RESEGNSE DIFFERENCES BETWEEN PROCESS AND REACTEVE SCHIZOPHRENiCS AS INDUCED BY MAGAZENE PHOTOGRAPHS Thai: for tho 009m :4 ”L D. MiCHIGAN STATE UNIVERSITY John Mark Reisman 1958 Q-~,‘ This is to certify that the thesis entitled Response Differences Between Process and Reactive Schizophrenics as Induced by Magazine Photographs presented by John M. Reisman has been accepted towards fulfillment of the requirements for Ph 'D ' degree in _P_S_y_gyCh°10 J] ; (T Méf’or professor 9/ Date March 3. 1958 0-169 LIBRARY Michigan State University 1/' —-\‘1 (1 T'W-I' “fl“ " 1" 1".- A )L“ - 7‘ lfi'l'f.’ ~V KT “n.9,,“ "'3'" "Ill“ .- 'f‘ .. RLDFON bk! J'l" ‘- “""“-*41' V--’ D 133-1 A. uJ—J..uvI l .LKJKJL‘JQQ 1i“. LJ Led-141V L4. 4.1 SOHILOEHALLICS AS IADUCED BY MAGAQINE thIOUuAfnS John Mark Reismen AN ABSTiAC submitted to the SchOol for Advanced Graduate Studies of Michigan State University ‘5 Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DQCTQi OF rs LOSOfHY Department of Psychology 1958 M ‘ 3/ flf/ John m. heisman ABSTRACT On the assumption that changes in motivation occur as symptoms persist over long periods of time, it was hypoth- esized that reactive schizophrenics would avoid while pro- cess schizophrenics would not avoid photographs considered to represent areas of frustration, conflict, and/or threat. To test this hypothesis, thirty-six male hospitalized veterans diagnosed as schizophrenic were classified as reac- tive and thirty-six as process. A comparison group consisted of thirty-six male veterans hospitalized for physical ail- ments and not considered psychotic. The three groups were controlled for age, IQ estimates, and length of hOSpitaliza- tion. The experimental task was to sort an ordinary deck of playing cards onto a board divided into quadrants for eleven trials. On trials 1 and 9 - ll, suercts were requested to sort fast. From trials 2 - 8, subjects were allowed to sort fast or slow under one of four conditions: FP subjects were told they would see pictures after each trial during which they sorted fast; 3? subjects were to see pictures after each trial during which they sorted slowly; FL subjects were to see a light turned on after each trial during which they sorted fast; SL subjects were to see a light after each trial during which they sorted slowly. Fast and slow were defined in re- lation to the subject's speed on the immediately preceeding - John M. Heisman trial. The pictures were magazine photographs, mainly of people, previously judged by five clinicians as to whether they represented an area of frustration, conflict, and/or threat. Pictures so judged were presented first, in sets of five, when appropriate. The light was a small flashlight bulb. It was predicted reactives would sort cards so as to avoid seeing pictures while process would not. In a sub-experiment, ten process and ten reactives who performed under the light conditions were used as subjects. A pile of 33 pictures was presented to each subject who was told to simply look at them for as long as he liked. Eleven of these pictures were photos of people judged to represent an area of frustration, conflict, and/or threat; eleven were photos of people judged non-threatening; eleven were photos of scenery or inanimate objects. It was predicted that pro- cess subjects would look at the three types of pictures for about the same length of time but that reactives would look at pictures judged threatening less than the other two types. Generally, the results supported the hypothesis. As was predicted, reactives who sorted under F? were slower than under any other condition. They saw significantly fewer pic- tures than process subjects and less than would have been ex- pected from their performance under the light conditions. Re- actives were significantly more variable under FF and SP tnan under FL or 3L. In the sub-eXperiment, reactives looked at photographs for a significantly shorter period of time than I John n. Reisman process subjects. As was expected, non-threatening and in- animate photos had a significant positive effect on the length of time that reactives loohed at pieturss ta; there was no differential effect with process subjects. It was also found that the sorting time differences between the groups, normals sorted faster than reactives who sorted_faster than process subjects, seemed to be more a reflection of differences in motivation than of a deficit in psychomotor ability in schiz~ ophrenia. The major conclusions with respect to the population employed were: 1. Schizophrenics may be fruitfully divided into process and reactives. 2. Reactive schizophrenics avoid pictures considered to represent areas of frustration, conflict, and/or threat while process schizophrenics do not. 3. Duration of symptoms is a crucial variable in the investigation and underscanding of schiZOpnrenia. 'hhSPONSfl DIFFEdBNCBS BLTNEEE PROCESS AND REACTIVE SCHIJOPHRENICS AS INDUCED BY MAGAAINE PHOTOGRAPHS BY John Mark Heisman A.THESIS submitted to the School for Advanced Graduate studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1958 II III IV VI TABLE OF CONTENTS INTRODUCTION Development of the Process-Reactive Concept Functions and Limitations of the Process-Reac- tive Concept Formulation of the Hypothesis PROCEDURE Main Experiment Sub-Experiment SELECTION OF PICTURES ARD LIGHT Main Experirent Sub-Experiment SELECTION OF SUBJECTS EXPERIXENTAL PREDICTIONS RESULTS Main Experiment Prediction 1 Prediction 2 Prediction 3 Prediction 4 Other Results Relevant to the Hypothesis Incidental Results Sub-Experiment Prediction 5 Prediction 6 Other; Results Relevant to the Hypothesis Page 15 21 21 24 26 26 28 29 37 39 39 39 41 43 43 45 51 54 54 55 55 VII VIII IX Incidental Results DISCUSSION Discussion of Results Suggestions for Future Research SUMMARY AND CONCLUSIONS BIBLIOGRAPHY APPENDIX 57 58 58 62 64 68 72 TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE TABLE 10 11 12 13 14 15 l6 l7 18 LIST CF TABLES Criteria For Selection of Process and Reactives Controlled Variables For The Twelve Groups Mean Sorting Time Per Trial During Trials 2 Through 8 Mean Number of Times Groups Saw Pictures or Light Pooled Mean Sorting Time Per Trial During Trials\2 Through 8 Mean Sorting Tire Per Trial During Trials 9-11 Pooled Mean Sorting Tirre Per Trial During Trials 9—11 Controlled Variables in the Sub-Experiment Seconds Spent Looking At the Three Types of Pictures Description of Experimental Pictures Analysis of Variance: Age Analysis of Variance: IQ Estimates Analysis of Variance: Length of Hospital- ization’ Analysis of Variance Trial 1 Analysis of Variance of Differences Between Time on Trial 1 and 2 Analysis Trials 2 of Variance Repeated Measures Through 8 Analysis Trial on of Variance Mean Sorting Time Per Trials 2—8 Analysis of Variance Square-Root Trans- formation of Mean Sorting Time Per Trial on Trials 2-8 Page 31 4O 41 44 48 50 53 54 57 73 76 76 76 77 77 78 79 79 TABLE 19 FIGURE 1 FIGURE 2 FIGURE 3 FIGURE 4 Analysis of Variance Repeated Measures Trials 9 Through 11 LIST OF Mean sorting times Mean sorting times Mean sorting times condition Mean sorting times FIGURES of the reactive group of the process group under the Fast-Picture of the normal group 80 46 47 52 ACKNOWLEDGEHENTS The author wishes to express his deep appreciation to the members of his committee: Dr. M. Ray Denny, who served as chairman, Dr. Albert I. Rabin, Dr. Milton Rokeach, and Dr. Paul Bakan. He only wishes it were possible to adequately acknowledge each of their rany suggestions which not only improved the quality of the experiment but the ability of the experimenter as well. He is greatly indebted to Dr. Rabin for his help in the design of the experiment, to Dr. Rokeach for giving the thesis a new perspective, to Dr. Bakan fer the design of the sub-experiment, and, especially to Dr. Denny for his constant guidance, encouragement, and interest which made more than this thesis possible. He wishes to thank Dr. Stewart Arnitage and his staff at the V. A. Hospital in Battle Creek and Dr. John Brownfain, chief clinical psychologist at the V. A. Hospital in Dearborn, for their assistance in obtaining subjects for this experi- ment. The author dedicates this thesis to his wife, Margo, his daughter, Hope, and her future siblings whose contributions have been measurable and all of whom have patiently waited for its completion. INTRODUCTION A. Development of the Process - Reactive Concept One of the stimulating facets of psychological