Summary and Research Agenda
In response to our broad charge, the committee examined many aspects of the complex and diverse set of institutions and programs through which the majority of the nation’s teachers are prepared. The bulk of our report focuses on the first three questions in our charge, about the candidates who enter teacher preparation programs, the nature of the pathways and programs those candidates select, and the extent to which the content of teacher preparation is consistent with scientific evidence. The first part of this chapter provides a summary of our findings about teacher preparation in the United States.
There is no lack of writing on teacher preparation, yet there are many gaps in the research base. The fourth part of our charge was to make recommendations regarding future data collection that would provide useful, valid, and reliable information. The second section of this chapter presents our conclusions about the research base, and the final section presents our recommendations for future research.1
SUMMARY: TEACHER PREPARATION IN THE UNITED STATES
We looked first for information about the first two parts of our charge, regarding the individuals who enter teacher preparation programs and their academic preparation, as well as the types of instruction and experiences
they receive. There is no system in place to collect data across the myriad teacher preparation programs and pathways in the United States. Thus, we can say little about the characteristics of aspiring teachers, the programs and pathways they follow, or the outcomes of their preparation. We found some information about general elements that most teacher preparation programs share, such as courses in pedagogy and the foundations of education and required fieldwork. We also found that both programs and pathways vary dramatically in their requirements, structure, and timing. Because of the paucity of systematic research as well as the enormous variation in virtually all aspects of teacher preparation programs and pathways, we cannot draw any specific conclusions about the characteristics of current teacher preparation programs.
Researchers have examined particular programs and pathways to look for differences among the people who pursue different routes, as well as differences in the effectiveness of graduates. The findings are slim. Some research suggests that there are differences in the characteristics of teacher candidates who are attracted to different pathways and types of programs. There is also some research that compares the outcomes for graduates of different kinds of programs. However, the distinctions among pathways and programs are not clear-cut, and there is more variation within categories such as “traditional” and “alternative”—and even within the category of master’s degree programs—than there is between the categories.
Conclusion 3-1: There is currently little definitive evidence that particular approaches to teacher preparation yield teachers whose students are more successful than others. Such research is badly needed. We believe that the highest priority research would be studies that examine three critical topics in relation to their ultimate effect on student learning:
comparisons of programs and pathways in terms of their selectivity; their timing (whether teachers complete most of their training before or after becoming a classroom teacher); and their specific components and characteristics (i.e., instruction in subject matter, field experiences);
the effectiveness of various approaches to preparing teachers in classroom management and teaching diverse learners; and
the influence of aspects of program structure, such as the design and timing of field experiences and the integration of teacher preparation coursework with coursework in other university departments.
Content of Teacher Preparation Programs: Research Evidence
The question of the extent to which the required course work and experiences in reading, mathematics, and science across teacher preparation programs are consistent with converging scientific evidence presented a somewhat different challenge for the committee. Within each of the three fields there is a range of material that is potentially relevant. This material includes a relatively small body of empirical studies that provide some evidence about the effects of particular kinds of instruction; it also includes an even smaller amount of evidence about the effects of particular approaches to teacher preparation.
The other kinds of research that are available include descriptive and qualitative studies, which explore many aspects of teaching and learning in the three subjects, as well as a substantial body of empirical work on learning and cognition, which has had an important influence on practice within each discipline. In addition, the professional organizations that provide leadership in the fields of reading, mathematics, and science have drawn on the available research and their own intellectual traditions and experience as educators to develop content and achievement standards for students, standards for teachers, and, in some cases, guidance or standards for teacher education.
Substantial work by educators and researchers has identified some strong arguments about the factors that are likely to influence teacher quality and student learning. Yet this work is only a starting point because the empirical evidence supporting the impact of these factors is limited. The research base varies across the three school subjects, and our conclusions about preparation in each field reflect these differences. Our discussions of the state of knowledge in these three areas also reflect the fact that we found no evidence in the literature that undermines the current recommendations of disciplinary experts, or calls into question the tradition, common to many fields besides education, of basing some decisions about professional education on such recommendations.
