What we know about teacher preparation at elite education schools
There are some 1,400 schools of education in the United States–schools that prepare the teachers who teach most of America’s elementary and secondary students. By virtue of their numbers, and the fact that some 70 percent of our three million public school teachers have attended these institutions as undergraduates, education schools seem a likely subject of study for reformers. Do we know what our future teachers are learning in our schools of education?
Surprisingly, neither the defenders nor the critics of education schools have produced research that answers that central question–which is why my colleague Susan Rozen and I embarked on a project several years ago to evaluate the course syllabi at selected schools of education. Our results were first presented as part of a Washington conference, sponsored by the American Enterprise Institute and the Progressive Policy Institute, then published as a chapter in the 2004 book, A Qualified Teacher in Every Classroom. We intended our work to be a first step in responding to the need for useful data about schools of education. What we hadn’t expected was the degree of interest–pro and con–and the passions that our conclusions would attract. In fact, I was invited back to the Progressive Policy Institute a year after our report came out to debate one of our critics. (See my discussion of Dan Butin below.) This piece addresses the responses to our work, and I hope it will continue to move the discussion about this important topic forward.
Almost all the research to date about the quality of teacher preparation has been based on highly aggregated data that makes no distinctions between education schools or on interviews with teachers-college faculty members and their students (future teachers) about their personal experiences in college.
Stanford education professor Linda Darling-Hammond and several colleagues, for instance, once surveyed 3,000 beginning teachers in New York City to examine the relationship between their accounts of their preparation program and their confidence in their readiness for teaching. The survey results suggested that “Teachers prepared in a single formal program of preparation feel better prepared than those who take a series of courses from different institutions, who in turn feel better prepared than those who enter through alternative programs . . . and those who enter without prior experience or training.” But the study did not correlate teachers’ feelings of preparedness with measures of effectiveness, or, indeed, with any evaluation (other than self-evaluation) of their teaching ability. Much earlier (1989), Ken Howey and Nancy Zimpher, then at Ohio State, analyzed six exemplary–so they concluded–preservice teacher-education programs, all in the Midwest, but did not correlate their interview findings with program effectiveness. The key assumption of the study was that claims of ownership and identification with a program could be linked to program effectiveness–but no external validation of such assumptions was offered in the study. Nor did Howey and Zimpher offer any detailed analysis of course syllabi at the programs they reviewed.
A rare exception to this trend was research conducted by University of Georgia professor Peter Smagorinsky and Melissa Whiting of the University of Southern Mississippi nearly a decade ago. Smagorinsky and Whiting looked at syllabi for English methods courses from 81 universities for their book How English Teachers Get Taught: Methods of Teaching the Methods Class (1995) and, based on their findings, recommended learner- and activity-centered (“constructivist”) rather than traditional teacher-led programs. Of all these researchers, only Smagorinsky and Whiting actually reviewed a collection of syllabi, all in the field of English, to understand what prospective teachers were learning.
What We Intended
Beginning our research in 2002, my coauthor, Susan Rozen (Director of Reading/Literacy at the Bedford Public Schools in Bedford, Massachusetts), and I proceeded in the direction suggested by Smagorinsky and Whiting and decided to assess the nature of our future teachers’ schooling through an evaluation of the courses they were required to take. We expanded on that effort, however, and included a broader range of subjects and greater number of syllabi in our purview.
We reviewed a sample of syllabi in 16 schools of education, 14 of which were ranked in the top 30 in the nation by U.S. News and World Report (see Figure 1). We looked at schools like Harvard, Stanford, and UCLA (top ranked by U.S. News) as well as Eastern Michigan University and Sonoma State (not in the magazine’s top 30). We focused on initial certification programs and, within those programs, on the professional sequence required for certification. As a result, we looked at undergraduate programs, with the exception of those universities in states that required certification programs to be completed at the postbaccalaureate level (such as California). In certain universities, the traditional four-year undergraduate certification program has been expanded to five years, and it was therefore the five-year program we reviewed (e.g. Michigan State University). Where a university offered both a graduate and an undergraduate certification program, we chose the undergraduate program. We chose courses that were part of the required professional sequence because these courses are designed to prepare prospective teachers for teaching. We did not include in our review courses that were part of the general education requirements or courses that were specific to majors.