research ... perhaps of all scientific research ...is that each new bit of knowledge welds into a key that unlocks more distant horizons of ignorance. Perhaps no other area better illustrates that advances may at once illuminate and perplex than the study of schizophrenia. While it has become the object of ever-increa- sing scientific scrutiny, it has also become the source of an alarming body of equivocal and contradictory experimental evidence, and the subject of a vast number of free-floating theories. It was only about a hundred years ago, in 1860, that Morel described a form of stupidite associated with progressive mental and emotional deterioration by the term demense precoce. He referred to a phenomenon in which young adults who as children seemed intelligent and teachable had matured to become stupid and dull (hO). Toward the end of the nineteenth century the term was popularized by Kraeplin in its Latinized form, dementia praecox (33). Under this rubric he subsumed disorders which until then had been considered separate clinical entities: catatonia, hebephrenia, and certain delusional states. The beauty of his classification ...of seeing a unifying thread coursing through diverse forms of behavior ...was an advantage 1 which.psychiatrists were loathe to give up even when they recognized that the illness did not always progress to dementia nor always begin in youth. Rather the term had to be changed and some new fundamental aspect brought into focus. In 192M. Bleuler advocated the term schizophrenia and stated that the fundamental symptom.of the disorder was a splitting of the personality ...a splitting among the thought, motor, and emotional processes ...a splitting of the mind (5). If this caused any satisfaction, it was short-lived. Just four years later, Strecker and Wiley stated that schizophrenics formed into two psychotic groups, a poor prognostic group which was composed of patients who seemed to have a basic and con- stitutional seclusive make-up, and a better prognostic group which.was composed of patients whose psychosis was acute with a stormy onset precipitated by some significant situation (we). Rumblings were heard from.the psychoanalysts who not only be- rated the term itself as superficial and grossly inadequate but who felt that it embraced a number of conditions which varied widely in psychosexual level and stage of libidinal fixation (16, 19). It became increasingly evident that a heterogeneous popu- lation was encompassed under the label of schizophrenia. Kraeplin's nosology was repudiated by many and all that remained was to find a new conceptualization. Thus there began to emerge 2 a distinction between schizophrenics which held great promise: process vs. reactive. The process-reactive concept did not spring full-grown from the head of Zeus but is an evolution from the thinking of a.number of men. It was Sullivan who believed that schizophrenia should be reserved as a label for those disorders of living which have an acute onset due to some situational stress and a good prognosis, while the term "dementia praecom’should be applied to that disorder which has an insidious onset, poor prognosis, and seems to be an organic, degenerative disease (AT). In a similar vein were the comments of Strecker and Wiley (we). Differences in recovery rates among schizophrenics impressed Langfeldt and he introduced the term."process" to designate those patients whose chances of improvement were worse than the acute or "schizophrenifonm" (35). D. E. Cameron focused on early symptoms and distinguished among schizophrenics on this basis: hypoactive behavior present for a long thme versus hyperactive behavior present for a relatively shorter time prior to hospitalization (8). The prepsychotic personality was ems phasized by Darrah, who felt that dementia praecox.implied not only an insidious onset progressing to deterioration but also a schizoid prepsychotic personality. Schizophrenia, however, implied, in addition to an abrupt onset with a milder nondeterio- rating course, a usually adequate prepsychotic personality (12). 3 {I ulill‘ III I] In a study of fifty-nine case histories, Wittman and Steinberg found some confirmation for Darrah's views. They noted that about 50% of these histories described a life-long introverted personality with no specific onset while the other half des- cribed disorders apparently based on environmental factors (52). The continuum.approach was applied by Bellak in consider- ing etiological factors in schizophrenia. He believed that these factors ranged from the completely organic to the comp pletely psychogenic with the two polar types having different prognoses. The favorable prognostic group is psychogenic and has an acute onset, an atypical (not schizoid) personality, precipitating factors and confusion are present, while affect is, more or less adequate. Many of these phenomena, he felt, could be understood in terms of psychological dynamics (3). Thus the conception of two different groups unfortunately sharing the same diagnostic label grew. Briefly stated, the process or true or chronic schizophrenic was considered to be the patient with a long history of poor psychlogical adjustment which eventually culminated in hOSpitalization; the reactive or benign or acute schizophrenic was one with an overtly good psychologi- cal adjustment prior to hospitalization whose psychotic onset was sudden with some traumatic factor usually evident. B. Functions and Limitations of the Process - Reactive Concept Utilizing this distinction, Kantor, et. al. found that process patients gave psychotic Rorschachs while reactiveis seemed to give relatively nonpsychotic Rorschachs (29); Brackbill and Fine reported that the Rorschachs of process and organic patients were virtually indistinguishable indicating the possibility of central nervous system.pathology in process schizophrenia (7); Devault found that differences in physiological responsiveness exist between the two groups (13). Though these results have been encouraging, the distinction.itse1f has been attacked. Becker has strongly made his point that process and reactive patients do not exist as pure types. Dismayed by difficulties in classification and overlapping etiological factors in case histories, he has advocated the abandonment of considering the process-reactive distinction as a dichotomy and has offered in its stead the view that they be considered end-points along a continuum.of personality adjustment (2). Becker's criticisms appear justified and hencefortthy "process" and "reactive", the author wishes to imply "predominantly process" and'pre- dominantly reactive." However, there is at least one alter- native to considering the distribution along a process-reactive continuum.as normal and that is to consider the distribution as'bhmodal. Such an alternative seems more in keeping with 5 the spirit of the work already done in this area and with the very conception of process and reactive schizophrenia itself. The experiment which is to be described in this paper was primarily empirical not because there is a lack of theoreti- cal formulations in this area but because there is a veritable host of differing eXplanations with varying degrees of verifia- bility and probability. Virtually eVery conceivable brain anomoly has been.proposed as an eXplanation for schizophrenic behavior. These have been dutifully recorded by Brackbill (6) and it is to his credit that he has emerged from.the welter of positive, negative, and ambiguous results with a feeling of optimism.about research in this area. Similarly, psychological eXplanations have ranged from.adjustive reSponses to painful problems of modern life (39) to the contention that the schizo- phrenic is preoccupied with his visceral sensations and that his symptoms lack purpose (38). It would appear logical that prior to championing any theoretical position with regard to schizophrenia, some attempt should be made to ascertain the nature of the beast and whether or not theorkas are being built on quicksand. The clarification of the nature of schizophrenia is one of the functions which it is hoped the process-reactiwe classification.will achieve; the eXplanation of conflicting ex- perimental findings is another. It is difficult to conceive of anything that is more frustrating or disheartening to the scientist than similar eXperiments with dissimilar results. Though research in schizophrenia cannot claim.any monopoly on such conundrums, it