What are the Data?

I experienced an old, familiar sinking feeling as a member of the audience directed a question to a panelist at a recent meeting of the American Psychoanalytic Association. A seasoned psychoanalyst had presented his position that analysis, rather than psychotherapy, was the treatment of choice for most patients he sees in his practice. The analyst said he would treat patients in face-to-face psychotherapy with an agreement that they strive to understand why the most effective treatment, psychoanalysis, was not chosen.

When pressed about his convictions regarding psychoanalysis, the analyst said, “It saved my life.” I admired his courage, his willingness to refer to his personal experience in the context of a professional meeting. I was quietly reviewing personal benefits of my analysis when I was alerted by the question, “What are the data?”

At face value, it's a reasonable question that I had heard, and asked, innumerable times when I worked as an experimental psychologist. But, if my memory is accurate, this was the first time I had heard it asked at an APsaA meeting. The audience member pressed the analyst about his approach and, if I recall correctly, got him to admit that there were some persons seeking his help for whom he would not recommend psychoanalysis. Next, he asked the analyst to articulate “the data” upon which he would base his decision to recommend psychotherapy instead of psychoanalysis. Eventually, the analyst responded, “I plead ignorance.” The interrogator returned to his seat.

I felt the return of an oppressiveness that had been one reason for my departure from experimental psychology, an area where all declarative sentences must be backed by reams of data, where data are the armaments of theoretical competitors. I imagined that the interrogated analyst was thrown by the word “data” and that he would have been able to articulate his experience contributing to his decision to recommend psychotherapy to a particular person, had he thought of his experience as data rather than construing data as tables and graphs of quantified symptoms, behaviors, and demographic characteristics of persons seeking treatment.

Scientific Status

Experimental psychologists probably know more about construction of data than adherents of other academic and professional disciplines. This is so because of psychology's struggles to become recognized as scientific, to cast off its historical ties to mentalism, and to gain legitimacy and status in a society placing increasing value on technological progress. Experimental methods and statistics are required entry-level courses for undergraduates majoring in psychology. When I taught the courses, students majoring in other areas of “natural science” were encouraged to take them. They learned how to operationally define concepts, observe and measure phenomena, design and carry out correlative studies and experimental investigations, display data, apply and interpret various statistical measures, and present their results most convincingly.

I felt the return of an oppressiveness that had been one reason for my departure from experimental psychology, an area where all declarative sentences must be backed by reams of data, where data are the armaments of theoretical competitors.

Students were taught that experience per se was an unreliable source of information, that introspection had failed as a method of inquiry. They were taught to focus on subject matter that was observable via sensory organs and capable of meeting requirements of interob-server agreement and measurability. Only through this methodology could psychological science be advanced.

It has not been determined whether these methods of generating data will help illuminate what is of most importance to psychoanalysis—the influence of unconscious mentation on experience and action as gleaned from analysands' descriptions of their experience of derivatives of unconscious mentation during sustained and disciplined introspective inquiry. One problem in applying the methods is observational. Since no analyst can sensorily observe an analysand's experience or unconscious mentation, the requirement of inter-observer agreement on the occurrence of instances of the subject matter, as specified by experimentalists, cannot be attained.

One approach to this problem is to work with data that can be observed sensorily, taking the analysand's verbal descriptions as the basic unit of observation. Words can be perceived auditorily and subjected to measures of inter-observer agreement about their occurrence, allowing their use as variables in correlational and experimental studies. The workings of unconscious mentation may be inferred from patterns of word usage, linguistic probes of semantics, and other systematic analyses.

Another approach posits an additional operation in analytic observation, and is usually referred to as empathy. This approach is two-pronged: (1) attending sensorily to analysands' descriptions of their experience, and (2) attending to our own experience of derivatives

of our own unconscious mentation considered to be resonant to the unconscious mentation of the analysand's. In this schema of analytic observation, words are considered to serve multiple functions: to describe an analysand's experience, and to evoke resonant derivatives of unconscious mentation of the analyst, which may be considered a form of data gathering unique to psychoanalysis, some would say the wellspring of psychoanalytic data. It remains to be determined whether data obtained using these operations of observation can meet requirements of interob-server agreement. I think requisite studies could be designed and implemented, but I would not recommend suspending psychoanalytic inquiry until the results were in.

Focus on Mentation

If we maintain a focus on unconscious mentation as our basic subject matter, we must find ways to articulate more clearly, and to systematically assess, the nature of the data with which we work. The phenomena of interest should determine the methods of investigation, not vice versa. From this perspective, “What are the data?” may be one of the most important questions with which we as analysts must grapple.

A concept that was entertained years ago when I was working as an experimentalist was called ecological validity. I don't know whether it was ever developed systematically into a measure of validity, but it made good sense. Does the experiment pertain to phenomena as they exist outside the rarefied confines of the laboratory or the research protocol? When I was an experimentalist, many phenomena of interest were so extensively modified by adapting aspects of them to fit experimental methodology that even the most sophisticated of investigations bore little resemblance, and I will add relevance, to the topic being investigated. I propose that if a set of data seems irrelevant to how you work in your practice, entertain the possibility that it is.

Getting back to my reaction at the panel, the question, “What are the data?”, can be used for purposes other than promoting inquiry. Ironically, the question can be asked in the service of stifling inquiry, refusing to accept the described experience of others as worthy of consideration if they have no graphs or tables of quantified data to bolster their statements. This usage of the question may be especially detrimental to psychoanalysis. If indeed the analyst's unconscious resonance to that of the analysand is a hallmark of psychoanalytic observation, the described experience of the analyst may be our most direct route of access to the data. From this perspective, it is imperative that analysts strive to articulate as clearly as possible their experience of working analytically

Data are constructed as approximate representations of aspects of phenomena being investigated. Data per se should not be granted the status of idealized substitutes for the phenomena they represent. It would be a mistake to replace the worn out disclaimer “That's not analytic!” with “What are the data?”