Conclusion 5-1: Successful beginning readers possess a set of foundational skills that enable them not only to continue growing as readers but also to progress in all academic subjects. A variety of instructional approaches that address these foundational skills can be effective when used by teachers who have a grounding in the foundational elements and the theory on which they are based.
Conclusion 5-2: It is plausible that preparation in the nature of the foundational reading skills and research-based instructional approaches would improve teachers’ practice to a degree that would be evident in learning outcomes for their students. However, there is currently no clear evidence that such preparation does indeed improve teacher effectiveness or about how such preparation should be carried out.
Conclusion 5-3: There are very few systematic data about the nature of the preparation in reading that prospective teachers receive across the nation. The limited information that exists suggests that the nature of preparation of prospective teachers for reading instruction is widely variable both across and within states.
Conclusion 5-4: Little is known about the best ways to prepare prospective teachers to teach reading. Systematic data are needed on the nature and content of the coursework and other experiences that constitute teacher preparation in reading.
Conclusion 6-1: It is plausible that to provide students with the instructional opportunities they need to develop successfully in mathematics, teachers need preparation that covers knowledge of mathematics, of how students learn mathematics, and of mathematical pedagogy and that is aligned with the recommendations of professional societies.
Conclusion 6-2: Many, perhaps most, mathematics teachers lack the level of preparation in mathematics and teaching that the professional community deems adequate to teach mathematics. In addition, there are unacceptably high numbers of teachers of middle and high school mathematics courses who are teaching out of field.
Conclusion 6-3: Both quantitative and qualitative data about the programs of study in mathematics offered and required at teacher preparation institutions are needed, as is research to improve understanding of what sorts of preparation approaches are most effective at developing effective teachers.
Conclusion 7-1: Systematic data are needed on the nature and content of the coursework and other experiences that currently constitute teacher preparation in science. Research is also needed to examine the
propositions regarding the teaching and learning of science contained in professional recommendations that have not been adequately examined empirically.
This was the picture of what converging evidence suggests about teacher preparation, against which one might measure what is currently happening. However, there is very little systematic research about current practice in the preparation of reading, mathematics, and science teachers. The limited information we found does not support broad conclusions about the nature and content of current teacher preparation programs.
As we describe in Chapter 8, our investigations of these issues led us to consider the accountability system, which is designed to ensure the high quality of teacher preparation programs. The accountability measures in place are diverse, and the gaps in the data available are large. If accountability for teacher preparation is to become more effective, a major assessment of the current situation would be needed.
Recommendation 8-1: The U.S. Department of Education should sponsor an independent evaluation of teacher education approval and accreditation in the United States. The evaluation should describe the nature, influence, and interrelatedness of approval and accreditation processes on teacher education program processes and performance. It should also assess the extent to which existing processes and organizations align with best practices in accountability and offer recommendations for how they could do so more effectively in the future.
The last part of our charge was to make recommendations regarding a model for data collection that would provide valid and reliable information about the content knowledge, pedagogical competence, and effectiveness of graduates from the various kinds of teacher preparation programs. The base of empirical knowledge about teacher preparation is thin. We believe the way forward is to build on what has been done by drawing on the professional consensus in each academic field for promising hypotheses about which features of teacher preparation are most promising and to subject those hypotheses to rigorous research. We were asked to develop an approach to future research that would provide a firmer foundation for policy and practice in the future. We organized our response around two overarching needs:
improved understanding of the relationships between characteristics of teachers’ preparation and students’ learning, and
a comprehensive, coherent system for collecting data about teacher preparation.
In discussing these two needs, we offer our assessment of the most important questions to pursue and the most productive means for doing so. Our discussion and recommendations draw on a study we commissioned (Crowe, 2007) to examine the current status and quality of data systems, as well as analysis of the available data related to the questions in our charge.
The Relationship Between Characteristics of Teacher Preparation and Student Learning
An obvious question to ask about teacher education is whether particular ways of preparing teachers lead to measurable improvements in student learning. Many researchers have worked hard to establish such connections. In Chapter 2 we discuss why it is difficult to establish clear causal links between aspects of teacher preparation and outcomes for the students teachers teach after they have completed their training. Programs may differ in the types of candidates they attract and in the types of knowledge and skills that candidates acquire. Programs may also differ in whether and where their graduates teach (e.g., what kinds of schools; urban or rural) and how long they remain teachers. And programs almost certainly graduate people who have different capacities to use their knowledge and skills to improve their students’ learning. Some programs may produce graduates who are more effective in some settings than others. We repeat as Figure 9-1 the model used in Chapter 2 to portray the complex interactions among different elements that influence teacher quality and student achievement.