has had more than its share. In particular, the question of whether schizophrenics suffer any real psychological deficit has long been a subject for investigation and has produced its quota of mixed findings. Layman investigated this problem.by administering the Stanford-Binet,Rorschach, and various performance tests to twenty schiZOphrenics before, during,and after the administration of sodium.amytal. In general, the performance of the patients under amytal was superior and afterwards, it tended to revert to their previous level of functioning. He attributed the changes to the removal and restoration of inhibitions (36). So far no irreversible deficit and this seemed to also be the conclusion of Shakow and Goldman.who studied speed of tapping. They found that schizophrenics tapped significantly less and were more variable than normals. However with practice the scores of the schizophrenics showed some improvement and became less variable. It was reported that the diagnostic types of schizo- phrenia distinguished themselves in various ways, e.g., cata- tonic patients did best and the authors considered a major part of the differences to be due to the attitudes and degree of 7 cOOperation of their schizophrenic subjects (uh). Essentially the same results were found by Huston, et. al. with respect to reaction time (28) and by Huston and Shakow with respect to performance on the pursuit rotor and prod learn- ing. The latter authors attributed the poor performance of their patients on the pursuit rotor to an impaired ability to assume and maintain a set (26). They gained indirect support for this view from Freeman, et. al, who claimed their patients had a heightened reaction to external stimulation which was greater than that of normals (l7) and from Chapman, who found that schizophrenics were more easily distracted than normals (10). However, when Huston, et. a1. directly tested it, they reported their schizophrenics apparently unimpaired in the ability to shift set and maintain goal orientation (25). A recent study by H. E. King produced familiar results but a different interpretation. Using chronic schizophrenics who performed such psychomotor tasks as reaction time, speed of tapping, and finger dexterity, he reported his psychotic group retarded on all tests (32). So impressed.was he by these findings that he concluded psychomotor capacity was disturbed in a fundamental sense, that it was ” ...a reflection of defect at the core"(32, 156)", and that " ...retarded and faulty psycho- motor response is regarded as a basic indication of this state of psychobiologic maladaptation" (32, 156). 8 A subsequent study (31) used acute schizophrenics with an operant motor task of pulling a plunger for rewards of candy and cigarettes. Its hypotheses were a logical ex- tension of H. E. King's conclusion: clinical improvement will be positively related to an increased Operant rate; severity of illness will be inversely related to operant rate of response. Neither hypothesis was supported. After reviewing a number of studies in this area, Hunt and Cofer concluded that the deficit of schizophrenics was not due to any loss in fundamental capacity but rather to defective motivation and control of performance (24). Defective motivation might explain Winer's findings that, except for paranoids, schizophrenics did poorer on incidental learning than normals (51). Yet defective motivation would not seem omnipresent, since Greenberg did not find any great impairment in incidental learning with paranoid and acute schizophrenics (20). It would be expected from Hunt and Cofer's cOnclusion that increasing motivation should result in better perfor- mance. Cohen reported that increasing motivation had a mild but insignificant facilitative effect (11); Stotsky reported that he considered his findings on the relationship between motivation and reaction time unclear (45); Lair found that praise was better than reproof but reproof was better than nothing in learning verbal materials but that praise and re- proof had no effect on card sorting performance (34). {The matter does not seem to gain too much in clarity by Considering what sort of stimulus constituted 9 an effective motivator since Garmezy, simply by illuminating the word "WRONG", produced sufficient punishment to cause avoidant responses (18) while Hirschman's personal entreaties had little effect on the patients in his eXperiment (23). The belief that conflicting eXperimental results may be largely ascribed to a heterogeneous schizophrenic pOpulation has given additional impetus to process -reactive research. This is a second function which it is considered this new form of classification will achieve: a.means of explaining divergent and ambiguous findings and of increasing reliability in experimentation. Nevertheless, dichotomization of a schizo- phrenic sample is still no guaranty of clear-cut results. Following the completion of this experiment, an article appeared by Rodnick and Garmezy (h2) in which they discussed a number of experiments conducted primarily at Duke University. 0n the assumption that psychological deficit in schizophrenia was due to faulty motivation, they predicted that social censure would produce deficit performances in schizophrenics. Subjects were divided into good and poor premorbids based on ratings of the case histories in terms of adolescent and recent sexual adjustment, social aspects of recent sexual life, and past and :recent adjustment in social relations. There is considerable similarity between good.premorbid schizophrenics and reactives and poor premorbids and process. This the authors recognized 10 and were of the opinion that the poor premorbids might be viewed as organic since there is some evidence to suggest that process patients are organic. Because the poor premorbids have been subjected to social censure over a long period of time, the authors hypothesized that these patients would be more sensitive to social censure and thus exhibit greater deficit in performance than good pre- morbids under some sort of punishment condition. The illogicality of this line of reasoning is that, if the poor premorbids were constitutional or organic, then less rather than more sensiti- vity to psycho-social factors would be expected. Nevertheless, they have reported several studies to be described below, which they believe support their hypothesis. Bleke (h) had assumed that reminiscence is a function of dissipation of interference and that censure would produce interference. Therefore, he hypothesized that poor premorbid schizophrenics would have more interference during learning under censure and so show more reminiscence than good premorbid schizophrenics. The subjects were presented with.a list of fourteen nouns and had to pull and push a lever in a certain pattern for each. Under the reward condition, when they mani- pulated the lever correctly, the word "RIGHT" was illuminated and, when they were incorrect, an empty slot was illuminated. Under the censure condition, when they manipulated the lever 11 incorrectly, the word "WRONG" was illuminated, and when they were correct, an empty slot was illuminated. Three measures of reminiscence were obtained: retention, relearning, and improvement in recall. The retention and relearning data were ambiguous-and not reported. ,Using improvement in recall as a measure, the poor premorbids under the censure condition showed more reminiscence than the other groups. Just as significant, though not commented on, the good premorbids under censure were the only group whose score was negative (avoidance?). Rodnick and Garmezy state later in their article that a flaw was found in the design of Bleke's list which caused subjects who were stereotypic in their movements to have higher scores than those who were variable.i Thus they now consider Bleke's results to be equivocal. An experiment by Alvarez, which is unpublished, is also reported by Rodnick and Garmezy (u2). The dependent variable was shifts in preference judgments as a measure of aversion toward previously neutral stimuli which had been associated with censure. Good and poor premorbids and normals ranked six pictures in terms of preference. The four of the six which were not ranked highest or lowest were assumed to be neutral. Two of these four were then associated.with "WRONG" and two with "RIGHT". Then rankings of the six pictures in terms of preference were again made. Results were that poor premorbids 12 