Note: This paper was presented at a symposium, "Scientific Foundations of Clinical Psychology at the Beginning of the 21st Century -- Victories, Setbacks, and Challenges for the Future" sponsored by Division 12 (Clinical Psychology), at the annual meeting of the American Psychological Association, Honolulu, July 31, 2004.
Links to other papers that discuss clinical training issues will be found in the reference section.
I greatly appreciate the invitation to join this symposium on the future of scientific clinical psychology. Although I have never been on the core faculty of a clinical training program, I do have clinical training, and I have tried though my research and teaching to help build bridges between clinical psychology and the other subfields of the discipline, as well as between clinical psychology as a science and clinical psychology as a profession – bridges that should run both ways (Kihlstrom & Canter Kihlstrom, 1998).
Let me begin with a little personal history. My own clinical training was in the classic scientist-practitioner model, with a decided emphasis toward science and away from practice. I applied to Penn because I was interested in doing hypnosis research with Martin Orne, and because I was also interested in personality – I wrote on my personal statement that I wanted to quantify the concepts of existentialist theories of personality, leading Burt Rosner, the chair of Penn’s Psychology Department at the time, to tell me that they had accepted me just to see what I looked like – so I sent my application to the Program of Research Training in Personality and Experimental Psychopathology. I was very happy, the next spring, to get a thick envelope from Philadelphia – but also dismayed to find that Penn’s offer of admission was signed by Julius Wishner as head of the Clinical Training Program. I immediately called Julie, who ran both programs, and told him I didn’t want to be in the Clinical Training Program; in that charmingly gruff way that Julie had, he replied "Don’t worry: you’re not", and hung up the phone. Nevertheless, the students in that program did everything that clinical students did – some proseminars, a weekly research seminar, a course on assessment, a course on treatment, a little practicum. It was enough to prepare us for our internships. Having completed our dissertation research before leaving on our internships, we wrote our dissertations at night and on weekends, and then we were out.
It’s been said that everybody favors the training model in which they themselves were trained, and I suppose that I’m no exception. For all these years, I have been a staunch advocate of the scientist-practitioner model, and dismissive of such alternatives as the practitioner-scholar model. To be honest with you, I’m not even all that thrilled with the clinical scientist model, which is probably the closest to the version of the scientist-practitioner model implemented at Penn at that time I was there – not just because the very term "clinical science" strikes me as scientistic (Noam Chomsky once said that you know a science is in trouble when it has to call itself a science; he had political science in mind, but the criticism might apply equally well to cognitive science and neuroscience), but because it's uninformative about what the person actually does: the person isn't a clinical scientist -- he or she is a clinical psychologist. Also, the clinical-scientist model, like the practitioner-scholar model, seems to dig a ditch, rather than build a bridge, between science and practice. However, it now strikes me that the scientist-practitioner model has outlived its usefulness, and creates more problems than it solves. In my view, it is now time to train some students for science, and other students for practice – even if this training must take place in quite different programs, and even if it must take place in quite different institutions.
The reason for my change of heart is that I have become increasingly convinced that the traditional scientist-practitioner model works to the disadvantage of both types of students. For example, the student training for an academic career in teaching and research must spend precious time preparing for a career of clinical practice that he or she will never pursue, and in which he or she has no particular interest. It’s one thing for budding clinical researchers to take a couple of courses in assessment and treatment methods, see a couple of patients during practicum, and then get full-time exposure to "the living material of the field" during their internships before taking up posts in universities and medical centers. It’s another thing entirely for students who are really interested in clinical research to spend as much as one or even two thousand hours during their graduate studies acquiring and polishing practical skills to make themselves more attractive in the competition for internship placements.
|
|
|
|
|
|
At Berkeley, for example, students in our "clinical science" program must complete, in addition to a two-semester clinical proseminar (no harm in that), and one or two courses in methods and statistics (everybody does that), two or three courses in assessment, two courses in intervention, and the equivalent of more than six courses of practicum – not to mention three to four "breadth" courses covering the biological, cognitive-affective, social, and individual bases of behavior. That’s a total of 13 to 15 courses – almost two full years’ worth.