The syllabi we analyzed were in the domains of educational foundations, reading, and general methods courses (which typically include a student-teaching experience). In our original conference paper, we also presented findings in mathematics, but we were persuaded that in this field the academic courses were so integral to the professional preparation programs that we should not publish those findings. These domains form the heart of the education mission. If schools of education are essential, as their defenders argue, it is here that we would find evidence of the fundamental pedagogical skills that such schools provide.
Analyzing the course catalog in each university, we determined which courses in the professional sequence were required within the domains that we were reviewing. Then we gathered the syllabi for those courses from various sources, including professors, education school administrators, and the Internet. In a few cases we could not acquire a full set of the course syllabi in the domain under analysis, and so did not include that domain in our formal findings. In total, we reviewed 165 course syllabi that fell within our research criteria. Of these, 45 were in the area of the foundations of education, 61 in reading, and 59 in general methods and/or student teaching.
A Crucial Decision
As researchers, we then faced an important choice. We could simply record which texts were used in these courses and how often they were used, or we could list the books most often required (with numbers) and indicate which books were included rarely (again with numbers) or not at all. Although identifying worthy books that were not included on syllabi would open us up to criticism (as we soon discovered), the first option would have resulted in a long inventory of no clear significance, except perhaps to those within the profession. Naming books not included enabled us to draw out the significance of the books that were included. By juxtaposing counts of the most frequently required readings with the absence or near-absence of others, we were able to provide a first portrait of what future teachers are–and are not–learning at leading schools of education.
In the domain of foundations of education, the books most often required by the programs we reviewed were authored by Anita Woolfolk, Jonathan Kozol, Henry Giroux, Paulo Freire, Joel Spring, Howard Gardner, and John Dewey. Woolfolk’s work is a textbook in educational psychology, and one of Joel Spring’s volumes is a textbook in educational foundations. The rest are well-known works that embrace a constructivist and/or progressive standpoint. Conspicuously absent from almost all such syllabi were works that took a very different approach to teaching, such as those by E. D. Hirsch or Diane Ravitch. (We found Hirsch on two syllabi, Ravitch on just one.) Equality of education is a central theme of these courses, as evident from the included authors. Nonetheless, not one of the foundations courses, in the 15 schools of education for which we had complete data sets for that domain, asked students to read The Black-White Test Score Gap, at the time of our review arguably the leading collection of scholarly writings on that subject. We also noted that eight of the programs of teacher certification we reviewed did not cover either the philosophy or the history of education among the courses required for certification.
In our review of the courses in the teaching of reading, we followed the same approach. Here, however, we could draw on the findings of the National Reading Panel (NRP) and the National Research Council (NRC) to determine what was missing from the syllabi. Although recent research on reading is incorporated into some programs, work by Louisa Moats, Jeanne Chall, and Marilyn Adams–whose books and articles have been referenced frequently and used to support conclusions by the NRP and NRC–is rarely required. Analyzing the assignments and assessments listed in the syllabi, we also noted how rarely students were required to demonstrate competence in teaching reading skills and strategies. They were infrequently asked to demonstrate knowledge through presenting a lesson on critical reading skills in class or through taking a quiz or test.
Finally, we looked at the general methods courses and student-teaching experience. Here the norm is that the student first takes a methods course and then spends time in a public school classroom. While we again looked at required readings, for a gauge of the teaching experience part of the curriculum we focused on the handbooks that were provided to students as they entered their practicum and the descriptions of the practicum available in the course catalogs and/or on the web. Of the 11 schools from which we could get full data sets, just two indicated that they videotape their students’ teaching. Handbooks and catalog descriptions also suggested that professors at schools of education do not go to the schools to observe their students teach, but typically rely on adjunct instructors to report on the student teachers’ instructional performance. Finally, we found no evidence that in the practicum student teachers were being assessed on the basis of measured student performance. However, we have since learned that a very small number of schools of education are beginning to introduce this mode of assessment.
