1960), is open to the interpretation that deprivation effects are not due to the inhibition of dreaming, which is a nonessential by product, but to the inhibition of a certain physiological stage of sleep with unknown func- tion. Pivik and Foulkes conclude that their results are in accordance with Freud's view of the dream as a safety-valve, and that there is still validity to the concept of "dream deprivation." In summary, the discovery of REM sleep and its characteris- tics has the potential of adding greatly to knowledge of psychopathology. Even though findings have not been totally consistent, some psycho- pathology and change in dream content evidently does result from REM (dream) deprivation. Perhaps the largest contribution from this area is the unique opportunity it has provided for the testing of psycho- analytic propositions of dreaming and the syntheses of biology and psychodynamic theory which have resulted. Hypnoticaliy inducedopsychopathology Research in this area falls into two categories (Reyher, 1962, 1967): 1) studies in which the behavior to be observed is produced 148 by direct suggestion (usually of affect); 2) studies in which the be- havior is produced by the induction of conflict. Direct suggestion Most investigations of this type involve suggesting to the hypnotized subject that he experience a particular affect or aspect of psychopathology (e.g., Levitt, den Breeijen, & Persky, 1960). The results are then observed. These studies are in general methodologi- cally weak, since they seldom employ control groups. But even more important, psychopathology produced in this way explains nothing about the natural etiology of symptoms--these investigations are little more than demonstrations that psychopathology can be suggested. This group of studies shares with chemical psychoses, sleep deprivation, and sensory deprivation the error of mistaking the phenotype of behavior for its genotype. The case for direct hypnotic suggestion in experi- mental psychopathology lacks even the saving grace of the minimal clinical relevance which can be claimed by sleep deprivation (involve- ment in disturbances due to "brainwashing,“ combat psychosis, and other possible sleepless conditions (Luby & Gottlieb, 1966) and sensory deprivation ("brainwashing," solitary confinement in prisons, anecdotal accounts of sensory deprivation effects while lost at sea or disori- ented in the visual Ganzfeld of the arctic, psychotic-like states developed by polio patients confined in tank-type respirators, and mental changes in temporarily blind cataract patients [Ruff, 1966]). That is, "Direct suggestion tests nothing but itself" (Reyher, 1962, p. 346). 149 Nevertheless, if the purpose is not to explore the natural etiology of symptoms, direct suggestion can be used to advantage. Wittkower's (1935) studies present concrete evidence of the profound bodily effects of various psychological states. After taking pre- hypnotic measures of his dependent variables, Wittkower suggested dif- ferent emotional states to his subjects and recorded the dependent measures again. He was able to cite X-ray evidence of increase or decrease in heart size (depending on the subject), changes in quantity and chemical content of saliva, increases or decreases in gastric motility and acidity (again dependent on the characteristic reactions of the particular subject), changes in amount and chemical content of bile, changes in leukocyte counts, serum changes in calcium, potas- sium, chloride, and iodine, and changes in urine and galvanic skin response (GSR). While Wittkower's work can be criticized for lacking control subjects who were not hypnotized, his purpose was not to in- vestigate hypnosis, but simply to "instruct" his subjects in the most effective manner. Again, Wittkower's intention was to demonstrate soma-psyche relationships, not to understand them, which is the reason the direct suggestion approach was appropriate. His methods have been adopted and expanded by Dudley, Holmes, Martin, and Ripley (1964). Kehoe and Ironside (1963), and Persky (1962). The direct suggestion literature, not systematically reviewed here, is more adequately summarized by Deckert and West (1963) and Weitzenhoffer (1953). 150 Induction of conflict The history of this line of inquiry began with the investi- gations of A. R. Luria (1932). Under hypnosis, Luria's subjects were led to believe that they had committed reproachable acts. It was further suggested that although the memory of these false stories would not be retained (suggestion of a paramnesia), subjects would feel quite disturbed by them posthypnotically. (After the session, of course, the experimental conflict was removed.) Critical words from the paramnesia were placed in a list of control words not specifically releated to the story, and the complete list was presented to subjects as an association experiment. Subjects were pt press a tambour with their preferred hand whenever they gave a verbal response to one of the words on the list. Luria presented this taSk both before and after the implantation of the paramnesia. In addition to the voluntary pressure