Certification is widespread in the professions, particularly in medicine and the medical specialties. The certifying body in American psychiatry is the American Board of Psychiatry and Neurology. I was a director of that board from 1982 to 1989 and its president in 1988. My comments on certification in the APsaA are informed by this experience.
The primary purpose of professional certification is to provide assurance to the public by (1) identifying those who can demonstrate at least minimal competence and (2) distinguishing them from those who cannot. Much of the recent discussion of certification in psychoanalysis seems to ignore this principle. Certification is not primarily for the members of the profession or for its internal political concerns; it is to assure the public.
Further, it is not particularly helpful in raising the quality of average practitioners. (Professional education is central here. Examinations during training may enrich education, while recertification may encourage continuing education.) Nor is it helpful in improving the very best practice; this requires scholarly inquiry, not certification. Certification assures minimal competence.
Certification is also not particularly helpful in assessing the quality of educational programs. The performance of graduates, often some years after training, is shaped by many factors other than the training program, and the information obtained in the certification process is generally too little, too late, and too poorly designed to evaluate the program.
Reliability refers to the consistency and reproducibility of the results, regardless of the circumstances or the examiners. It is essential for validity and for fairness. A profession should evaluate the reliability of its certification procedures, share the results with its members and the public, and strive to improve those procedures. There are many ways to do this. For example, different evaluators could blindly rate the same applications and compare their conclusions.
Validity refers to the relevance of the evaluation process to the real thing in which we are interested — the “gold standard,” in this case clinical competence. It is easy to get reliability without validity; measures of height or weight would work this way (plenty of reliability, but no relevance to what we're trying to measure). However, there can be no validity without reliability.
At times the quest for ultimate validity so interferes with reliability that it endangers the entire process. This is a particular problem in a field like psychoanalysis, where so much in which we are interested is subjective, imprecise, and unquantifiable. Mixed examination methods — some of higher reliability, some of lower reliability but higher face validity — can sometimes be helpful. For example, the American Board of Psychiatry and Neurology employs a mixture of multiple-choice questions, oral examinations based on videotaped interviews, and oral examinations based on interviews of patients, with all oral examinations involving simultaneous independent evaluators. This means that methods with high reliability and marginal validity are mixed with strategies of lower reliability but presumptively higher validity.
The certification process evaluates an individual at one point in time. Individuals change over time, and knowledge and practice in the field change as well. As a result, all medical certification is time-limited and is linked to processes for recertification or the maintenance of certification. For example, certification by the American Board of Psychiatry and Neurology is limited to ten years.
One could argue that psychoanalysis changes more slowly than psychiatry, and perhaps even that the personal capacities and characteristics essential for clinical competence in psychoanalysis decay more slowly, but it is hard to imagine that the public would feel much assurance about a psychoanalytic practitioner who completed certification twenty or thirty years earlier. If psychoanalysis is to embrace certification, it must also embrace recertification or explain how it is different from all of the professions that do so.
The addition of a recertification procedure would have an interesting effect on the politics of certification. Every analyst, certified or not, would be a candidate for future certification. Roles in the profession that required certification would logically require recertification and the continued maintenance of certification. Anonymity in the procedure would allow valuable studies of validity, comparing the performance of first-time and repeat applicants. Finally, the exposure of both senior analysts and younger members of the profession to the procedure would assure close attention to the sensitivity and the comfort of the candidates.
It is possible that certified members would refuse to participate in recertification. However, this would raise serious questions about the public assurance mission of certification and the appropriateness of using it as a requirement for various professional roles.
Negative responses to examinations and evaluation procedures are common. Such procedures also offer opportunities for personal and institutional enactments that may interfere with the goal and provoke such negative responses. In other words, examinations are projective tests for examiners, and examiners' sadism may be expressed during the tests. It is important to protect against these enactments, both out of respect for colleagues and because of the potential long-term negative effects on the profession.
Some simple principles have been developed by all certifying professions:
The purpose of certification should be explicit and clear.
The evaluation procedure should only evaluate, not attempt to teach or demonstrate. Admixtures of teaching in the absence of an educational alliance are inappropriate.
The procedures and criteria should be transparent. Safeguards should ensure that information about the candidate's identity, background, or training doesn't contaminate the results.
Robert Michels, M.D., is Walsh McDermott University Professor of medicine and psychiatry at Cornell University. He is a training and supervising psychoanalyst at the Columbia University Center for Psychoanalytic Training and Research.
It is important to recognize that there will be candidates who fail (otherwise the process has no value) and that some who fail will be critical of the process.
Certification is a vital activity for any profession that serves the public. It requires time and attention, and it represents an ongoing burden and responsibility for the members of the profession. The American Psychoanalytic Association has developed a certification procedure, but major problems of reliability, validity, recertification, and the examination process have not been solved.
Discontinuing certification would raise questions about our claim to be a profession. Continuing it without change seems unsatisfactory. Improving it will increase the burdens on the profession. We must choose.