This doesn’t count an unspecified number of courses on diversity issues, and the requirement that each clinical student take a minor outside clinical. Nor does it take into account the fact that clinical "students... are encouraged to take as many as possible of the major courses and seminars offered by the core Clinical Science Faculty": at last count, there were approximately 15 such courses in the General Catalog. All this while their research-oriented peers in cognitive, social, and developmental psychology, having completed their proseminar and methods requirements in their first year or so of graduate study, are happily working in their advisors’ laboratories developing their individual research portfolios, and taking the occasional seminar offered by their advisor or someone in a closely related area of research. Research, including research-related courses and seminars, should have pride of place in the training of these research-oriented students. Instead, both research training and research itself are subordinated to the requirements of training for practice.
N.B.: As of 2004-2005, the UCB area in Developmental Psychology has reduced its proseminars from 4 to 2, further increasing the disparity between Clinical Science students and their Developmental counterparts.
Students
heading for careers in clinical practice face a different set of issues. Because
they take the same courses as the other graduate students do, they have to sit
through "breadth" and methods courses that are expressly intended for
students preparing for scholarly careers in various subdisciplines of the field.
I don’t doubt that budding practitioners would benefit from a concise overview
of cognitive psychology (at the very least, such a course might have spared us
the excesses of the recovered-memory movement). But that’s not what they
usually get, because for the most part clinical students take breadth courses
that are designed not for them, but rather for graduate students in other
subfields of psychology. Instead of a concise overview of the cognitive
psychology that every clinician ought to know, they get 15 weeks’ worth of
lectures comparing prototype and exemplar theories of conceptual structure, or
connectionist and rule-based theories of the acquisition of the past tense of
verbs. These courses are not listed because they are designed and taught
for the purpose of providing budding clinicians -- practitioners or scientists
-- with a breadth of exposure to psychology. They are listed simply
because they are available.
Something
similar happens with the methods and statistics requirement. Again, clinical
practitioners-in-training are subjected to exercises in counterbalancing and
Latin squares, discussions of the comparative advantages and disadvantages of
the Bonferroni and Tukey multiple-range tests, principal components versus
principal factor analysis, and how to calculate the proper degrees of freedom in
discriminant function analysis. Clinical researchers need this kind of
instruction but it is simply lost on those headed for clinical practice – not
because they aren’t smart enough, or even because they’re not particularly
interested (who in their right mind is?), but because they will never have
occasion to apply this knowledge in their entire lives. Just as budding
schizophrenia researchers don’t have to know how to do psychotherapy, budding
psychotherapists do not need to be taught how to do statistics. Instead,
they need to be taught how to consume statistics – and, in particular,
to respect statistical evidence as the core of the scientific base of clinical
practice.
Part of the problem here, frankly, is the APA accreditation scheme, which continues to be based, at least implicitly, on the scientist-practitioner model. Because it doesn’t distinguish between research-oriented and practice-oriented students, it imposes the same requirements on both. The result is that research-oriented students are forced to take courses that they don’t need, and practice-oriented students are forced to take courses that aren’t appropriate for them. Part of the solution, I think, is for research-oriented, "clinical science" programs to simply to drop their accreditation. That will permit them to develop their own curricula, just like their colleagues in cognitive, developmental, and social psychology, free of outside infringements on academic freedom. It will also make them less attractive to applicants who are not really interested in research, but who say they are in order to gain admission to high-prestige programs. Without the pressures of accreditation, clinical research training programs would look more like their counterparts in cognitive, developmental, and social psychology.
Part of the problem, too, is the changing nature of the clinical internship. When internships were originally proposed, by David Shakow (Shakow, 1938), they were intended to be opportunities for graduate students interested in clinical problems to get out of the classroom, and out of the laboratory, and become acquainted with "the living material of the field". Most of the student’s clinical experience was to take place on the internship itself. Actually, I’ve always thought that every graduate student ought to do some sort of clinical internship, for just this reason: students of visual perception could spend some time in an optometry clinic, and students of memory could spend some time with Alzheimer’s patients – but I digress. But that’s not the case anymore, because internships are increasingly construed as sources of revenue rather than vehicles of training, and so pre-doctoral students must devote increasing amounts of time preparing to provide reimbursable services with a minimum of costly supervision. The solution to this problem is for research-oriented programs to develop their own, in-house or "captive" internships, to give their students the kind of broad and deep encounter with "the living material of the field" that is consistent with the research-training goals of their PhD programs.
Actually,
I have long believed that every clinical psychology program, whether
science-oriented or practice-oriented, should have its own in-house or
"captive" internship program (Kihlstrom & Canter Kihlstrom, 1998).