showed significant decrements in preference for the censured photographs and did not react as effectively to the "rewarded" photos as goods and normals. The authors interpret this second result as generalization of aversive tendencies among the poor premorbids. However, since the subjects saw the four "neutral" pictures more often than the two extreme ones and there was no control for this variable, it is just as possible to inter- pret the results in terms of the poor premorbids showing less preference for pictures they have seen more frequently. In a third experiment by Dunn (1h) the schizophrenics were not classified into good and poor premorbids. However, after the experiment was completed, Rodnick and Garmezy made this classification. Three silhouetted scenes depicting a mother and a young boy in a scolding, feeding, and whipping relationship were used along with a control picture of a house and tree. Five variations of each standard picture were made by, for example, varying the position of the mother's arm over a #50 arc. The problem.which confronted the subjects was one of discrimination. If the subject thought the comparison pic- ture was the some as the standard, he pulled a lever; if he did not, he pushed the lever. Results of this experiment were that there were no significant differences. However, on the scolding pictures the schizophrenics were less discriminative than the normals which was significant between the 5% and 10% 13 levels. Dunn believed that in those cases where the relation- ship with.the mother constituted a problem area, the subject tended to perform less adequately on the scolding series. He does not state why this prOblem.area should only be manifested in scolding and not in feeding and whipping pictures. When Hbdnick and Garmezy classified Dunn's subjects into good and poor premorbids, they found that the poors tended to dis- criminate less than the goods. This difference was not signi- ficant nor can it be interpreted since not only could it have been due to aversion but also to deficit in discriminative ability and/or indifference on the part of the poor premorbids to the taSk. These, then, are the eXperiments which Rodnick and Garmezy offer in support of their hypothesis that poor premorbids are more sensitive to social censure than good premorbids. However, they also report several studies which are in line with the results of this eXperiment. An unpublished study by Ussery involved the analysis of TAT stories. He found that the good premorbids were more productive, were more socially and emotion- ally oriented, and showed more stivings for independence than the poor premorbids. A study by Harris, which is as yet un- published had the following results: good.premorbids tended to underestimate the sizes of mother-child pictures (avoidance?) while poor premorbids tended to overestimate the sizes of all 11+ pictures including a neutral one (approach?). The hypothesis of Rodnick and Garmezy seems to have little evidence to sup- port it and some evidence which is against it. C. Formulation of the Hypothesis The essence of this process-reactive concept is that the former represents a chronic schiZOphrenic illness which becomes progressively more severe, while in the latter, an individual with an apparently satisfactory adjustment "suddenly" becomes psychotic. In dividing a schizophrenic sample according to this concept, two groups are formed: 1. Predominantly process: those patients whose case histories indicate reason for and behavior characte- ristic of unsatisfactory psychological adjustment over a long period of time. 2. Predominantly reactive: those patients whose case histories indicate a relatively satisfactory psycholo- gical adjustment with a mental illness seeming to occur over a short period of time. It was this difference in length of time during which symptoms are manifested which.was regarded as a key variable in this eXperiment and in the formation of its hypothesis. Duration of symptoms had been recognized as a probable means of eXplaining the continuation of maladaptive behavior 15. long after the individual had been placed in a safe, permiss- ive environment and the conditions which precipitated these symptoms were removed. It was during this length of time that certain changes were postulated to occur in the motivation of the symptomatic behavior. Cameron and Magaret stated: "The persistence of regressive behavior, once it has developed, will depend in part upon the per- sistence of the conditions precipitating it and in part upon the degree to which it satisfies the individual as a way of life or leads to secondary gains. If the conditions responsible for regres- sion last a long time, the regressive pattern are likely to become established through long practice and to be integrated with the rest of a person's organized habitual behavior. And even though these conditions disappear, there is always the possibility that regressive reactions may in themselves prove rewarding (functional autonomy)." (9, p. 225). Similarly, Tilton attempted to explain the persistence of withr drawn behavior from the framework of elicitation theory. He assumed chronic schiZOphrenics had a strongly conditioned avoidant attitude and that a strong relaxational pattern elici- ted approach responses to the safety of withdrawal (h9). What is to be noted is again the relevance of the time variable. It is over a long period of time that certain changes were assumed to occur in maladaptive behavior: integration, func- tional autonomy, or/and a change in direction from escape or avoidance to approach. If these assumptions were valid, if chronic deviant be- 16 havior became over-learned, intrinsically satisfying; or/and altered in aim, then it was logical to assume that differences in the character of schiZOphrenic behavior existed between pro- cess and reactives. For since the process pattern of regres— sion and withdrawal is considered to be of long duration where- as the reactive pattern is supposed to occur in a brief period of time, then the above changes which were believed to take place should be found in the process group but not in the re- active. In other words, though the process schizophrenic origi- nally reSponded to his noxious environment with regressive and withdrawn behavior, as time passed, this behavior either became habitual or/and more elicited by the relaxation and safety it afforded than by existing environmental pressures i.e. it changed from.predominantly escape and avoidance to approach. However the withdrawn and regressed behavior of the re- active schizophrenic is not only of insufficient duration to allow such changes to take place but virtually by the definition of the concept itself, is a response made in adulthood to an environment that is over-whelmingly unbearable i.e. it is pre- dominantly escape and avoidant. Some support for this view was afforded by an experiment by G. F. King, who investigated physiological responsiveness to mecholyl among process and reactives. He found that the process but not the reactives seemed similar to normals in their reactions to the drug. King 17 exxflained these results with the assumption that the gradual v .05‘110). These results indicate that the normals sorted the cards significantly faster than either the reactives or the process but that the two schizophrenic groups barely approached significance between their mean sorting times. However, since it was predicted that the reactives under FP would sort slowly (Prediction 1), their times were removed to see what effect this would have upon the inter- pretation of the data. The reactive mean and standard de- viation were now 94 and 31.8, respectively. Again t tests were conducted using the formula recommended by Edwards (15) when there is heterogeniety of variance. The t ratio be- tween normals and reactives becomes 1.362 and is not signi- ficant (R>.10). However, the t between process and reactives now becomes 2.260 which is clearly significant (P<.OS). To summarize, when all four reactive conditions were pooled, the results indicate that the reactives sorted about as slowly as the process subjects and significantly slower than the normals. When the times of the reactives under FF 49 ‘were removed from the pooled data, the results indicate that the reactives sorted about as fast as the normals and significantly faster than the process. In view of the im- portance of the reactive FP condition, it was wondered why the Groups X Conditions Interaction was not significant. There are two reasons which may explain why this Interaction was not significant: the relatively small num- ber of subjects in each group and the high variabilities. To test the validity of this explanation, it was decided to compare the mean sorting times of subjects on trials 9-11. Variability was less on these trials because there was no longer any significant amount of learning taking place (Be- tween Trials 2-8 F = 56.29, P(.01; Between Trials 9-11 F = 1.63, P).05) and because everyone was supposedly sorting fast. Reactive subjects who had formerly sorted under FP were compared with reactive subjects who had formerly sorted under the two light conditions. A similar comparison was made with the process subjects (See Table 6). The signifi- cant differences which were found within each of the two TABLE 6 Mean Sorting Time For Trial During Trials 9-11 ann "FL" & "SLN t (N #10) ’ .