Contents
This article presents some of the most important aspects of a large-scale study conducted at the University of Chicago Counseling Center in 1950-1954. It was made possible by the generous support of the Rockefeller Foundation and its Division of Medical Sciences. In 1954, I was invited to give a talk at the 5th International Congress on Mental Health in Toronto, in which I decided to describe parts of the research program. A month after this report, the University of Chicago Press published our book, which described the entire program. Although most of the work on the book was done by Rosalind Diamond and myself, as managing editors and authors of separate parts, the other authors of the collection also deserve thanks for their contributions. They are John M. Butler, Desmond Cutwright, Thomas Gordon, Donald L. Grammon, Gerald W. Haig, Eve S. John, Essilyn K. Rudikoff, Julius Seaman, Rolland R. Tages and Manuel J. Vages.
The particular reason this work is included in this book is that it provides a partial representation of the exciting progress we have made in measuring this changing, obscure, highly significant, and defining aspect of the personality that the individual calls «the self.» .
* * *
The purpose of this paper is to present some features of the experience of my and my colleagues, when, with the help of objective scientific methods, we were able to evaluate the results of one of the types of individual psychotherapy. To help you understand these features, I will briefly describe the conditions under which the present study was conducted.
For many years I have worked with my fellow psychologists in the field of psychotherapy. In our experience of conducting psychotherapy, we tried to find more effective methods for achieving constructive changes in the personality and behavior of a person who turned to us for help, poorly adapted to life or with a disturbed psyche. Gradually, an approach to psychotherapy based on this experience was formulated. It has been called variously — non-directive or customer-centered. This approach and its theoretical justification have been described in many of my books ([1], [2], [34], [5], 6]) and numerous articles.
One of our constant goals has been to subject the development of psychotherapy and its results to rigorous scientific scrutiny. We believe that psychotherapy is a deeply subjective direct experience experienced by both the therapist and the client. This experience is filled with complexities, subtleties and nuances of personal relationships. However, we are convinced that if this experience is so significant, if deep acquaintance with it causes changes in personality, then these changes should be amenable to scientific research.
Over the past 14 years, we have done a lot of research on both the process and the results of this form of psychotherapy.7. In order to shed light on the results of this type of psychotherapy, over the past five years we have expanded our research by conducting a number of related studies. I want to present some salient features of this particular research program now under way.
Three aspects of our research
In my opinion, the following three aspects of our research would be of great importance to our readers:
- The criterion used by us in the study of psychotherapy, which does not coincide with the traditional opinion about the criteria in this area.
- A study design that overcomes some of the difficulties that previously stood in the way of obtaining accurate results.
- Our progress in the objective measurement of elusive subjective phenomena.
These three elements of our program could be used in any attempt to measure personal change. Therefore, they are applicable to any kind of psychotherapy or to the study of any method used to change personality or behavior.
Let’s now return to these three elements that I mentioned and consider them one by one in more detail.
Research Criteria
So, what serves as a criterion in psychotherapy research? This question causes many difficulties, which we encountered primarily in the planning of the work. It is widely believed that the purpose of research in this area is to measure the «success rate» of psychotherapy, or the «rate of cure achieved.» Although we have not escaped the influence of this opinion, after careful consideration, we have abandoned these concepts, since they are indefinable, are essentially value judgments, and therefore cannot be part of the science in this area. There is no general agreement on what “success” means—whether it is symptomatic relief, conflict resolution, improved social behavior, or other changes. The concept of «cure» is completely unacceptable, since in most of these disorders we are not dealing with a disease, but with a behavior developed as a result of learning.
As a result of these considerations, we did not ask in our study, “Was psychotherapy successful? Was the client cured? Instead, we asked a question that is more scientifically justified, namely: “What are the circumstances that accompany psychotherapy? »
To get a basis for answering this question, we turned to the theory of psychotherapy that we are developing and took from there a theoretical description of the changes that probably occurred during psychotherapy. The aim of our study was to determine whether the changes we predicted could be measured. Thus, based on the theory of psychotherapy, centered on the client, we put forward the following hypotheses. During psychotherapy, feelings that were previously not allowed into consciousness are experienced and included in ideas about oneself, in the “I-concept”. During psychotherapy, ideas about oneself approach the ideas about the “ideal self”. During and after psychotherapy, the observed behavior of the client becomes more socialized and mature. During and after the psychotherapy process, the client begins to more accept himself, his «I», and this correlates with an increase in acceptance of others.