curves generated, verbal reaction time, nonvoluntary activity of the preferred hand, and respiration were recorded. Because the experimental words produced lengthened reaction times and irregular hand pressures and respiration, Luria concluded that he had demonstrated signs of conflict. Since he used no controls, standardized procedures, or statistical analysis, the value of Luria's work lies in the demon- stration that hypnotic methods are useful in the laboratory study of psychopathology. Huston, Shakow, and Erickson (1934) repeated one of Luria's experiments with 12 normal subjects. Nine "appeared" to accept the paramnesias, and Luria's motor tests revealed signs of conflict in Six. Huston, et al., noticed that under hypnosis these six subjects 151 tended to give verbal responses directly related to the conflict and showed few nonverbal disturbances, while the reverse was true in the waking state. The authors suggested that if the "excitation" aroused ‘ ‘ by the conflict is not discharged verbally, disturbances of the volun- tary and nonvoluntary motor systems appear. Their speculations are consonant with the traditional psychoanalytic view that unconscious (i.e., nonverbalizable) material is potentially symptom-producing until it becomes conscious (i.e., until it can be verbalized), but confirmation of this intriguing notion must await more stringent test. Bobbitt (1958), also using Luria's techniques and measures, hypothesized that disturbances would increase as repression weakened. To mimic repression, she gave her six subjects total amnesia for the paramnesia, partial amnesia, or full awareness. She tested pre- hypnotically, while the paramnesia was operative, and again after the removal of the paramnesia. To her surprise, she found the greatest degree of disturbance in the partial awareness group. Her findings, though, are not unexpected from at least two points of view: 1) The most parsimonious interpretation is that subjects with full awareness realized the conflict was implanted--the paramnesia would then have lost its anxiety-inducing properties (Reyher, 1962). 2) Again, like the interpretation suggested for the results of Huston, 53555,, it is possible that full awareness did away with the pathogenicity of the conflict simply because of the properties of consciousness, while partial awareness, analogous to the situation in which ego functions fail and a partial breakdown in repression occurs ("the return of the repressed"), produced a more intense conflict because the impulse 152 (i.e., the memory of the acting out of the impulse) was trapped between awareness and unconsciousness. Total amnesia, like a successful re- pression, prevented the conflict from influencing conscious experience. Reyher (1962, 1967) proposed a paradigm by which the clini- cal relevance of hypnotically induced psychopathology could be deter- mined, and expressed concern that previous investigations had not met the paradigm's requirements: "In the absence of adequate criteria, the data of instrumental hypnosis research tend to be either rejected or consigned to the limbo of ambiguity" (Reyher, 1962, p. 344). The paradigm is composed of four points: 1) The hypnotically induced process must not contain clues about how the experimenter expects the subject to respond. Reyher bases this criterion on the work of 0rne (1959), who showed that hypnotic subjects are particularly susceptible to experimenter and situational demand characteristics. Since the research reviewed directly above (Luria, Huston, 22.21:» Bobbitt) used Luria's technique of suggesting to subjects that they will feel dis- turbed by the induced conflict, none of these studies meets the cri- terion. Furthermore, an ideal design, says Reyher, should activate the subjects' natural conflicts instead of inducing an artificial one and suggesting an amnesia for it. Artificial conflict, though, is still used by Reyher and his co-workers (see below) because of the opportunity it affords for standardization of the anxiety-provoking stimulus and because activation of natural conflicts in the laboratory does not produce as intense or as wide an array of symptoms (Reyher, 1962). 2) The induced process, of itself, must produce other "processes and behavior; that is, it must be response-producint" (Reyher, 1962, 153 p.345). The fact that direct suggestion studies specify in their instructions to subjects the behavior to be observed as the dependent variable leads Reyher to conclude that they lack relevance. 3) At least some of the responses produced by adherence to steps 1 and 2 must meet the criteria for inclusion into some classification of psycho- pathology. This requirement also is not met by the Luria-type studies, since hand pressure and nonvoluntary movements have an unknown rela- tionship to naturally occurring psychopathology. (It might be noted that at least one of the subjects of Huston, e;_51,, showed inability to sleep, constant headache, poor appetite, antagonism toward the hypnotist, and irrational behavior surrounding his smoking habit-- smoking was specifically implicated in the conflict implanted in this subject--for the 24 hours before the paramnesia was removed.) 