This is a debut. The decision to run a column in TAP that is devoted to psychoanalytic research springs from a realization that although fascinating work is being done, those who conduct it labor for the most part in relative obscurity, struggling against the common misapprehension that their work is far removed from the daily concerns of the clinician. For a long time, members of The American have been asked to contribute annually to the Fund for Psychoanalytic Research, yet very few know much about the projects these funds support. To rectify this situation, and convince readers of the relevance and importance of this work, The Research Corner will in each issue of TAP feature the work of one investigator sponsored by the Fund.
In this issue we feature the work of Dr. Enrico Jones of Berkeley, who recently received a grant from the Fund to study process and outcome in the intensive psychotherapy of depressed patients. The current study, which builds upon a preexisting research program, the Berkeley Psychotherapy Research Project, uses several intensive case studies in an attempt to model causal relations between process and outcome with these patients. Individuals meeting Research Diagnostic Criteria for major depression are seen in twice-weekly psychotherapy for two years; all sessions are videotaped, and periodic assessments are made by the therapist, the patient, and an external evaluator. These assessments constitute the data base.
In keeping with current requirements set by both intramural and external granting agencies, Jones's project stays abreast of the latest technologies in psychotherapy research. Following recent mandates by NIMH that no funding for long-term psychotherapy treatments, or any verbal psychotherapy, will be forthcoming without some sort of descriptive or operational manual, Jones has created a manual, to be used by participating therapists, that operationalizes the psychoanalytically informed treatment of depressed patients. A few manuals depicting the psychodynamic treatment of particular populations do in fact already exist, having been conceived by the research groups surrounding Lester Luborsky in Philadelphia and Otto Kernberg at New York Hospital in White Plains. Manualization remains, however, a very controversial and hotly debated topic among psychotherapy researchers. On the one hand, it is argued, manuals limit the freedom of the therapist to respond. On the other, when a therapy is manualized, it becomes possible to evaluate consistency between treatments, allowing a researcher to say that one treatment is comparable to another. The researcher can then go on to investigate a key question, namely, whether the treatment being delivered is indeed the treatment that is said to be delivered. The capacity to gauge this simple fact goes a long way toward legitimizing research in intensive psychotherapy. The more intensive the psychotherapy, however, the more difficult it is to manualize, and therefore to research using empirically sound methods.
Another controversial issue in psychotherapy research concerns the relation between process and outcome measures and the nature of the treatment. Twenty years ago Hans Strupp articulated the principle of “P-T-O congruence,” by which he meant that there should be an interwoven theoretical network linking problem, treatment, and outcome variables. In psychoanalytic treatments, then, one looks for measures that lend themselves to psychoanalytic processes. The problem is that very few such measures have been developed. Where previous research projects have opted for measures such as the Rorschach, Jones has chosen another route, using a mix of measures which, though not necessarily linked theoretically to clinical psychoanalytic processes, are demonstrably reliable and valid across a number of settings. Thus, he employs therapist ratings (on instruments such as the Inventory of Interpersonal Problems), patient self-report ratings (on instruments such as the Beck Depression Scale and the Minnesota Multiphasic Personality Inventory), and an external clinical evaluator's ratings (on instruments such as the Schedule for Affective Disorders and Schizophrenia and the Hamilton Rating Scale for Depression). In addition, several measures are used to assess more qualitative characteristics of the therapies. The advantage of these measures is that the instruments are psychometrically sound. The potential disadvantage is that they will not map well onto change in psychoanalysis as it is usually conceived. For example, since the deepening of a transference neurosis suggests that change does not occur in a linear fashion in analysis, it might be expected that in some cases an increased experience of depression or anxiety is the sign of something going right rather than wrong.
Another problem that has scuttled several projects concerns the competence of the therapist. Several previous projects have used psychiatry residents, psychology trainees, or volunteers side by side with more experienced therapists to conduct both paid and unpaid therapies, and then treated all the cases as if they were equivalent. In short, therapies were handled as if they were comparable to a drug dosage. Again, extramural granting agencies are clear and firm; long-term psychotherapies that are to be funded must use senior, well-trained clinicians who represent the best the field has to offer. In line with this requirement, Jones has elected to use only experienced clinicians in largely full-time practice as the therapists in his study.
Two aspects of the project are worthy of special note. The first is the use of a technique Jones has introduced called the Q-sort method. The other is the use of time series analysis. The Q-sort technique was adapted by Jones for psychotherapy research from a method pioneered at UC Berkeley by Jack Block. Fascinating, creative, and very complicated, it is well described in a paper published by Jones in the Journal of the American Psychoanalytic Association in 1990, entitled “The Psychoanalytic Case Study: Toward a Method of Systematic Inquiry.” With respect to time series analysis of single cases, there is a widely shared view that standard between-group research designs simply will not suffice in psychotherapy research, as they fail to provide the necessary methodological leverage for answering the questions typically asked by researchers of clinical psychoanalysis. New methods are necessary, and are being developed all the time. Time series analysis lends itself well to the processes of therapy because it permits the modeling of fluctuations in process throughout the course of a treatment. It also allows the researcher to identify causal relationships over long periods of time and to study bidirectional influences. Previous studies have segmented therapy hours into equivalent sections, compared them to one another, and demonstrated causal relations through demonstrating a time-bound association. By contrast, time series analysis allows the assessment of process and outcome change over long periods.
There is much more to be said. Research in psychoanalysis, demonstrations of efficacy as well as investigations of complicated microprocesses, are becoming increasingly important. This does not at all diminish the significance of clinical psychoanalysis or its ways of posing and answering questions. There are many ways to deepen our understanding of what it is we do. The philosopher Imre Lakatos once noted that in the modern world no single method stands alone as science. Depending on the questions asked, there is room in the house of science for a variety of methods. Case studies, clinical/theoretical reports, and empirical investigations can all aid us in our quest for a deepening understanding of the human psyche.