Such an arrangement would avoid the situation that exists now, and has existed
at least since the 1990s, where there are more students graduating from clinical
training programs than there are internship slots. For example, data from the
Association of Professional Psychology Internship Centers (APPIC) indicates
that, from 1986 to 1997, the number of internship applicants unplaced three
weeks after Uniform Notification Day increased more than 700%, from 65 to 469,
while the number of internship vacancies decreased by more than 50%, from
79 to 34.
Things
improved somewhat after 1998, possibly by the addition of more unaccredited
internship slots, but still, from 1999 through 2004, unmatched applicants
outnumbered unfilled vacancies by a ratio of 2:1. Note added 11/30/06:
The trend continued, and worsened, in 2005 and 2006. In 2006, there were 731
unplaced applicants, an increase of almost 21% over 2004, and only 300 unfilled
vacancies -- a decrease of 3%.
If every program had its own internship, every graduate of every program would be guaranteed an internship slot – more like the match system for medical internships, where the number of slots actually exceeds the number of American graduates. But equally important, from a pedagogical point of view, if every program had its own internship, the pressure on interns to provide reimbursable services would be relieved, and the training goals of internships could be more closely articulated with the training goals – towards science or towards practice -- of the student’s predoctoral program.
As a footnote, let me say that I also believe it is a mistake for clinical training programs to "farm out" practicum experiences to community practitioners, who do not necessarily share the scientific values of the doctoral training program. It’s a prescription for disaster when students learn in the classroom that the Rorschach isn’t worth the paper it’s printed on (Wood, Nezworski, Lilienfeld, & Garb, 2003), and then they go out into the world to score Rorschachs for someone who thinks that they are the epitome of clinical assessment. Medical students do their clerkship rotations in academic health-science centers; similarly, clinical training programs should keep tight control of their students’ externship and practicum experiences.
So
what would training look like if we separated training for practice from
training for science? Frankly, it would look a lot like the training provided in
medical schools, which offer quite different curricula for research-oriented
students headed for the PhD, and practice-oriented students headed for the MD.
Berkeley doesn’t have a medical school, but there is a very nice one just
across the bay at the University of California, San Francisco, which also offers
PhDs in some 17 health-science fields from Biochemistry and Molecular Biology to
Sociology, and there is no overlap between the two sets of programs. For
example, medical students take an "Essential Core" of 9
interdisciplinary courses, beginning with an 8-week review of basic biological
and behavioral science followed by 8-week blocks devoted to the various organ
systems, cancer, infection and immunity, and life-span human development. These
block courses run parallel to a foundational course in patient care that runs
for two years, covering clinical skills, professional issues, and clinical
reasoning; this course then leads to a longitudinal clinical experience in the
third year and advanced rotations in the fourth. Medical students get some
biochemistry, but it is not the same basic biochemistry course taken by
biochemistry PhD students; and they get some neuroscience, but not the same
basic neuroscience course taken by neuroscience PhD students.
That’s essentially the vision I have for the future of training in clinical psychology. Training for clinical research should look more like research training in the rest of psychology, and training for clinical practice should look more like medical school. If the proposal on the practice side looks like a PsyD program, that’s intentional. I always thought that the PsyD was a good idea, even if I also thought it was usually a poorly implemented one (Yu et al., 1997). The PsyD recognized that there is an inherent difference between training for science and training for practice. If some PsyD programs were established to enable students (and faculty) to escape from science, and I am sure that they were, the solution is not to abandon the PsyD format, but to reform it so that future generations of practitioners are trained to respect science as the base of practice, and of the status and autonomy of their profession as well – just as current and future generations of physicians are.
There is no reason that training for science and training for practice cannot proceed on parallel tracks within the same department – though with different curricular requirements. Schools of public health and social work have no problems with these divided functions – nor, for that matter, do schools of law and business. At Berkeley we have a School of Optometry that trains both researchers in vision science and professional optometrists, as well as a College of Chemistry that houses separate departments of chemistry and chemical engineering. In both places everybody seems to get along just fine, but the two curricula are very, very different. Housing the two programs under the same institutional roof would allow current and future scientists and practitioners to benefit from contact with each other, but it would require some adjustments: clinical faculty would probably have to expand, to cover all the various aspects of clinical practice; and nonclinical faculty would have to agree to mount basic-science courses that are geared to the needs of future clinical practitioners.