4) Ade- quate control groups must be used. These are to be composed of subjects with hypnotic susceptibility equal to that of experimental subjects (Reyher, 1971). The controls, whose identity is not known to the ex- perimenter, receive the same experimental treatment and are instructed to simulate hypnosis by a co-experimenter (0rne, 1959). Again, far from using well-devised controls, the Luria-type studies used no control at all. Other than his own researches, Reyher cites a study by Wolberg (1947) as coming close to meeting the paradigm's criteria. Wolberg implanted a conflict about the consumption of someone else's chocolate bar; the candy was found by the subject beside the chair after the hypnosis was terminated. Wolberg instructed his subjects that, al- though they would forget the experimenter had suggested craving for the 154 chocolate and consequent feelings of guilt, they would "react" to the suggestion nevertheless. "The distinctive aspect of his instructions is the posthypnotic suggestion of an overwhelming impulse which is rendered anxiety-producing by pitting it againSt conscience" (Reyher, 1962, p. 347). Wolberg's important result was subjects' reports of posthypnotic psychopathology, both somatic and psychological, includ- ing dizziness, tachycardia, and negative hallucination. Although Wolberg neither used controls nor refrained from instructing his sub- jects how to perceive the impulse, the symptoms he produced were sufficiently intense and corresponded closely enough to naturally occurring psychopathology to lead Reyher to conclude that they may well have represented valid phenomena. Taking as his starting point the previous studies and his own paradign, Reyher and his co-workers have generated over the inter- vening years the largest body of empirical data in this field (Reyher, 1958, 1961, 1967; Perkins & Reyer, 1971; Moore, 1964; Rokeach, Reyher, & Wiseman, 1968; Sommerschield & Reyher, 1973; Burns & Reyher, in press). The method has involved the implantation of a sexual or ag- gressive paramnesia: the sexual or aggressive feelings are described as intense, the urges are not satisfied in the story before it comes to an end, and the subject is told, "Later you learned that the only way you could attain peace of mind was to completely push the whole experience into the back of your mind" (Sommerschield & Reyher, 1973; Burns & Reyher, in press). The subject is then told that he will re- experience the intense sexual or angry feelings upon psothypnotic cue (notice: the amnesia is specific for the events of the paramnesia, 155 not the affect), and that these feelings will be directed toward a person in the laboratory. The person toward whom the impulses are to be directed is said to be the individual in the paramnesia toward whom the subject felt such strong impulses. The posthypnotic suggestion also includes instructions that the subject, upon experiencing the intense feelings posthypnotically, will tell the experimenter how he (the subject) would like to express those feelings. Used as posthyp- notic cues have been dey words from the paramnesias presented tachis- toscopically in a list of control words (Reyher, 1958, 1961, 1967); the experimenter's question, "How are you doing?" (Burns & Reyher, in press), and the experimenter's verbalization of the numbers one, two, and three--in this case it was suggested that the feelings would in- crease in intensity with the increase in number (Sommerschield 8 Reyher, 1973). If, as Reyher predicted, the hypnotically induced sexual or aggressive feelings were highly anxiety-provoking (i.e., response- producing in their own right), then, Reyher further hypothesized, two related processes should spontaneously be set into motion: 1) When the posthypnotic cue is given, there should be a spontaneous inhibition (repression) of the previous hypnotic suggestion to direct the intense sexual or angry feelings toward the co-experimenter. 2) Because of the anxiety which motivated the repression, associated with the subject's reception of the posthypnotic cue should be spontaneous reports of psychopathology; and the symptoms should correspond in form to clini- cally observed manifestations. 156 These predictions have been confirmed in the studies cited above, and consistently, control subjects simulating hypnosis, compared to experimental subjects, have shown less spontaneous inhibition of the impulse and symptoms of significantly less intensity and frequency. In each study, symptoms have ranged over most of the spectrum of psy- chopathology seen in clinical practice. Thus, a spontaneous inhibition, or repression, has been demonstrated experimentally, and a link be- tween this inhibition and the presence of symptoms has been shown. However, the most important results for the field of experimental psy— chopathology have been those which suggest a mechanism by which fre- quency and type of symptom are caused. In the first studies (Reyher, 1958, 1961), Reyher reported a relationship between type of psychopathology and the degree to which a subject inhibited the expression of the hypnotically induced anger and associated destructive impulse. The categories of symptoms, labelled the symptomatic reaction scale, are reproduced below from Sommerschield and Reyher (1973). The order of the categories was de- termined by correlating each category with the degree of repression (represented by Reyher's self-devised repression index) and ranking the categories according to the degree of repression. 