Thus far, some attempts have been made to compare the learning of students whose teachers were prepared in one way to that of students whose teachers were prepared in a different way. Unfortunately, we found that the existing studies have generally been insensitive to the details of teacher preparation that are most likely to result in differences in quality. Theoretically, the best way to do this sort of investigation would be experimental field trials, in which teacher candidates are randomly assigned to different programs and students are randomly assigned to program graduates. When randomization is not possible, however—which is frequently the case in studies of education and other complex human behaviors (see Chapter 2) —other means of estimating the effects that programs have on participants and their students can provide valuable information. Other approaches include regression discontinuity designs, instrumental variables, or natural
experiments with appropriate controls. Researchers are still in the process of working out an array of practical approaches to providing reliable answers to questions about teacher preparation.
Research in teacher preparation would also be much easier to conduct if researchers had better measures of student outcomes than standardized achievement scores in mathematics and reading. Although scores are readily available and easy to use, they provide incomplete measures of both students’ learning and the effects of teachers (though assessment issues differ across the school subjects). We also believe there is much to be learned regarding the links between teacher preparation and the knowledge and skills teachers display in the classroom. Recently, there has been substantial interest in the development of observational protocols that measure various domains of teaching that have been linked to student outcomes (e.g., Mashburn et al., 2007; Matsumara et al., 2008; Grossman et al., 2009). Observational protocols offer a vehicle for exploring the contributions of teacher preparation and evaluating teachers’ effectiveness.
A strong research program designed to illuminate the ways teacher preparation influences outcomes for students would include evidence drawn from a variety of different perspectives, with the goal of establishing not only whether a particular feature of preparation is important to student
outcomes, but also why it is important. At present, research has provided only a fragmented and limited picture of how characteristics of teacher preparation improve student outcomes.
In our judgment, the simplest and most effective way to produce a clearer picture would be to focus research on the aspects of preparation that have the highest potential for effects on outcomes for students. Existing research provides some guidance on three aspects of teacher preparation that are likely to have the strongest effects: content knowledge, field experience, and the quality of teacher candidates.
There are strong reasons to believe that teachers need relevant content knowledge to be effective. Nevertheless, there is surprisingly little research that establishes clear and strong connections between teacher content knowledge and student learning. Throughout the report we discuss the challenges of isolating these connections—not only that the measures of both content knowledge and student learning are weak, but also that the relationships among learning, learners, classroom practice, and teacher preparation are complex. Nevertheless, we believe understanding how content knowledge influences student outcomes is very important.
The conclusions we drew about the research that was needed related to preparation in reading, mathematics, and science focus on this point. Looking across these three fields, we note several topics that would be fruitful for research:
Clarify what is meant by teacher knowledge and how that construct can best be measured, and how content knowledge interacts with knowledge of the pedagogical application of that knowledge.
Develop better measures of student learning of academic content.
Establish the extent to which measures of teacher content knowledge can predict student learning.
Conduct intervention studies in which teacher content knowledge is enhanced and the intervention group is compared with one or more control groups established by a rigorous research design, such as randomized trials.
Most observers agree that aspiring teachers should have field experience as part of their training. Yet reviews of previous research have failed to reveal any distinct relationships between the way field experiences are structured and implemented and teacher effectiveness. Recent work sug-
gests that teachers benefit from preparation programs that provide significant oversight of field experiences and from field experiences that are congruent with candidates’ eventual teaching positions (e.g., Boyd et al., 2008a). Although this research is suggestive, there is no systematic causal evidence on what aspects of field experiences have the greatest effect on teacher effectiveness.