But there are lots of alternative arrangements possible. Universities might develop "schools" of psychology, like Berkeley’s School of Optometry, which would train both scientists and practitioners with an expanded faculty. Or, training for practice could be removed to academic health-science campuses like UCSF, which already house schools of medicine, dentistry, and nursing; this would bring practice-oriented students into closer contact with patients, but it would also require expansion of the basic behavioral science faculty in these institutions – not incidentally, creating jobs for the products of the research training programs.
Training for practice might also be removed to free-standing schools of professional psychology. As with the PsyD, there is nothing wrong with such schools in principle. UCSF is for all intents and purposes a free-standing medical school, and it is not inferior to Harvard because it lacks departments of classics and political science. In fact, UCSF is the equal to Harvard (or its superior) precisely because, in addition to training medical students, it hosts a world-class cadre of basic scientists who do basic and applied research relevant to health care. This is a feature that most free-standing professional schools lack – and which may contribute to their relatively poor performance (Cherry, Messenger, & Jacoby, 2000; Maher, 1999; Yu et al., 1997). If professional schools are going to train practitioners properly, they are going to have to invest a lot more in their basic-science and research infrastructure than most of them seem inclined to do at present.
Now, I’m told that this room holds 113 people, so I bet that there are at least 113 objections to this proposal. I have answers for all of them, but time permits me to respond to only four or five.
First, I seem to be advocating the segregation of science and practice within clinical psychology, when clinical practice has already veered far from its scientific base, and managed care is putting a greater emphasis on evidence-based practices (Kihlstrom & Kihlstrom, 1998). But I’m not: science and practice are not like science and religion, Steven Jay Gould’s "Non-Overlapping Magisteria" that have nothing to say to each other (Gould, 2003). Science needs input from the real world of practice, and practice must be placed on a firm scientific base. But science and practice are different, and they deserve curricula that respect these differences. The fact that MDs don’t get the same coursework and research experience as biochemistry PhDs doesn’t make medicine any less science-based. Medicine is "scientific" not because physicians are scientists, but because medical practice is based on scientific evidence.
Second, my view of clinical practice may be outmoded. I have heard it said that the future of clinical psychology is not in the direct delivery of clinical services, but rather in the design and evaluation of assessment and treatment programs that will be delivered by other professionals. If so, we might have to invent an entirely different practice-training model, but it still wouldn’t be the model of PhD research training. Moreover, we’re probably going to need a lot fewer clinical psychologists than we’re training at present, and someone should inform the thousands of undergraduates who will be applying for positions in clinical training programs next year, and the thousands more each year after that, with the expectation that their careers will be devoted to testing and therapy of individuals and groups.
Third (and maybe fourth, depending on how you count), I seem to be going against history, abandoning the scientist-practitioner model at precisely the time when more and more physicians are training to do research, and basic researchers in fields like cognitive neuroscience are becoming more interested in studying patients. As a matter of fact, that’s why I enrolled in an experimental psychopathology program to begin with – because long ago I realized that psychopathology offered a unique perspective on normal mental life. But you don’t have to be trained to practice to take pathology seriously, or to do research with patients. There’s nothing that cognitive neuroscience does that physiological psychology didn’t do before it, and I have yet to see a piece of medical research that couldn’t have been done just as well by a "mere" PhD. My graduate advisor was an MD/PhD, and I have two MDs in my own department. Martin was a wonderful researcher, and so are my Berkeley colleagues; but such exceptions merely test the rule, and I’m generally an advocate of the division of labor. To take people who are trained to treat patients, and then divert them into research, simply diminishes the human resources available for healthcare (Kihlstrom, 2000). Scientists who want to study patients should collaborate with the practitioners who treat them, and practitioners who want to do research should collaborate with the scientists who know how.
Fourth (or fifth, depending on how you counted), I've abandoned the thing that I, as a devout generalist within psychology, ought to prize most: the breadth of exposure to the entire field of psychology that clinical psychologists must get, whether they are training for science or training for practice, by virtue of the APA accreditation standards. Now, I freely admit that the one thing I really like about the standards is that they contain a breadth requirement. And I think that a breadth requirement should be preserved in training for practice, just as there is a broad basic-science requirement in medical school. But as much as I don't like it, the fact is that psychology as a science is going the way of specialization, if not super-specialization. We have students in visual perception who don't know anything about memory, students in working memory who don't know anything about autobiographical memory, students in psycholinguistics who are so focused on the processing of individual words that they don't know what a sentence is. It's a shame, indeed it is, and I think it works to the detriment of psychology as a science that we graduate students who are so narrowly specialized that cognitive students know nothing about social psychology and social students know nothing about development. It won't kill budding clinical researchers to take four breadth courses -- what kills budding clinical researchers is training for practice. But in the final analysis, responsibility for breadth of training should fall on departments as a whole, not on clinical psychology alone.