1. The first set of symptoms is characterized by the dominance of autonomic nervous system innervation such as nausea, gastric distress, headache, tiredness, sleepiness, tachycardia, pres- sure in head, sweating, flushing, skin disturbances, organ dysfunctions, heaviness, temperature alterations, and such feelings as "queasy" and “antsy.” 2. The second set of symptoms include those which are dominated by innervation of the somatic nervous system, such as stiff- ness, aches, pains, tension, tics, tremors, physical discom- fort, etc. 157 3. The third set includes disturbances of affect such as (a) flattening: lack of feeling, apathy, etc.; (b) superego re- actions: feelings of being alone, abandoned, ashamed, depressed, disgusted, guilty, worried, etc.; (c) inversion: definite feeling of well-being; and (d) alienation: feelings that seem weird, strange, odd, unreal, unnatural, foreign, etc. 4. The fourth type of symptoms include unspecified distress that cannot be categorized clearly as either physical or emotional in nature such as feeling upset, nervous, etc. 5. The fifth set of symptoms include anxiety, fear, apprehension, etc. 6. State of confusion, doubt, and disorientation which include statements that one's thoughts are being pushed or pulled and that the content of thought cannot be specified represent the next stage in the continuum. 7. Then there are dissociative reactions: (a) somatic and idea- tional delusions, such as limbs feeling detached, "crazy" thoughts, and intruding paranoid ideas; and (b) strong compul- sive urges not carried out in behavior, such as wanting to move hands around, scratch at something, etc. 8. Next comes disturbance or distortion in perception of the anxiety—producing posthypnotic cues (words). 9. Finally, symptoms include derivatives of the induced conflict. These are symbols of the induced experience and/or the repres- sive forces themselves. Memories of personal experiences that are congruent with, or similar to, the induced experience are also symptomatic. (Sommerschield & Reyher, 1973, p. 279). Corroborating evidence for this relationship between degree of repression and type of psychopathology has come from the results of Reyher (1967), Burns and Reyher (in press), from an analysis of data collected by Perkins (reported in Sommerschield & Reyher, 1973), and from the results of Sommerschield and Reyher (1973). In each of these investigations, the sequence of symptom categories reported by subjects who achieved progressively greater awareness of the anger and destruc- tive impulse correlated significantly with the sequence of categories in the symptomatic reaction scale. None of these studies found a 158 significant correlation for those subjects who did not achieve progres- sively greater awareness. In addition, in two independent samples, Reyher and Basch (1969), using self-report, found significant relation- ships of .43 and .57 between degree of repression and proportion of psychosomatic symptoms. This study, too, supports the order in which the categories of the symptomatic reaction scale are organized, since psychosomatic symptoms are said to accompany a high degree of repres- sion. There is evidence, then, that degree of repression is an influ- encing variable in the process by which a particular type of symptom is developed and expressed. Perkins and Reyher (1971) demonstrated that frequency of symptoms is related in a curvilinear manner to degree of repression. This finding means that, as repression weakens, not only are different types of symptoms manifested, but these types are experienced 555551 _w_i__t_h the types which have come before: there is something of a snow- ball effect. Then, with full awareness comes conflict resolution and the reduction of symptom frequency. It may be noted that this finding is consistent with the interpretations offered above for the results of Bobitt (1958) and Huston, eg_51,, (1934). The curvilinear relationship of frequency of symptoms and degree of repression has been partially replicated by Sommerschield and Reyher (1973), who found a significantly negative relationship between number of symptom types and degree of repression for a hostile paramnesia, but not for an oedipal paramnesia. Burns and Reyher (in press), in another independent study, again replicated the effect. 