As a first effort to look at syllabi of required courses, our research represents a beginning. Fair concerns can be raised about technical matters. We were careful to name the 16 schools of education, and to define and list the number of courses and programs in each of the three domains we studied (foundations, reading, and methods), thus allowing other researchers to fully replicate our work. But we did not list the individual syllabi or name the particular schools (out of the 16) from which we were able to get full syllabi sets in each subdomain. It was clear when we sought syllabi from faculty members that, had we not assured their anonymity, we would not have obtained many of those syllabi: thus our choice to present aggregated data. Our data set was relatively small; certainly students take more than those courses that are required for certification, and a full survey would examine the many electives. We hope to address these issues in follow-up studies. When we conducted the research, more than half the schools of education listed many of their course syllabi on the web. Since that time, more are doing so. Ideally, this will allow future studies to include greater disaggregation of the data set.
Technical questions were not, however, the subject of most of the criticism we received at the Washington conference and in subsequent written and oral critiques. Kati Haycock, executive director of the Education Trust, said that we shouldn’t have focused on this “proxy stuff” and should have instead studied the links between individual teachers and their students’ achievement. We could agree with that–and have said that we encourage such research, although it is statistically very challenging. Meanwhile, our study focused elsewhere, on the quality of teacher preparation. Arthur Wise, president of the National Council for Accreditation of Teacher Education (the group that accredits almost half of the nation’s schools of education), argued that he didn’t think our survey sample was representative of “what’s going on at the institutions that prepare most of the nation’s teachers.” This is a fair objection, which we (and others) are now addressing, but a review of the top-ranked schools tells us what the profession looks toward as its best models. The faculty at elite schools are very influential in national policy debates, and they also train the future faculty at many less elite schools. Thus an analysis of elite schools, we believe, gives us important data from which we can draw some, albeit tentative, conclusions.
Other critics were harsher. Stanford education professor Darling-Hammond, author of The Right to Learn and well-known in our field for her emphasis on the importance of schools of education and professional certification for teachers, accused me of using “personal and political” standards and castigated our study as one of those “diatribes that come at the problems ideologically.” David Labaree, also from Stanford and author of The Trouble with Ed Schools, argued that we were wrong to take syllabi as serious evidence of anything: they are “more an ideological portrait of a course than actual substance.” As an observation about the nature and limits of syllabi, Labaree’s point is not to be denied: syllabi vary in their level of detail, ranging from a schematic account of texts, topics, and course requirements to a week-by-week specification of readings and assignments, along with supplementary readings, grading rubrics and formulas, and lengthy expositions of the perspective of the instructor. The thinner syllabi capture less information, and even the most complete will not capture what videotapes, for instance, would–the manner of presentation, the mode of teaching, the reactions of the students in class. But none of this undercuts our findings. The choice of readings is explicit in every syllabus and clearly indicates the formal content that the student should expect to encounter in the course. And by the same token, it is unlikely that much time in a class is devoted to texts that are not on the syllabus.
The critics were not done. The most recent critique of our efforts, from Dan Butin of Gettysburg College, calls me “disingenuous” for distinguishing empirical research about learning to read from opinions of professors who teach foundations courses. Professors speaking off the cuff at conferences or writing letters to others in the profession have fed into the controversy. Internet education blog Eduwonk.com eventually decried what it called “a nasty whispering campaign.” Frequently, we have had to point out that courses cited to us as evidence against our findings were not required courses but electives, just as our paper had made clear.
The strength of the discourse can, in all fairness, partly be blamed on us. Our complaint about the syllabi is a classic liberal one: students were not being exposed to a variety of important arguments and points of view. But it is true that we were deeply surprised by what our research found, and we expressed that surprise in our conference paper in strong language. We were perhaps too quick in using provocative phrases in our initial paper. Such phrases as “intellectually barren” and “too focused on indoctrination” unsurprisingly fueled incendiary comments. However, a problem of language cannot be taken to invalidate our conclusions.
Whispers–and personal attacks–aside, it is worth responding to the more substantive arguments of our critics. In a piece published electronically by Columbia University Teacher College’s TCRecord last August, Dan Butin raised an important issue of causality in an extensive critique. “There is an extremely loose coupling between what is written and what is taught,” he argued, “between what is taught and what is learned, and between what is learned and what is subsequently enacted in an efficacious way.” And though Butin rightly points out that we did not address those linkages, he says about his own paper, “This study will not address these assumptions either.” Butin’s point, of course, is equally applicable to all formal education: taken to its logical conclusion it suggests–counterintuitively–that we should not care about what is taught since education makes no difference to our future practice. At the very least, if future teachers are taught to believe in one set of principles and are subsequently forced to abandon them in practice, their professional preparation is problematic. It is worth adding that Butin criticized me for insisting on balance: “Steiner’s criteria of balance . . . focus on the wrong agenda.” Butin argues that I should have realized that reading books from only one perspective will enable “deep analysis.” Butin’s intellectual embrace of fundamentalism, the absolutism of his antiliberal stance, once again comes as something of a shock.