Here are a few hypotheses that we were able to explore. It is probably clear that in our research we have completely abandoned the idea of one main criterion, and instead used many well-defined variables, each of which was precisely formulated for the hypothesis under study. This means that we hoped that the study would be able to express our findings something like this: client-centered psychotherapy leads to measurable changes, for example, variables a, b, d, and f, but does not cause changes in variables c and e. When such statements obtained, both the specialist and the non-specialist will be able to make a value judgment about whether they consider the process that brought about these changes to be successful or not. Such value judgments, however, will not change the hard facts in our slowly increasing scientific knowledge of the effective drivers of personality change.
Thus, in our study, instead of the usual global indicator of «success», many special variables are used as criteria, each of which is operationally defined and derived from our theory of psychotherapy.
This solution to the problem of evaluation criteria helped a lot in choosing the means of research wisely for conducting a series of tests. We were not looking for an answer to a question that cannot be answered: by what means can success or cure be measured. Instead, we asked specific questions related to each hypothesis. What means can be used to measure the «I-concept» of an individual? By what means can the maturity of behavior be satisfactorily measured? How can you measure acceptance of yourself and others? Despite the difficulty of these questions, it is possible to find operational answers to them. Thus, our decision regarding the criterion of evaluation helped us a lot in solving the whole problem of means of investigation.
Research Design
Many thinking psychologists have noted the fact that there is no objective evidence that psychotherapy causes constructive personality change. Hebb8 argued that «there is no evidence to show that psychotherapy is useful» [9, with. 271]. After reviewing several works at his disposal, Eysenck10 pointed out that the data «do not prove that Freud’s psychotherapy or any other therapy hastens the recovery of neurotic patients» [11, pp. 322].
With this situation in mind, we were keen to design our study rigorously enough to establish two points when confirming or rejecting our hypothesis: 1) whether there was a significant change; 2) if it happened, whether it was due to psychotherapy, and not some other factor. In such a complex area of research as psychotherapy, it is not easy to create a research design that will achieve these goals, but we believe that we have achieved genuine success in this direction.
Having chosen the hypothesis to test and the means most suitable for operational measurements, we were ready for the next step. A selected set of objective research tools were used to measure various characteristics of the client group before psychotherapy, immediately after its completion, and 6-12 months later (see Fig. 1). A medium-sized group of twenty-five specially selected people, like those who come to the University of Chicago Counseling Center, was formed. The study of a medium-sized group allowed for a more detailed analysis. With a larger number of subjects, the analysis would be more superficial.
Ris.1
Study scheme
Part of the group, which then carried out psychotherapy, was chosen as the control group. This group was subjected to exploratory testing. Then a two-month break was made, which served as a control period. After that, clients were tested again before psychotherapy. A reasonable explanation for this action is that if changes occur in individuals only because they are motivated by psychotherapy, or because they possess a certain type of personality structure, then such a change must occur during this control waiting period.
The other part of the non-psychotherapy group served as an equivalent control group. In terms of age indicators, socioeconomic status, the ratio of men and women, students and non-students, both groups were equivalent. At appropriate time intervals, this group was subjected to the same testing as the control group participating in psychotherapy. Part of the second group was tested four times to be strictly comparable with the control group participating in psychotherapy. The rationale for creating this equivalent control group is that if changes in individuals are due to exposure to time, random variables, or as an unnecessary consequence of repeated testing, then such changes will show up in the results of that group.
The general logic of this dual control design was that if changes occurred in the psychotherapy group during and after psychotherapy that were significantly greater than those experienced during the waiting period in the psychotherapy control group or in the equivalent control group, then it was reasonable to attribute these changes to the influence of psychotherapy.
I cannot sum up in a nutshell the complex and ramified details of the various studies that have been carried out in these experiments. A more comprehensive report [7] describes 13 projects completed to date. Suffice it to say that complete data were obtained from 29 clients with whom 16 psychotherapists worked, as well as complete data from the control group. Careful evaluation of the research data enables us to come to the following conclusions: during and after psychotherapy, the client undergoes profound changes in self-perception; constructive changes are taking place in the qualities and structure of the client’s personality — the qualities of his personality are approaching the characteristics of a fully functioning person; there are changes in the direction of increasing personal integrity and adaptation; the client’s friends report changes in his behavior towards maturity. In each individual case, it was found that the changes that occurred were more striking than those observed in the control group or in the same clients in the control «waiting period». Only in relation to the hypothesis put forward in connection with the acceptance of other people and a democratic attitude towards them, the data obtained are ambiguous and unclear.
In our opinion, the research program just completed is sufficient to change the claims made by Hebb and Eysenck. At least in relation to psychotherapy, there is now objective evidence of positive changes in the client’s personality and behavior in directions that are usually considered constructive, and these changes are due to psychotherapy. It is the adoption of specific complex research criteria and the use of a highly controlled research design that enables these claims to be made.