159 Reyher's own integration of his empirical findings follows: The successful replication of our earlier investigations reinforces the theoretical formulation presented at the beginning of this communication. The significant correlations between the empirically derived sequence of symptoms for individual subjects and the order of symptom categories on the symptomatic reaction scale show that psychosomatic symptoms are the first symptoms to appear as repres- sion begins to weaken, and as the drive or drive-related impulses progressively become more completely and blatantly represented in the subject's awareness, other types of psychopathology appear. These symptoms are psychological rather than psychosomatic and are produced by the drive in conjunction with the anxiety it produces. When the intensity of anxiety reaches a critical degree, one or more mechanisms (defenses) are activated to prevent further in- creases in intensity and the disorganization of adaptive behavior. These defenses express indirectly the drive through dissociative reactions, symbolically represent the drive by phobias, obsessions, and compulsive behavior, or externalize the drive (projection). However, the anxiety that is producing these defense and other re- sponses is still producing psychosomatic symptoms; and as the drive achieves an unknown degree of representation in awareness and in- tensity, it may produce directly a psychosomatic symptom, as in the case of gastric hyperactivity due to the frustration of oral in- corporative impulses (Alexander, 1950). The systematic principle being explicated here is that different degrees of representation of an anxiety-producing drive in an individual's awareness, via derivatives, confers different response-producing properties upon it (RéyRer, 1963). (Sommerschield & Reyher, 1973. pp. 287-288). There has been no lack of controversy surrounding Reyher's theory and methodology. As Reyher (1963) has said, his techniques are useful only for testing theories which postulate relationships between the phenotype of psychopathology (the symptom itself) and certain underlying processes, or genotypes. Levitt (1963) criticized this limitation as an acceptance, without empirical test, of psycho- analytic hypotheses of the etiology of psychopathology. More impor- tantly, the mysticism of Mesmer still seems to hand heavy over the image of hypnosis in the eyes of many investigators and, in addition, many psychoanalytically inclined researchers, who would be most likely to find Reyher's results intriguing, are still influenced by Freud's 160 categorical rejection of hypnosis in favor of free association. Many of these criticisms and attitudes, however, are obviated by Moses' (1974) recent finding that a slight modification of Reyher's method produces as much and as intense psychopathology in the waking state as it does in hypnosis. Moses (personal communication) has also collected data showing that control subjects who sit and do nothing for the dura- tion of time it took the experimenter to introduce the false story to the experimental subjects show fewer symptoms. A similar reduction in symptoms is shown by subjects who are read a "neutral" paramnesia. Stern and Reyher (1975), who requested subjects to describe for 30 minutes their verbal associations and visual imagery, used the sympto- matic reaction scale to score the subjects' verbalizations. They found that the reactions observed were so much less frequent and more restricted in range than those seen in Reyher's earlier investigations that the data were described as anxiety signs rather than psychopatho- logical sysmptoms. Thus, it appears that while hypnosis may not be essential to the production of intense laboratory psychopathology in normal subjects, anxiety-inducing paramnesias are. Of course, hypnosis and posthypnotic amnesia are still necessary if the purpose of the experiment is to document relationships between repression and psycho- pathology. Reyher's adoption of psychoanalytic concepts and terminology has grated on more behaviorally oriented researchers and has probably prevented his work from having greater impact. Specifically, his self-styled index of repression, an important dependent variable in his own investigations and those of his co-workers, is open to the 161 criticism that it is a concrete measure of a hypothetical process. Reyher (1967) has countered this argument by operationally defining repression as the opposite of verbalized awareness. Whenever a subject receives the posthypnotic cue for the experience of the intense angry or sexual feelings, he can report full awareness of the feelings, partial awareness (e.g., annoyance), or no awareness. The repression index is the sum of each subject's expression of full, partial, or no awareness, biased conservatively by giving greater weight to reports of full awareness than to partial awareness, and greater weight to partial than to no awareness. In this section, two types of hypnotically induced conflict have been reviewed. The first type, direct suggestion, was criticized for the lack of clinical relevance of its results, and because it ap- pears to have no potential for the uncovering of information regarding etiology. Direct suggestion can test only the hypothesis that it is effective in producing psychopathology. The second type, hypnotically induced conflict, was dealt with in more detail. The early work of Luria (1932) and his followers (Huston, e5_51,, 1934; Bobitt, 1958) were also criticized for lack of clinical relevance. Further, these investigators suggested to their subjects the effects of the implanted conflicts, while Reyher's (1962) paradigm requires that the suggested affects and impulses must be allowed to be response-producing in their own right before clinical relevance of results can be claimed. Reyher's experimental production of spontaneous impulse inhibition (repression) and his replicated findings that the type and frequency of psychopath- ology are related to the degree of repression were reviewed. 