(N = 16) R 88 65 2.19; P<.05 P 70 115 2.30; P<.05 schizophrenic groups between "FF" and "FL" and "SL" indicate 50 that the lack of a significant Groups X Conditions Inter- action was due more to the variability and small N's of the groups and less to the absence of any effect of the con- ditions. Assuming that the effects established on trials 2-8 may be in part permanent or learned, here again is evidence in support or Prediction 1. F. incidental Resultsv Of particular interest: is that the normals sort rapidly under all four conditions (See Fig. 4). This in- dicates not only that they were indifferent to the pictures and light but also that they were motivated to do well despite the fact that they were free to sort slowly. Con- trast this with the curves of the reactives who sort slowly under FF and rapidly under the other three conditions (See Fig. 1) or with the process subjects who sort rapidly under FP and slowly under the other three conditions (See Fig. 2). Both the reactives and process seemed indifferent to the light, i.e., reactive subjects performed the same under both light conditions and so did the process subjects. Yet the reactives sorted rapidly while the process did not which would indicate that the reactives, unlike their process brethren, came equipped with their own motivation when none was explicit in the experimental procedure. Another indication that the process and reactives were drawn from different population was evidenced in the heterogeniety of variance found on trials 2-8. Since no 51 Mean sorting times of the normal group. 5?. predictions had been made about variability, two-tailed tests were used and the tabled probability levels of F doubled. There was no significant difference between the variability of the normals and reactives (F = 1.16). How- ever, process subjects were significantly more variable than either reactives (F = 2.59, P(.O2) or normals (F = 3.03, Pe§02). 0n trials 9-11, the process group was significantly more variable than the reactives (F = 2.095, P(.05) and the normals (F = 9.647, P<.Ol). In addition, the reactives were significantly more variable than the normals (F = 3.266, P<.01). The mean sorting times on trials 9-11, which are presented in Table 7, also differed significantly (Reactives vs. Process, t = 2.830; Reactives vs. Normals, t = 2.137; Process vs. Normals, t = 4.328). Table 7 Pooled Mean Sorting Time Per Trial During Trials 9-11 Reactives Process Normals (N = 36) (N = 36) (N = 36) Mean 76 104 63 s. D. 31.8 54.7 17.6 To summarize the results on these last three trials, the normals as a group sorted the cards significantly faster and were significantly less variable than the schizophrenic sample but within the schizophrenic sample, the subjects con- sidered reactive sorted the cards significantly faster and 53, were significantly less variable than the subjects considered process. II. Sub—Experiment The purpose of the sub-experiment was to determine whether reactives were responding to the pictures which represented an area of frustration, conflict, and/or threat 'in particular or to pictures in general. Those subjects employed do not differ significantly with respect to age, IQ estimate, and length of hospitalization (See Table 8). TABLE 8 Controlled Variables in the Sub-Experiment Mean Age Mean IQ Mean Mos. No. Hospitalized Schiz. Undif. P (N = 10) 29 106 5 R (N = 10) 30 105 5 In describing the results, the following designations will be used: T stands for pictures of people considered to represent areas of frustration, conflict, and/or threat; NT for pictures of people not so considered; I for pictures of inanimate objects. A. Prediction 5 It was predicted that reactives would look at NT and I pictures longer than at T pictures. The mean time that each subject looked at each type of picture was computed. Sign tests were then employed. With an N of ten, 1.25 or 54 fewer minus signs among the ten differences were needed for significance at the 2%% level using a one-tailed test. Nine of the reactives looked at NT pictures longer than at T. All ten reactives looked at I pictures longer than at T. Both of these positive effects were significant and in the expected direction. On the other hand, only six of the ten reactives looked longer at I pictures than at NT (not signi- ficant). These results do not necessarily indicate that reactives avoid only T pictures (since, as seen below, the process spend longer time with all photos)but that reactives avoid T more than NT and I pictures, i.e., the avoidance of the reactives is not wholly generalized. B. Prediction 6 It was predicted that the process subjects would loek at the three types of pictures for about the same length of time. Among the ten process subjects, five looked at T pictures longer than at NT, six looked at I pictures longer than at T, and five looked at I longer than at NT. In short, for the process subjects, as was predicted, there are no significant differences between the times looked at threatening, non-threatening, and inanimate pictures. These results indicate that the process subjects would have probably performed about the same in the Main Experiment regardless of the type of picture that was used. C. Other R§sults Relevant To the Hypothesis While the results seem to point clearly to the con- clusion that the reactives avoid while the process do not, 55 it is more difficult to determine whether the process were indifferent to (Prediction 4a) or approached the pictures (Prediction 4b). Some help in evaluating this issue is afforded by these final results. The mean time the process group looked at the pictures was significantly longer than the mean time of the reactives (t = 3.25; P<.01; See Table 9). Run tests were then used to determine if the differences between the groups with respect to the times they looked at any one type of picture were significant. The statistic derived by this non-parametric technique is C, which in this case would have to equal -l.645 for significance at the 5% level using a two-tailed test. With T pictures, C = ~3.2l; with NT pictures, C = -l.83; with I pictures, C = -l.83. Thus the process group looked at the pictures significantly lon- ger than the reactives and differed significantly from the reactives in the times they looked at each of the three types of pictures. It might be argued that the above differences are a reflection of generally slow performance by process sub- jects. This is a point which cannot be lightly dismissed. However, evidence that the process subjects may have appro~ ached the pictures includes: (1) their learning curves; (2) the fact that the subjects who sorted under FP were significantly faster on trials 9-ll than the subjects who sorted under the light conditions; and (3)0bserxations that these subjects did seem motivated under the picture 'f/ ,/V conditions. TABLE 9 Seconds Spent Looking At The Three Types of Pictures T NT I TOTAL Mean Mean _ Mean V Mean P. 6.8 7.8 6.7 7.1. R. 3.2 4.2 . . 