Measuring changes in «I»
Since I can only present a very small sample of results, I will select from the area in which the greatest advance in research methods has been made and where the most intriguing results have been observed. This area includes our attempts to measure changes in the client’s self-perception (self) and the relationship of self-perception to some other variables.
To obtain an objective measure of the client’s perception of themselves, we used the recently proposed Q-method developed by Stephenson [12]. From conversation recordings and other sources, a large set of statements has been generated by which clients describe themselves. Here are some typical statements: “I am a submissive person”, “I do not trust my emotions”, “I am calm and nothing bothers me”, “I am afraid of sex”, “I usually like people”, “I am an attractive person” , “I worry that other people will think badly of me,” etc. A random sample of one hundred of these redacted statements was used as a research tool. In theory, we had a sample of all the ways in which a person can perceive himself. These statements were reprinted on cards. The client was asked to sort the cards into nine approximately equal parts, starting with those cards on which the written statements best matched his personality at the moment, and ending with those that are less characteristic of her. The client did this before psychotherapy, after, in the subsequent period, and also several times during psychotherapy itself. Along with this, clients were asked to sort the cards in such a way as to imagine the “ideal self”, which he would like to be.
As a result, we received detailed and objective data on the client’s perception of himself at different points in time and on his ideas about the «ideal self». The correlation coefficients between these different sorts were then determined. High correlation between two classifications showed similarity between them or no change, low correlation — difference between them or significant changes.
To show how this method has been used to test some of our self-related hypotheses, let’s see how some of the data from a single client study relates to several hypotheses [of 7, ch. fifteen]. I think this will better show the intriguing nature of the results than if I presented the general conclusions from our study, although I will try to mention these general results in passing.
The client, from whose data I will take the material, was a very unhappy forty-year-old woman in marriage. She felt guilty about her daughter’s nervous breakdown in her teens. The woman was quite deeply disturbed, and, according to the diagnostic scale, she was considered seriously neurotic. She was not part of her own control group, so she entered psychotherapy immediately after the first test. This woman came to talks 40 times within 5,5 months, after which she completed psychotherapy. Follow-up tests were done 7 months later, at which time she decided to come back for 8 more interviews. The second study was conducted 5 months later. The consultant decided that a very significant advance had been made in the process of psychotherapy.
On fig. Figure 2 presents some data related to the change in the client’s perception of himself. Each circle represents sorting data for certain qualities of the «ideal self» or «real self». Sorting was carried out before psychotherapy, after the 7th and 25th interviews, at the end of psychotherapy and 6 and 12 months after psychotherapy. Correlations between many of these sorts are given.
Ris.2
Changing the relationship between the «real self» and the «ideal self»
Now let’s consider this data in relation to one of the hypotheses we were interested in testing, namely the hypothesis that the «I» perceived by the client changed to a greater extent during psychotherapy than during the time when psychotherapy was not carried out. In this case, the change was greater during psychotherapy (r = 0,39) than during any of the subsequent periods (r = 0,74, 0,70) or in the subsequent 12 months (r = 0,65). Thus, in this case, the hypothesis was confirmed. In this respect, the woman’s test results were typical of other clients; the overall result was that changes in self-perception during psychotherapy were significantly greater than during the control or follow-up periods, and significantly greater than changes occurring in the control group.
Let’s consider the second hypothesis. It has been predicted that during and after psychotherapy, the perceived «I» is evaluated more positively, that is, more in line with the «ideal self» or highly valued «I».
The client shows a significant discrepancy between «real self» and «ideal self» on the first test (r = 0,2 1). During and after psychotherapy, this discrepancy decreases, and the high degree of agreement shown in the final follow-up study (r = 0,7) thus confirms our hypothesis. This is typical of all major outcomes, which for the group as a whole show a significant increase in fit between themselves and the ideal during psychotherapy.
A careful study of Fig. 2 shows that by the end of the study, the client in her perception comes very close to the person she would like to be when she came to psychotherapy (rIB.SF2 = 0,70). It can be noted that, compared to her initial ideal, her final ideal has become a little closer to her initial self-image (rSB.IF2 = 0,36).
Let me briefly consider another hypothesis, that changes in the perceived self will not be random, but will occur in a direction that expert judges would call «adjustment.»
As one of the goals of our study, Q-sort cards were presented to a group of clinical psychologists not associated with the study, and they were asked to sort the cards as a well-adjusted person would sort them. This gave us a sorting criterion against which to compare the perception of any client. A simple scoring score was developed to express the degree of similarity between the client’s self-perception and this idea of an «adjusted» person. It was called the «fitness score», with a higher score indicating a higher degree of fitness.