A substantial research program could be built around hypotheses regarding field experiences. A program that included theoretical work, qualitative analysis, statistical analysis, and randomized experiments could provide strong causal evidence of the effects and mechanisms by which various components of field experiences—such as coplanning, coteaching, scaffolded entry into practice, seminars with mentors, a mentor with relevant content and grade level experience, and the like—affect teachers’ classroom practices and student achievement. For example, each primary component of field experiences could be systematically manipulated in a randomized control field trial to examine the relative effects on teacher classroom practices and student achievement outcomes.
It is also likely that statistical analysis that exploits the substantial differences in current practice would yield insights on relative effectiveness, although this analysis would require controls for selection effects. Qualitative analysis that examined the implementation of the field experience components would provide important insights on how and why these components may influence teacher effectiveness and could offer some suggestions on whether there may be important interactive effects. For example, one interactive effect that is worth examining is whether teachers who work in low-performing schools benefit more from certain field experiences than others.
Quality of Teacher Candidates
The quality of new teachers entering the field depends not only on the quality of the preparation they receive, but also on the capacity of preparation programs to attract and select academically able people who have the potential to be effective teachers. Attracting able, high-quality candidates to teaching is a critical goal, and there is reason to believe that some pathways and programs are much more attractive than others for such potential teachers. Less clear are the factors that attract the best candidates, the way program selectivity and preparation interact and the effect of each on student learning, and the extent to which the importance of these factors vary depending on the attributes (such as grade level and ability) of the students whom these teachers ultimately teach. That is, though some programs are more selective and attractive to academically accomplished candidates,
researchers have not clearly established whether those candidates make the best teachers.
A Comprehensive Data Collection System
A primary obstacle to investigating these and many other important aspects of teacher preparation is the lack of systematic data collection, at both the national and state levels. Crowe (2007) found that, apart from the methodological problems we have discussed, there is a problem with the “availability and quality of data about nearly everything having to do with teacher preparation” (p. 2).
The many basic questions that are at present difficult to answer systematically include the following:
What are the characteristics of candidates who enter teacher preparation programs?
How do those characteristics vary by program or pathway?
Where do entrants and graduates of preparation programs ultimately teach?
How long do teachers with different types of preparation continue to teach? Are differences in preparation associated with differences in teachers’ career trajectories?
Where do teachers with different types of preparation teach?
How do the knowledge and teaching practices of teachers with different types of preparation differ?
What have been the effects of states’ policies regarding program approval and teacher certification?
A more comprehensive approach to data collection would provide both baseline monitoring of the status of teacher preparation (and improved opportunities to link that information with other aspects of the public education system) and a common foundation on which to build research efforts that investigate important aspects of teacher preparation. Moreover, it would provide the basis for much-needed national attention to the importance of teacher preparation and the urgency of improving it.
What would a more comprehensive approach look like? A comprehensive data system for teacher preparation would provide meaningful information about teacher candidates, preparation programs, practicing teachers, the schools in which those teachers teach, and the students they teach. For example, with respect to teachers, observational measures of their skills and practice would provide information about the content of preparation that goes beyond degree title, courses taken, or certifications attained. Similarly, with respect to students, the standardized performance
measures that many states currently use provide important information, but it will be essential to bring other kinds of information about student learning into the systems used to track trends and evaluate the effects of teachers. The assessment community has made important strides in developing richer measures of achievement, such as portfolios of student work and assessments that are embedded in classroom instruction and in developing ways to standardize them. Measures of other important aspects of learning, such as persistence and motivation, are also important, but at present these issues have a very limited presence in large-scale data collection efforts and accountability systems.
The new data would be integrated so that information about teacher candidates and their preparation could be connected with the knowledge, teaching practices, career paths, school environments, and student outcomes of the teachers who are prepared in different ways. One key to integration will be consistent definitions of key indicators. At present, states each develop their own teacher licensure categories (which may change from year to year), determine which assessments teachers must pass—and most use many different ones—and what performance level will constitute passing. States differ in the way they define teaching assignments and identify out-of-field teachers, and they even have differing ways of counting years of teaching experience. There are countless other sources of variation that make it extremely difficult for researchers to compare across states or generalize from the available information. Some information is also needed on a national basis because substantial numbers of teachers move among states during their careers.