Frankly, I’m sorry to see the scientist-practitioner model go, because it was a lovely idea. But it was also a rhetorical device constructed, in large part, to aid an upstart clinical psychology’s professional competition with the psychiatric establishment. Psychiatrists had at least the cosmetic advantage of medical training (although as far as I can tell not many of them ever practiced that much medicine), and the scientist-practitioner model seemed to say, "We’re doctors too, but we’re not just doctors, we’re also scientists". (The practitioner-scholar model adopts the same conceit -- "We're not just practitioners, we're also scholars": Funny, but physicians don't have to defend their professional status by referring to themselves as scholars.) Neither the scientists nor the practitioners need to do that anymore. Clinical scientists have enough to learn without learning practice as well, and clinical practitioners have enough to learn without acquiring research skills as well – especially in the new era of managed care and evidence-based healthcare. I’m simply proposing that our training programs reflect that fact of life. People who want to treat patients should be trained to treat patients, with the best methods that science provides, and people who want to do research should be trained to do research.
References
Cherry, D. K., Messenger, L. C., & Jacoby, A. M. (2000). An examination of training model outcomes in clinical psychology programs. Professional Psychology: Research & Practice, 31, 562-568.
Gould, S. J. (2003). The hedgehog, the fox, and the magister's pox: Mending the gap between science and the humanities. New York: Harmony Books.
Kihlstrom, J. F. (2000). Personal statement concerning research training in the behavioral and social sciences. In National Research Council (Ed.), Addressing the nation's changing needs for biomedical and behavioral scientists. Report of the Committee on National Needs for Bomedical and Behavioral Scientists, Education and Career Studies Unit, Office of Scientific and Engineering Personnel (pp. 101-107). Washington, D.C.: National Academy Press. Read text at: www.nap.edu/openbook/0309069815/html/101.html. The text is also available, with a prefatory note, at www.institute-shot.com/national_research_council_report.htm.
Kihlstrom, J.F. (2000, October). "Several things went wrong": Commentary on the NRC report on research training in the behavioral and social sciences. APS Observer, 13(10), 1, 17-18. Link to text at: http://www.psychologicalscience.org/newsresearch/publications/observer/nihcomment.html. Link to expanded commentary at: www.institute-shot.com/More_on_training_health_researchers.htm.
Kihlstrom, J.F. (2001, February). Response to Sutton: Further commentary on the NRC report on research training in the behavioral and social sciences. APS Observer, 14(2), 5.
Kihlstrom, J. F., & Canter Kihlstrom, L. (1998, August). The living material of the field. Paper presented at the American Psychological Association, San Francisco. Link to text at www.institute-shot.com/clinical_psychology.htm.
Kihlstrom, J. F., & Canter Kihlstrom, L. (1998). Integrating science and practice in an environment of managed care, The science of clinical psychology: Accomplishments and future directions. (pp. 281-293). Washington, DC, USA: American Psychological Association. Link to text at: www.institute-shot.com/integrating_science_and_practice_in_a_changing_environment.htm.
Maher, B. A. (1999). Changing trends in doctoral training programs in psychology: A comparative analysis of research-oriented versus professional-applied programs. Psychological Science, 10(6), 475.
Shakow, D. (1938). An internship year for psychologists (with special reference to psychiatric hospitals). Journal of Consulting Psychology, 2, 73-76.
Wood, J. M., Nezworski, M. T., Lilienfeld, S. O., & Garb, H. N. (2003). What's Wrong with the Rorschach? Science Confronts the Controversial Inkblot Test. New York: Jossey-Bass.
Yu, L. M., Rinaldi, S.A., Templer, D. I., Colbert, L. A., Siscoe, K., & Van Patten, K. (1997). Score on the Examination for Professional Practice in Psychology as a function of attributes of clinical psychology graduate training programs. Psychological Science, 8, 347-350.
This page last revised 03/07/07 02:43:54 PM.