In the case of a 2-year-old woman, we considered fitness scores for six successive sorts in relation to the perception of self shown in Fig. 35, starting with the perception of the self before psychotherapy and ending with the perception in periods after psychotherapy. They were as follows: 44, 41, 52, 54, 51, XNUMX. Based on the operational definition of fitness, there is an obvious trend towards better fitness. This is also true for the group as a whole, with a marked increase in fitness scores occurring during therapy and a very slight regression in scores thereafter. In control individuals, in essence, no changes were observed. Thus, both for this woman and for the group as a whole, our hypothesis was confirmed.
When a qualitative analysis of various sorts was carried out, a result was revealed that further confirms this hypothesis. When the initial self-image is compared with that after psychotherapy, it is found that after therapy it changes in various ways. The client feels that he has become more self-confident, can rely more on himself, understands himself better, feels calm, his relationships with other people have improved. She feels less guilty, less touchy, less overworked and insecure, less secretive. These qualitative changes observed in the woman we observed are similar to those found in other clients and are generally consistent with client-centered psychotherapy theory.
I would like to highlight some additional interesting results, which are shown in Fig. 2.
Obviously, the idea of the «ideal self» is much more stable than the idea of the «real self». All correlations between them are above 0,70, and the idea of the person the client would like to be changes relatively little over the entire period. This is observed in almost all clients. Although we did not formulate hypotheses about this, we expected that some clients would achieve a better fit between their «real self» and «ideal self», mainly due to changes in their values, and others due to a change in self-image. So far, our data show that this is not true, that, with a few exceptions, it is the «real self» that changes the most.
The client described in this case is indeed undergoing some changes in his «ideal self». Interesting direction. If we compute the previously described fitness score for this client’s successive ideas of «ideal self», we find that the mean score in the first three dimensions is 57, and the mean score in the post-treatment dimensions is 51. In other words, the «ideal self» becomes less ideally fitted or more achievable. To some extent, this is already a more realistic goal. This customer’s data in this regard also reflects the trend of the whole group.
Another result relates to the «remembered self» (see Fig. 2). This sorting was obtained when the woman was asked, during a second, follow-up study, to re-sort the cards in the same way as she did at the beginning of psychotherapy. This «remembered self» was very different from the «I» given at the beginning of entry into psychotherapy. The correlation with the «real self» given at this time was only 0,44. Further, it was a much less favorable self, much more different from the «ideal self» (r = 2,1). It had a low fitness score of 26 compared to an original self score of 35. This suggests that this triage of remembered self-images approximates the objective change in the decrease in defensiveness that occurred over the 18 months of our study. At the last meeting, the woman was able to give a much more truthful portrayal of the ill-adjusted, disturbed person she was at the beginning of psychotherapy. This mapping, as we shall see, is supported by other evidence. Thus, the degree of change in the «I» over the entire period of one and a half years is perhaps better represented by a correlation of -0,13 between the «remembered self» and the final «self» than by a correlation of 0,30 between the initial and final «self». «.
Let me now turn to another hypothesis. Client-centered theory of psychotherapy holds that, in the psychologically safe atmosphere of the psychotherapeutic relationship, the client can allow himself to be aware of his experience, which he usually suppressed or rejected. This previously rejected experience is now becoming part of the self-image. For example, a client who has previously suppressed all feelings of hostility may begin to freely feel his hostility. Then his self-image is reorganized to include the realization that he sometimes has hostile feelings towards other people. Accordingly, his ideas about himself (his «I-concept») become more true, representing his entire experience as a whole.
We have tried to transform this part of our theory into an operational hypothesis, which was formulated as follows: during and after psychotherapy, there will be an increasing correspondence between the self-concept of the client and the diagnostician. The assumption is that an experienced person who makes a psychological diagnosis of a client, more than the client himself, is aware of the totality of his experience — both conscious and unconscious. Hence, if the client assimilates in his conscious self-image many of the feelings and experiences that he previously suppressed, his self-image should approach the diagnostic psychologist’s.
The method of investigating this hypothesis was to take a projective test (TAT13) that was used by the client in each measurement period, and allow the diagnostician to analyze these four tests. To avoid any warnings, the diagnostician was not told about the order in which the tests were to be applied. He was then asked to sort the Q-cards during each of these tests to get an idea of the client’s diagnosis at that time. This process gave us an unbiased diagnostic assessment, expressed in terms of the same method that the client used to express his «I». Thus, the possibility of direct and objective comparison through correlations of different Q-sorts was obtained.