As for Linda Darling-Hammond’s claim that our standards of review were personal and political, we couldn’t disagree more. And we have explained and defended our standards in our published chapter: where there is empirical evidence for the positive impact of good teaching practices, we used it to infer what should be included in a teacher preparation course–thus the standards we suggested for the teaching of reading. Scholars disagree, naturally, about which empirical studies are most persuasive. However, our work is based on a broad set of major national research findings. In this domain, at least, the charge of playing personal politics looks a little shaky. In other cases, however, such as in our review of the foundations of education courses, no such empirical data are available. So, for example, we looked for readings that have stood the test of time and/or have proved influential in our current education debate. And we sought evidence that professors who were teaching their students about good teaching practices were also analyzing video of their students actually engaged in their student teaching. We were basing those desiderata not on research-based evidence, but on the assumption that education-school faculty would naturally wish to give direct feedback to this critical part of their student’s preparation for their profession.
It is therefore in these two areas–foundations courses and teaching methods–that one imagines that Darling-Hammond’s critique was intended to cut most sharply. Only a personal or political decision, apparently, could explain our hope for finding Plato as well as Dewey, or works by critics as well as defenders of multiculturalism. (I intend “multiculturalism” to be understood here not as describing our condition, but as embodying a particular progressive political and pedagogical program.) To ask that students be required to read E. D. Hirsch’s critique of Dewey while also asking that they read Dewey is thus to render a personal–since it is not an empirically or research-based–opinion. It is not, however, a political decision, unless, of course, one believes a call for balance is in itself political.
It is also important to note, in this context, that at the time of writing up our research, the coursework Stanford’s school of education required for teacher certification was on the web. Heading up the material was a list of suggested readings for new students in the program. At the time we reviewed the list, it comprised works by Gloria Ladson Billings, Elizabeth Cohen and Rachel Lotan (eds.), Linda Darling-Hammond, Howard Gardner, Jonathan Kozol, Deborah Meier, Jeannie Oakes, Laurie Olsen, Vivian Paley, Mike Rose, Frank Smith, David Tyack and Larry Cuban, and Guadalupe Valdes. There is as little empirical evidence that reading these books will improve one’s teaching as there is for any other list of foundational texts. To take one example, if reading Deborah Meier’s moving book (The Power of Their Ideas) about turning around a school enabled her readers to duplicate her performance as principal of Central Park East, urban education in the United States would be very different. What standards, therefore, did Darling-Hammond and her colleagues use in recommending their own books and those of the others they place on the list? Most likely, personal and political opinion, grounded in deeply held convictions.
But this response may occlude a deeper set of issues. The premise of Darling-Hammond’s charge is that there are standards for the choice of foundation course readings that are neither personal nor political. What could such standards be? On what “objective ground” does one choose Dewey over Plato, or ask for X rather than Y? One reasonable response is to suggest that such choices, at their thoughtful and considered best, represent not simply opinions but judgments, and to insist on the difference between the two. The judgment that future teachers should have read a Socratic dialogue is based on the experience of two millennia of readers, who have found in those dialogues an inexhaustible set of questions, problems, and insights. By contrast, a close review of that list of readings that Stanford University recommended to its new student-teachers is instructive. All of these authors are American, and all but one of their books was written in the past 30 years. In short, one of our top schools of education is convinced that nothing from China or Europe or Africa, and nothing written before 1974 merits inclusion. Rather than characterize the Stanford list (or our own call for balance) as the result of personal and political opinion, I would suggest it embodies an effort of judgment: a serious effort to shape the fundamental worldview of future teachers. It is this judgment that we should be discussing. My own is that as teachers, we join a community and a craft that stretch back for millennia. We owe it to the next generation of teachers to introduce them properly to the richness and depth that is the legacy of those who have come before.
David Steiner was until recently the chairman of the Department of Education Policy at the School of Education at Boston University. He is currently on leave to serve as Director of Arts Education at the National Endowment for the Arts. The opinions expressed here are his own.