4.4 3.9 L- D. Incidental Results As an afterthought, subjects were required to sort tie pictures they had seen into "Like" and "Dislike" cate- gories. This was to determine if there was any relationship between the time a subject looked at a picture and whether ‘ or not he expressed a preference for it. Since the cate- gories do not represent genuine dichotomies, point biserial correlations were computed (21). For the reactive group, rpbi = +.275 which is not significant. For the process group, rpbi = +.33l which is not significant. The differ- ence between these two correlations is also not significant. However, it should be noted that these correlations are in the same direction and about the same size. It is probable that with larger samples, these correlations would be sig- nificant. If so, it would indicate some positive relationship between the length of time process and reactives looked at pictures and whether they expressed a "liking" for them. 57 Discussion ‘A. Discussion of Results This experiment began with the hypothesis that re- aactive subjects would avoid and that process subjects would runt avoid stimuli which might be considered to represent areas of frustration, conilict, and/or threat. In the pre- sent study, the stimuli were pictures judged as represen- ‘tative of such areas by clinical Isychologists while avoidance ‘was manifested by either sorting slowly or rapidly, which- ever would cause these pictures not to be shown. The re- sults supported the hypothesis. (See Figs. 1 and 2). That the reactives did avoid the pictures was in- dicated not only by their learning curves under the four conditions but by other statistical evidence. They sorted slowly under FF and saw significantly fewer pictures than would have been expected from their performance under the light conditions. In addition, it was found in the Sub- Experiment that this avoidance was not wholly generalized to all pictures since they looked at non-threatening and inanimate photos longer than at threatening ones. It seems reasonable to assume that the experiment would have been more sensitive in eliciting avoidance if there had been threatening pictures selected specifically for each subject rather than a variety of pictures covering a variety of problem areas. However, such a procedure would have detracted from the generality of the conclusions which 58 might be drawn. 0f concern was not whether the reactive with a mother problem avoids pictures of elderly females but whether the reactive with a mother problem avoids pictures of virtually everyone. The psychotic nature of withdrawal lies in its excess. Thus the pictures were composed mainly of people, both men and women, and were judged to represent a diversity of possible areas of frustration, conflict, and/or threat. That the reactives avoided such pictures was in accord with the clinical conception of schizophrenic withdrawal as being an over-generalized response. There seemed little doubt from the results that process subjects did not avoid the pictures. Instead the question became whether they were indifferent to or appro- ached them. Though the evidence was somewhat equivocal, it favors the interpretation that they were motivated to see them. During the experiment, it was noted that these subjects took interest in the photographs and a few of them even expressed displeasure when, after trial 8, they were informed that no more pictures would be shown. In contrast, several reactives requested that they be permitted to sort the cards without seeing the pictures. Normals sorted rapidly under all four conditions which meant that they were relatively indifferent to the photos and light. The schizophrenics were also indifferent to the light. Yet reactives under the light conditions sorted rapidly and were similar to normals in their perform- ance. They showed interest in their sorting time from trial 59 to trial, verbalized goals as to the time they wished to attain, and inquired as to how their performance compared with that of others. On the other hand, process subjects under the light conditions sorted slowly. They appeared bored or indifferent. Some of them whistled or hummed tunes as they sorted the cards, there were inquiries as to how many trials were left, and, as was mentioned earlier, a few had to be excused from the experiment because they simply refused to continue it. However, under the FP condition, the Process subjects sorted the cards about as fast as any other group. There- fore, the significant difference between the mean sorting times of the process and the reactives and the process and normals is not interpreted as a deficit in psychomotor per- formance but as a deficit in motivation. It would seem reasonable that such a deficit must be considered in the evaluation of virtually all psychological experiments with schizophrenic groups. Otherwise, experimental differences found between process and reactives and normals which are a reflection of differences in motivation may erroneously be ascribed to differences in the variable being measured. In other words, care must be taken to insure that in exper- iments, for example, on reaction time or critical flicker fusion, all Stbjects are about equal in motivation. Though experimental directions may be sufficient to insure the cooperation of normals and reactives, additional incentives 60 ‘may be required fOr process subjects. In view of the present results, it seems that an effective incentive is the showing of pictures. As stated in the Introduction, the experimental hypothesis was based on the assumption that as schizophrenic symptoms persist over long periods of time, changes occur in their motivational basis. Although withdrawal may or- iginally have been in response to a noxious environment, it later may become habitual or motivated by secondary gain. Therefore, the difference in prognosis ascribed to process and reactive schiZOphrenics need not be explained in terms of organic factors. A psychological explanation is possi- ble in which the duration of the symptoms is a crucial variable. Another explanation for the present results is based on the assumption that reactives are more prone to experience anxiety than process schizophrenics. Thus when reactives see pictures of the T type, they experience anx- iety and avoid whereas process subjects do not become anx- ious and so do not avoid. However, rather than being contra- dictory to the interpretation offered above, the two seem compatible. The reactives are more prone to experience anxiety than the process because their withdrawal is still in response to an environment which they perceive as noxious. Still another explanation may be derived from Selye's. views on stress (43). This would probably state that the reactive, who was recently subjected to sudden stress, had 61 his homeostasis upset whereas the process, who has had longer to adjust, is more stable homeostatically. Though this may be a partial explanation, there are several points which should be considered in its evaluation; the imbalance of homeostasis must be viewed relatively since no living organism ever achieves a state of perfect balance; it is not possible to predict from this whether reactives would approach or avoid noxious stimuli, since in achieving equi- librium, homeostatic mechanisms usually operate in an oscill- atory fashion (22); a relatively stable homeostasis would seem to have difficulty in explaining the progressive changes observed in schizophrenics by Arieti (l). B. Suggestions for Future Research Since reactive schizophrenics, unlike process, appear to be avoiding a noxious environment, it would be expected that these two groups should respond differently to various psychiatric treatment procedures. Because shock is noxious, it should be less effective than tranquilizers in the treat- ment of reactives. 0n the other hand, the very punitive nature of shock may elicit from process patients responses other than their symptoms while it is difficult to see any therapeutic value that might be gained by using tranquilizers. Therefore, it would be expected that shook would be more e- ffective than tranquilizers in the treatment of process schi- zophrenics. To be tested, recently hospitalized patients would have to be used as subjects. It would be predicted from the first explanation 62 offered that reactive schizophrenics who exhibit their sym- ptoms over a long period of time should have a change in motivation similar to process patients. These reactives should not avoid stimuli considered to represent areas of frustration, conflict, and/or threat. This could be tested by using schizophrenics hospitalized for, let us say, at least two years and classified by their case histories into process and reactive. These patients could perform in an experiment similar to the one described here and it would be expected that reactives would sort more or less like the process. 