Education schools periodically aspire to be like medical schools. Twenty years ago, three influential reports envisioned novice teachers as interns, methods-course instructors as clinical faculty, and student-teaching sites as the equivalent of teaching hospitals. If the best practices of medical schools could be copied, then ed schools would be able to recruit and retain better candidates. That was the hopeful message in “Tomorrow’s Teachers,” “A Nation Prepared,” and “Time for Results,” three major panel reports issued in 1986.
Twenty years later, it is time to look again at this strategy. The current emphasis on testing aspiring teachers deserves historical perspective. What happened to tests in medical schools in the early 20th century, the era when medical education improved dramatically?
Many medical schools changed nothing. The familiar routine of written, oral, and “practical” exams seemed adequate. Each department created and graded its own tests, and nearly everywhere they required rote recall of facts and procedures.
The few innovations occurred before and after, not during, medical education. Selective admissions, rare before the 1920s, eliminated the unqualified, as did tougher state licensing exams. Rather than focus on new tests, medical educators concentrated on other paths to improvement: closer ties with local hospitals, better laboratory facilities, more full-time clinical faculty members, tougher graduation requirements, and much larger budgets.
But at one of the best medical schools—Harvard University’s—the policies for testing and grading students changed in the first half of the 20th century, during the presidency of A. Lawrence Lowell (1909-1933). Assessment was only one of several strategies to improve the school, and that is one cautionary lesson for educators to consider. There is, however, another important lesson in the Harvard Medical School story: the value of looking carefully at the full scope of what students know and do.
What held together several dozen courses in a four-year medical curriculum? In medical schools before World War I, the relationships between the basic science and the clinical courses were poorly specified, especially in an age when many preclinical and clinical faculty members considered themselves superior to each other. (President Lowell compared the rift, circa 1910, to the tension between the North and the South in the 1850s.) Many of the older, part-time, clinical faculty members knew less laboratory science than most of their students, and they downplayed its importance for the work of a general practitioner. For them, the mantra of “science at the bedside” was not as urgent as it was for David Edsall, who was the dean of Harvard’s medical school from 1918 to 1934 and became one of the pioneers in relating physiology to disease.
In undergraduate teacher education today, we rarely ask our students to correlate the mass of information they acquire.
The direction of change throughout American medical research was toward laboratory discoveries with clinical significance, as well as clinical investigation carried out in the hospitals’ own laboratories. Harvard was no exception to that trend. The former wariness between clinical and preclinical faculty diminished in the 1920s, when all fourth-year medical students had to take and pass a comprehensive examination spanning the entire curriculum.
“Discuss epigastric pain; its physiological, anatomical, and clinical significance” was one of the three essay questions on the eight-hour 1920 comprehensive exam. Usually the questions focused on a specific illness, injury, or organ, but the discussion of those slivers of the curriculum had to include both scientific and clinical knowledge. Furthermore, three professors could pose any question imaginable as the student examined a patient at a teaching hospital. Rather than test memory, the goal was to see how well the students could correlate and apply what they had learned in various courses.
In addition to grasp, scope, and power (three of Lowell’s favorite words), the comprehensive examinations fostered self-reliance. The current enthusiasm in teacher education for tight alignments between curriculum and assessment would have worried Dean Edsall. For him, the comprehensive exams encouraged independent reasoning. Students could not cram for the capstone exam by taking certain courses, reading particular books, or hearing some lectures. Edsall resisted any curricular changes that would foster “spoon-fed correlation,” where the teachers “do for the student what he should do for himself.” When Abraham Flexner referred to the comps as a “system,” Edsall told him they were the opposite. The point was not to set up more requirements; medical students instead needed the freedom to cross course and department boundaries to find and use whatever knowledge applied to the problems at hand.
In undergraduate teacher education today, we rarely ask our students to correlate the mass of information they acquire. We hope that the sequence of coursework by itself will establish connections among courses, but we offer few incentives for candidates to integrate coursework in or before the senior year. With increasingly detailed specifications from the National Council for Accreditation of Teacher Education and the states, we are prone to tell candidates exactly what they must do, rather than encourage them to develop their own clinical judgment, to use, as Edsall said, “what he has gained from any and all sources—not to determine simply how much of this he can remember, but how well it serves him in comprehending and elucidating problems” of professional practice.
Edsall and Lowell faced another challenge early on. Very few professors ever changed the short-answer, factual course exams that had been given since the 1870s. “Gangrene of the toe,” “Discuss treatment of an infected knee joint,” “The surgical bearings of choked disc”: Third-year surgery exams, for instance, posed the same list of items created 40 years earlier. Tests sought definitions and summaries, as the most frequent first words of the questions indicate: “What is … ?” “Give an account of … ,” “State … ,” “Explain … ,” “Name … ,” “Describe … ,” “Enumerate … .”
But the importance of the grades derived from those tests lessened in the 1920s. Admission to fourth-year elective courses no longer turned on high grades. Special honors were awarded for excellent work in one particular area, and regular honors required good work on the comprehensive examinations along with good grades. Individual class rank gave way to placing students in the top, middle, and lower thirds.
Most crucially, promotion now depended on the recommendation of a faculty committee that could change course grades as it awarded one cumulative mark for the year’s work. The faculty relied on its impressions of the entire scope of the students’ work, rather than calculating grade point averages or using the standardized tests that were sweeping the country at that time. Watching performance in many settings—labs, outpatient clinics, hospital wards, class discussions, seminar meetings—provided a better gauge of fitness than a clutch of test scores. “Careful and close examination of the individual student” is essential, one professor told the dean, who, after 1925, asked each instructor to send written comments on the students to his office.
Don’t teacher-educators today also scrutinize a wide range of student work? For sure. But do we discuss student work with each other as often and as thoroughly as the Harvard medical faculty did? By creating three “promoting boards,” along with a large “committee on examinations,” the medical school set up unprecedented organizational support for the frequent discussion of students. Before the 1920s, such discussion happened informally, usually when problems arose, but now there were structures in place to make sure it took place.
Rather than file bits and pieces of NCATE-driven tasks in electronic portfolios, shouldn’t we get to know our students as unique individuals and then ground our assessments on our overall sense of their strengths and weaknesses?
Personal attention, knowing students well, could not be avoided or delegated to others. That would have been easy to do in a decade when enrollments rose and research productivity became more important than ever. It is not surprising that some faculty members yearned for external examiners, or a paid, in-house staff. But to his credit, the dean refused to subcontract this work. He knew that it was not easy. Each spring, the readers of the comps devoted from 35 to 45 hours to their job. They read answers in which some students still piled up long lists of facts, rather than thinking critically. The readers received exam questions from colleagues that were much too broad (“Discuss milk”). The comments on individual students varied in length and cogency, including one remark on a man who had never enrolled.
Even so, Dean Edsall and the Harvard medical faculty persevered. They believed in the importance of knowing each and every student well enough to share and defend those impressions with their colleagues.
Shouldn’t we know our preservice candidates well enough to do the same? Rather than file bits and pieces of NCATE-driven tasks in electronic portfolios, shouldn’t we get to know our students as unique individuals and then ground our assessments on our overall sense of their strengths and weaknesses?
A version of this article appeared in the October 19, 2005 edition of Education Week as Doctoring Schools