162 Conclusions In introducing this review, it was stated that much of the research in experimental psychopahtology had neglected to consider adequately the validity of generalization beyond the experimental sample and techniques. The field of animal neurosis is the most glar- ing example of this neglect--and yet, an experimenter can obviously use techniques with animals which cannot be used with human beings. This is, of course, the argument of those who are engaged in animal investigations. Today, though, comparative research attracts little interest in clinical circles. Nevertheless, the bulk of the compara- tive literature shows one important ingredient commonly missing from human investigations in this area: the use of theory to plan design and predict outcome. Animal experimenters are theory testers, and this facet of their work is what has kept the field alive since Pavlov's initial observations. In contrast, the area of sensory deprivation was almost wholly empirical. Probably the only theory of consequence to arise from it was Rapaport's (l9 ) contribution to the theory of ego auton- omy, and even this theory did not stimulate a further line of research. The problem (at least from the vantage point of experimental psycho- pathology) was that sensory deprivation was never convincingly related to any valid and relatively common factor in the etiology of psycho- pathology. If it had been, many theoretical frameworks could have used isolation techniques to test hypotheses, and furtherhypotheses would have been generated by the results of such studies. In retro- spect, the sensory deprivation literature seems to have died out when 163 those interested in it were able to explain to their satisfaction its effects. The dead-end future of the field was documented by the fact that 1eye1§_of the independent variable were almost never used: it is usually only theory which leads an investigator to discriminate the levels of a seemingly unidimensional variable. The fate of sleep deprivation would have been the same with- out the discovery of REMs and their relationship to dreams: early sleep deprivation studies, like much of the sensory deprivation litera- ture, were merely demonstrations of an interesting phenomenon. Now, though, with data linking REM deprivation and dream content (Pivik & Foulkes) and, perhaps, behavioral disturbance (Dement, 1960; Dement eg_51,, 1970), the field is abristle with psychoanalytic and psycho- physiological intergrators of theory and research. Chemical psychoses, once thought to hold promise for the understanding of the etiology of psychosis, have now been rejected by those who originally touted them (Hollister, 1964). Psychopathology produced by direct suggestion has also been criticized for lack of relevance (Reyher, 1962). Nothing is gained by the demonstration of phenomena we do not understand. What is needed is research in which naturally occur- ring etiology is specified. And these studies will probably consist of experimental tests of theories which were originally formulated in an attempt to explain the development of symptoms as they are seen by clinicians. An alternative expression is that the potential contri- bution of experimental psychopathology is the unravelling of geno- types. 164 There is little doubt that the studies of REM sleep and psychopathology meet these criteria, since every human being sleeps, shows nightly REMs, and dreams. However, except for the observation of Dement e;_51,, (1970) of lack of REM rebound in schizophrenics, this literature has not yet established clear connections between the phenomena under study and psychopathology as seem by clinicians. There have been attempts in this direction (Fisher & Dement, 1963; Dement, 1960), but these studies have been shown to be inconclusive. The comparative work, with its emphasis on the test of learning theo- ries of psychopathology, is also satisfying with respect to the in- vestigation of genotype, but it falls short of meeting the criterion of validity. The only paradigm which investigates genotypes, employs theory, strives to make its procedures correspond to the theory's description of naturally occurring etiology, and accepts as data only those manifestations which are also clinically observable, is the method of posthypnotic stimulation of hypnotically implanted paramnesias (Reyher, 1962) and the research which has been generated by this technique (e.g., Moses, 1974). Typed and Printed in the U.S.A. Professional Thesis Preparation Cliff and Paula Haughey 144 Maplewood Drive . ._" East Lansing, Michigan 48823 Telephone (517) 337-1 527 "I1111111111111111.1111s