63 SUMMARY AND CONCLUSlONS The essential difference between process and reactive schizophrenics is that the former manifests schizoid sym- toms from childhood while the latter appears to suddenly become psychotic when an adult. Assuming that changes in motivation occur as behavior persists over long periods of time, such changes would be more likely in process than in reactives. Therefore, while the withdrawal of the reactives may be in response to a noxious environment, the withdrawal of the process may be habitual or motivated by secondary gain. Thus it was hypothesized that reactive subjects would avoid stimuli considered to represent areas of frustration, con- flict, and/or threat while process subjects would not. The stimuli were magazine photographs judged by five clinicians as to whether they represented an area of frus- tration, conflict, and/or threat. A small flashlight bulb was used as a neutral stimulus for control groups. The experimental ta5k was to sort an ordinary deck of playing cards onto a board divided into quadrants for eleven trials. On trials 1 and 9-ll, all subjects were requested to sort rapidly. From trials 2-8, subjects were free to sort fast or slow under one of four conditions: FP subjects were told they would see pictures if they sorted fast; SP subjects were to see pictures if they sorted slowly; FL subjects were to see a light if they sorted fast; SL subjects were to see a light if they sorted slow. Fast and slow were defined in relation to the subject's speed on the preceeding trial. The light 64 or five pictures were presented after each trial during which the subject had sorted according to directions. The question was whether subjects would sort so as to see or so as not to see the pictures. Thirty-six male hospitalized veterans diagnosed as schizophrenic were classified as process and thirty-six as reactive. A comparison group consisted of thirty—six male hospitalized veterans not considered psychotic. The three groups were controlled for age, IQ estimates, and length of hospitalization. It was predicted that reactives would sort slowly under FP, rapidly under SP, and perform about the same under the two light conditions (avoidance). For the process group, it was predicted that either they would per- form about the same under all four conditions (indifference) or that process subjects would sort rapidly under FP, slow- ly under SP, and perform about the same under FL and SL (approach). No predictions were made for the normals who were used only to make comparisons more meaningful. The results indicated that the reactives did avoid while the process did not. While it was not conclusively demonstrated whether the process subjects were indifferent to or appro- ached the pictures, the evidence favored the interpretation that the process approached. Because reactives had only seen pictures considered to represent areas of frustration, conflict, and/or threat, a Sub-Experiment was conducted to determine whether they were 65 avoiding these pictures in particular or pictures in general. Ten of the reactives and ten of the process who performed under the light conditions were used as subjects. A pile of thirty-three pictures was presented to each subject who was told to simply look at them for as long as he liked. Eleven of these pictures were judged to represent an area of frustration, conflict, and/or threat; eleven were judged not to represent these areas; eleven were photos of inan- imate objects. It was predicted that the process subjects would look at the three types of pictures for about the same length of time but that the reactives would look at the threatening pictures less than the other two types. The results of the Sub-Experiment were in accord with these pre- dictions. Other results indicated that reactives and normals are motivated to perform effectively under conditions where process schizophrenics perform with indifference. Thus the significant differences between the mean sorting times of the three groups were interpreted in terms of differences in motivation rather than as any deficit in psychomotor ability on the part of process or reactives. The major conclusions with respect to the population employed are: l. Schizophrenics may be fruitfully divided into process and reactives. 2. Reactive schizophrenics avoid pictures con- sidered to represent areas of frustration, 66 conflict, and/or threat. Process schizophrenics do not avoid pictures considered to represent areas of frustration, conflict, and/or threat. Duration of symptoms is a crucial variable in the investigation and understanding of schizophrenia. e7 1. 9. 10. 11. 12. 13. 14. 15. BIBLIOGRAPHY Arieti, S. Interpretation of_schizophrehia. New York: R. Bruner, 1955. Becker, W. C. A genetic approach to the interpretation and evaluation of the process—reactive distinction in schizophrenia. g. abnorm. soc. Psychol., 1956, 53, 229-2360 Beléak, L. Dementia praecox. New York: Grune & Stratton, 194 . Bleke, R. C. Reward and punishment as determiners of reminiscence effects in schizophrenic and normal sub- jects. g. Pers., 1955, 23, 479-498. Bleuler, E. Textbook pf psychiatry. New York: Mac- millan, 1924. Brackbill, G. A. Studies of brain dysfunction in schizo— phrenia. Psychol. Bull., 1956, 53, 210-226. Brackbill, G. A. & Fine, H. J. Schizophrenia and central nervous system pathology. g. abnorm. soc. Psychol., 1956. 52. 310-313- Cameron, D. E. Early schizophrenia. Amer. i. Psvchiat., 1938. 95. 567-582- Cameron, N. & Margaret, Ann. Behavior pathology. Cam- bridge, Mass. Houghton Mifflin, 1951. Chapman, Loren. Distractability in the conceptual per- formance of schizophrenidcs. g. abnorm. soc. Psychol., 1956, 53, 286-291. Cohen, B. D. Motivation and performance in schizophren- Darrah, L. Shall we differentiate between schizophrenia and dementia praecox? g. nerv. ment. Dist., 1940, 91, 323-328. Devault, S. Physiological responsiveness 1p reactive and process schizophrenia. Unpublished doctoral disser- tation, Michigan State Univer., 1955. Dunn, W. L., Jr. Visual discrimination of schizophrenic subjects as a function of stimulus meaning. Q. Pers., 1954, 23, 48-64. Edwards, A. L. Experimental design ip psychological pg- search. New York: Rinehart, 1950. 68 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. BIBLIOGRAPHY (cont.) Fenichel, O. The Psychoanalytic theory pf neurosis. New York: Norton, 1945. Freeman, T., McGhie, A., & Cameron, J. L. The state of the ego in chronic schizophrenia. Brit. j. med. Psychol., 1957. 30. 9-18. Garmezy, N. Stimulus differentiation by schizophrenic and normal subjects under conditions of reward and punishment. J. Pers., 1952, 20, 253-276. Glover, E. Psycho-analysis. 5(2nd ed.) London: Staples Press, 1949. Greenberg, A. Directed and undirected learning in chronic schiZOphrenia. Dissertation Abstr., 1954, 14, 1457—1458. Guilford, J. P.’ Fundamental statistics lg s cholo and education. (2nd ed.) New York: McGraw-Hill, 1950. Guyton, A. C. Textbook 9f medical ppysiology. Phila- delphia: W. . Saunders, 1956. Hirschman, W. A study of the communication of learned meaningful material by schizophrenics and non-psychotics and its relationship to interpersonal involvement. Dissertation Abstr., 1954, 14, 186-187. Hunt, J. McV. & Cofer, C. N. Psychological deficit. In J. McV Hunt (Ed.), Personality and the behavior disorders. Vol. II. New York: IRoland, 1944. Huston, P. E. Cohen, B. D. a Senf, Rita. Shifting of set and goal orientation in schizophrenia. J. ment. Sci., 1955, 101, 344-350. Huston, P. E. & Shakow, D. Learning in schizophrenia: .I. pursuit learning. J. Pers., 1948, 17, 52-74. Huston, P. E. & Shakow, D. Learning capacity in schizo- phrenia. Amer. J. Psychiat., 1949, 105, 881-888. Huston, P. E. Shakow, D., & Riggs, L. A. Studies of motor function in schizophrenia: reaction time. J. gen. Psychol., 1937. 16, 39-82 Kantor, R. E. Wallner, J. M., & Winder, C. L. Process and reactive schizophrenia. J. consult. Psychol., 1953, 17, 157-162. 59 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. BIBLIOGRAPHY (cont.) King, G. F. Personal Communication, June 1957. King, G. F., Merrell, D. W., Lovinger, E., & Denny, M. R. Operant motor behavior in acute schiZOphrenics, J. Pers.. 1957. 25. 