The result of the study for this particular client is given in Fig. 3. The upper part of the diagram is a summary of the information from fig. 2. The bottom row shows the sorting data made by the diagnostician; correlations give us the opportunity to test our hypothesis. It should be noted that at the beginning of psychotherapy there is no connection between the client’s self-image and the diagnostician’s self-image (r = 0,00). Even at the end of psychotherapy, the situation remains the same (r = 0,05). But by the time of the first measurement after psychotherapy (not shown in the figure) and the second, subsequent measurement, the client’s idea of himself became much closer to the diagnostician’s idea of him (the first measurement after psychotherapy r = 0,56, the second r = 0,55). Thus, the hypothesis is clearly confirmed: the correspondence between the client’s idea of himself and how he is perceived by the diagnostician has increased significantly.
Ris.3
The relationship between the «real self», «ideal self» and the idea of a diagnostic psychologist’s client.
Numbers indicate correlation coefficients;
commas separating whole numbers from tenths and hundredths are omitted.
There are other results of this aspect of the study that are of interest. I would like to note that at the time of the beginning of psychotherapy, the diagnostician’s perception of the client is very different from the «ideal self» of the client (r = -0,42). By the end of the study, the diagnostician sees that the client is very close to his ideal at this time (r = 0,46) and even more similar to the “ideal self” that he had at the time of entry into psychotherapy (r = 0,61) . Thus, we can say that objective evidence shows that the client, in his perception of himself and in his general idea of his personality, basically becomes the person he wanted to be at the beginning of psychotherapy.
Another noteworthy problem is that changes in the diagnostician’s perception of the client’s self are significantly larger than changes in the client’s self-perception (r = -0,33 — compare with r = 0,30). This fact is interesting from the point of view of the generally accepted opinion of specialists that clients overestimate the degree of changes that have occurred in them. It is suggested that it is possible that an individual can change so much in 18 months that at the end his personality becomes more dissimilar than similar to his personality at the beginning.
The last remark to fig. 3 refers to the «remembered self». It should be noted that this «remembered self» is positively correlated with the diagnostician’s representation (r = 0,30), thus confirming the previous statement that it is more true and less associated with defensive reactions of self-image than that which the client could give at the beginning of psychotherapy.
Conclusion
In this article, I have tried to present at least the main outlines of the extensive psychotherapy research currently under way at the University of Chicago. Several salient features were noted.
The first feature is the rejection of a global criterion in the study of psychotherapy and the adoption of a special, operationally defined criterion of change. It is based on a detailed hypothesis that grows out of the theory of the dynamics of psychotherapy. The use of many specific criteria has enabled us to make scientific progress in identifying the types of changes that are or are not associated with client-centered psychotherapy.
The second characteristic feature is a new approach to the hitherto unresolved problem of control in psychotherapy research. The design of the study included two types of control: 1) an equivalent control group, explaining the effect of time, retesting, random variables; 2) own control group, in which each client participating in psychotherapy is compared with himself in the absence of psychotherapy. This group explains the influence of personality and motivation variables. The use of a dual control scheme made it possible to conclude that changes during psychotherapy that are not explained by the controlled variables occur due to the influence of psychotherapy itself.
The third characteristic feature of this work was that by conducting a rigorous scientific study, the development of subtle elements of the subjective world of the client was shown. The following developmental evidence was presented: a change in the client’s self-concept; the degree to which the «I-real» becomes similar to the «I-ideal»; the degree to which the self-image becomes more comforting and accommodating; indicators of the correspondence of the client’s perception of himself to the client’s perception of the psychologist-diagnostician. These results tend to confirm the theoretical claims that have been made about the place of the self-concept in the dynamic process of psychotherapy.
In conclusion, I want to present you with two conclusions. The first is that the research program I have described probably makes it clear that objective evidence, inherent in the canons of exact scientific research, can be obtained for changes in personality and behavior as a result of psychotherapy. They were obtained for one type of psychotherapy. This means that in the future, similar reliable evidence may be obtained regarding whether personality changes occur as a result of other types of psychotherapy.
The second conclusion, in my opinion, is even more important. Advances in research methods in recent years mean that many of the intricacies of the psychotherapeutic process are now widely open to scientific research. I have tried to show this by examining changes in the self-concept. But similar methods make it possible to objectively study the changing relationship between client and therapist, the study of «transference» and «countertransference» relationships, the changing locus of the client’s value system, and so on. I think it can be said that almost any theory that is supposed to be related to personality change or the process of psychotherapy is now open to scientific research. This opens up new horizons in scientific research. The search for this new road should shed a bright light on personality dynamics, especially on the process of personality change in interpersonal relationships.