317- 326 King, H. E. Psychomotor aspects 9; mental disease: an exploratory study. Cambridge, Mass. Harvard University Press, 1954. Kgagplin, E. Lehrbuch der psycriatrie. (6th ed.) Leipzig, 1 9 Lair, C. V. The effect of praise and reproof on learn- ing and retention of non-psychotics and schizophrenics. Dissertation Abstr., 1954, 14. Langfeldt, G. Prognosis in schiZOphrenia and the factors influencing the course of the disease. Acta psychiatrica and neurologica supplementum, 1937, 13, 2I8. Layman, J. W. A quantitative study of certain changes in schizophrenic patients under the influence of sodium amytal. J. ggp. Psychol., 1940, 22, 67- 86. Haier, N. R. F. Frustration: the stud of 4behavior without a goal. New York: McGraw-HiII, Maier, N. R. F. Frustration theory: restatement and extension. Psychol. Rev., 1956, 63, 370—388. Miller, D. R. Responses of psychiatric patients to threat of failure. J. abnorm. soc. Psychol., 1951, 46, 378-3870 Morel, B. A. Traite des maladies mentales. Paris: V. Masson, 1860. Richman, J. The effect of the emotional tcne of words upon the vocabulary responses of schizophrenics. J. gen. Psychol., 1957, 56, 95—119. Rodnick, E. H. & Garmezy, N. An experimental approach to the study of motivation in schizophrenia. In M. R. Jones (ED.), Nebraska Symposium on motivation 1957. Lincoln, Neb.: University of Nebraska Press, 1957. Pp. 109-184. Selye, H. Th3 stress pf life. New York: McGraw-Hill, 1956. 70 44. 45. 46. 47. 48. 4'90 50. 51. 52. BIBLIOGRAPHY (cont.) Shakow, D. & Goldman, R. Studies of motor function in schizophrenia: I speed of tapping. J. gen. Psychol., 1936, 15, 63—106. Stotsky, B. A. Motivation and task complexity as factors in the psychomotor responses of schizophrenics. J. Pers., 1957. 25. 327-343. Strecker, E. A. & Willey, G. F. Prognosis in schizo- phrenia. In Schizophrenia (dementia praecox). Assoc. for research Jp nervous and mental diseases, Vol. V. New York: P. B. Hoeber, I928. Sullivan, H. S. Conceptions 9: modern Psychiatry. Wash- ington: The William Allanson White Foundation, 1947. Thorpe, L. P. & Katz, B. ‘The Psychology 2: abnormal pg- havior. New York: Ronald Press, 1948. Tilton, J. R. The use of instrumental motor and verbal learning techniques Jp the treatment pf chronic schizo— hrenia. Unpublished doctoral dissertation, Michigan State Univer., 1956. Webb, W. W. Conceptual ability of schizophrenics as a function of threat of failure. J. abnorm. soc. Psychol., 1955: 50. 221' Winer, H. R. 224. Incidental learning in schizophrenics. Dissertation Abst., 1954, 14, 1002-1003. Wittman, M. P. ' Steinberg, D. L. A study of prodromal factors in mental illness with special reference to schizophrenia. 811-816. Amer. J. Psychiat., 1943/1944, 100, 71 APPENDIX 72 2. 3. 4. 10. 11. 12. 13. 14. 15. 16. 17. TABLE 10 Description of Experinental Pictures The faces of a middle-aged womar with eyes wide and a younger woman in the background.. (2) An attractive young woman wearing a sweater. (5)* Miners in a pit from an above perspective. (8) A young man holding a bowling ball apparently looking tensely down the alley; another young man and an attrac- tive woman are in the background looking in the same direction. (ll)* A man with mouth agape and blood on his head apparently trying to rise from his bed with a nurse in the back- ground restraining him. (14) An attractive young woman posed in an evening gown against a leopard skin; the head of a man facing her is to the left. (20) A negro runner at a track meet grimacing; other runners and a filled stadium are in the background. (l7)* A group of four men standing around a blueprint; one of the men is pointing at something in the distnace. (24) Interior of a paper pulp factory; two workers are visible. (33)* A man about to submerge another man in an icy pit of water. (26) A man seems to be talking seriously to a young boy on his lap. Group of weather-beaten tomb stones.* An infant with only the hand and arm of the woman holding the baby visible. Scene of a crowd at a night baseball game. * A semi-clothed young woman on a tiger skin. (28) A collection of adolescents in a teen—canteen. A smiling young man and woman in swim suits running on a beach. * 73 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 3C. 31. 32. 35. 36. TABLE 10 (cont.) A group of semi-formally attired men and women at a cock- tail party. Ferris wheel and another amusement ride at a carnival. * Group of smiling foreign soldiers gathered about an officer. Interior of a church whose walls are painted with pictures of saints and Christ.* A young man and woman asleep in bed. A waterfront scene with factories and a ship visible.* A baby in a high chair sticking its hand into the mouth of a man. (16) A hunter and his dog walking through snow.* A child about a year old with his hand around a dog. Group of young men and women seated at a barn dance. (25)* Group of adolescents laughing and singing. (10) Group of men and women in a sailboat being splashed with spray. (31)* Family group of four holding puppies. (4) Young woman fishing at a stream in the woods. (7)* Golfers walking down the green. (l9)* Men and women around a campfire at night. (23)* Groups of yo1ng men and women seated on grass in the Fall. (29)* A country scene; mountains and horses are visible. (l)* Two people fishing from a boat on a placid lake. (l3)* Inanimate Pictures Squares of pastel colors. (3)* Expressway intersection with cars. (6) 74 TABLE 10 (cont.) Canoes on a beach; palm trees are prominent. (9)* Scene of a wooded hillside. (12) Camera on a fishing basket. (15)* Rooftops in a city. (18) Pillars with plants on them. (21)* A guided missile on its launching pad. (22) An expanse of sky with four jetstreams from a plane. (27)* A house of colonial type architecture. (30) Concrete-domed airport terminal. (32). Note-Numbers in parentheses indicate order of picture in sub- experiment. Asterisk indicates colored photographs. Pictures are numbered in order of their presentation. TABLE 11 Analysis of Variance: Age Source Sum of Squares d.f. Mean Square F Methods 11.62 3 3.87 .11 Groups 15.50 2 7.75 .22 Interaction 71.00 6 11.83 .33 Within 3433.55 96 35.77 Total 3531.67 107 TABLE 12 Analysis of Variance: IQ Estimates Source Sum of Squares d.f. Mean Square F Methods 1147.55 3 382.52 .93 Groups 553.86 2 276.93 .67 Interaction 2428.77 6 401.80 .98 Within 39434.90 96 410.78 Total 43565.08 107 TABLE 13 Analysis of Variance: Length of Hospitalization Source Sum of Square d.f. Mean Square F Hethods 6.03 3 2.01 .25 Groups 42.06 ‘ 2 21.03 2.59 Interaction 26.75 6 4.46 .55 Within 780.08 96 8.12 Total 854.92 107 TABLE 14 Analysis of Variance Trial 1 Source Sum of Squares d.f. Mean Square F Conditions 943.83 3 314.61 .09 Groups 15194.13 2 7597.06 2.09 Interaction 3721.16 6 620.19 .17 Within 348897.65 96 3634.35 Total 368756.77 107 TABLE 15 Analysis of Variance of Differences Between Time on Trial 1 d 2 Source Sum of Squares d.f. Mean Square F Groups 6587.39 2 3293.69 2.38 Conditions 4117.09 3 1372.36 1.00 Interaction 22723.26 6 3787.21 2.74* Within 132381.18 96 1378.97 Total 165808.92 107 * Significant at less than the 5% level. TABLE 16 Analysis of Variance Repeated Measures Trials 2 Through 8 Source Sum of Squares d.f. Mean Scuare F Conditions 23412.60 3 7804.20 .57 Groups 222245.63 2 111122.82 8.17** Interaction 97045.98 6 16174.33 1.19 Between subjs. in same group % 1304981.03 96 13593.55 condition . T _ Total between subjects 1647685.25 107 Trialsz2-8 131591.96 6 21931.99 56.29** Interaction: trials X 4036.66 18 224.26 ‘ .58 conditions , Interaction: trials x 6720.42 12 560.04 1.44 groups ‘ Interaction: pooled subjects 238459.53 612 389.64 X trials Total within subjects 380808.57 648 Total 2028493.82 755 ** Significant at less than the 1% level. 78 TABLE 17 Analysis of Variance Mean Sorting Time Per Trial On Trials 2-8 Source Sum of Squares d.f. Mean Square F Groups 31813.47 2 15906.74 8.20** Conditions 3375.39 3 1125.13 .58 Interaction 13871.00 6 2311.83 1.19 Within 186,213.22 96 1939.72 Total 235273.08 107 F max. = 12.42; P is less than .05. TABLE 18 Analysis Of Variance Square-Root Transformation of Mean Sorting Time Per Trial On Trials 2-8 Source Sum of Squares d.f. Mean Square F Groups 66.09 2 33.04 8.00** Conditions 7.92 3 2.64 .64 Interaction 25.22 6 4.20 1.02 Within 396.60 96 4.13 Total 495.83 107 F max. = 8.67; P is greater than .05. ** Significant at less than the 1% level. 79 TABLE 19 Analysis of Variance Repeated Measures Trials 9 Through ll Source Sum of Squares d.f. Mean Square F Groups 95666.98 2 47833.49 ll.l2** Between Subjs. 451607.54 105 '4301.02 in same group Total between 547274.52 107 subjects Trials 393.94 2 196.97 1.63 Interaction: trials x 1943.85 4 485.96 4.01** groups Interaction: pooled subjects 25434.21 210 121.12 X trials Total within 27772.00 216 subjects Total 575046.52 323 ** Significant at less than the 1% level. 80