A few states have developed exemplary systems for capturing data. Florida, for example, has the PK20 Education Data Warehouse (see http://edwapp.doe.state.fl.us [October 2009]), which is a nationally recognized model. This system collects comprehensive information about the entire educational system and has built-in linkages so that researchers do not need to create cross-files to investigate specific questions. Texas, Utah, and Louisiana are developing similar systems. Unfortunately, few states collect a significant amount of data about the teacher preparation programs and pathways that are not based in their university systems (Crowe, 2007).
The U.S. Department of Education has focused on the problem of education data. The Institute for Education Sciences (IES) awarded grants to 14 states in 2005 to develop “well-designed, comprehensive statewide longitudinal data system[s] with the capacity to follow individual students’ performance over time, to transmit student information both within and between States, and to provide educators and education researchers with the data needed to improve outcomes for students” (see http://nces.ed.gov/Programs/SLDS/ [October 2009]). High-quality research on teacher education will require extending those data systems to information about
teachers, their background and education, preparation, and career paths across and within school systems.
At the national level, there are other data sources available, though none are linked to each other. They include a website maintained by the National Student Clearinghouse, which provides electronic verification of enrollment, degrees earned, and other information; the website of the National Association of State Directors of Teacher Education and Certification (NASDTEC), which lists approved college and university teacher preparation programs; a website mandated as part of the Title II of the 1998 Higher Education Amendments accountability system; the National Center for Education Information, which collects information on alternative routes; and the Core of Common Data (CCD), a project of the National Center for Education Statistics that collects a variety of relevant data on schools and students (Crowe, 2007).
The Data Quality Campaign (see http://www.dataqualitycampaign.org/ [February 2010]) has examined the data collection systems in every state and developed a set of recommendations to guide states in collecting comprehensive longitudinal data from the entire educational system (preschool through higher education) and using it to improve student achievement. To provide trustworthy answers to the questions about the connections between teacher preparation and student learning for which this committee could not find answers, data collection related to teacher preparation that is integrated into this type of system would be extremely valuable. Useful data collection will cover all levels of the education enterprise, from local school districts to states and the federal government. This means that a data network, rather than a single monolithic data system is needed. The federal government can play a critical role in coordinating definitions and standards and helping to ensure that measures are common across the nation.
Ideally, there would be a high-quality, well-defined state data system in every state that gives explicit attention to collecting baseline information about teacher education and its effects. Each state data system would use variables defined in the same way and measured in the same way. The network would include data analysis files that allow researchers to perform secondary analyses to look for causal relationships among the natural variation in approaches to teacher preparation captured in the data file. Most states are now building such databases, and with a reasonably modest expenditure of money and effort they could be expanded to collect data on the individuals who enter different types of teacher preparation programs and the achievement of the students they later teach. A significant sum of federal money has recently been targeted for state data systems related to education as part of the American Recovery and Reinvestment Act of 2009. A key goal for this federal funding is to make data collection more efficient and integrated so that it can better support improvement: thus, it is an
ideal time to ensure that states incorporate information related to teacher preparation in their data collection efforts.
Finally, a targeted longitudinal nationally representative study—similar to those that the National Center for Education Statistics has conducted in other areas—would make it possible to track individuals from before they enter teacher education through their teacher education experiences and into the classroom. We recognize that designing such a study would be difficult, primarily because it is difficult to anticipate which high school students will ultimately pursue teacher education and become teachers. But the feasibility of such a study could be explored: if it proved feasible, it would provide important information that could not be learned from either a national indicator system or existing state databases.
In order for policy makers and teacher educators to have a stronger empirical basis for decisions about teacher preparation, a much clearer and more detailed picture is needed of teacher candidates and how teacher preparation is delivered, as well as a means of tracking changes in this picture over time. A body of evidence, developed from multiple perspectives and using an array of research designs, that establishes links between teacher preparation and learning—both teachers’ learning and K-12 students’ learning—would also be of great value to those who are responsible for teacher preparation. Some evidence would come from research intended to identify causal links between specific aspects of preparation and students’ achievement. Other evidence would come from more systematic collection and analysis of both data about teacher candidates and the steps they take as they work to become teachers, and descriptive information about programs and pathways (such as analysis of accreditation materials, syllabi, course descriptions, and other program documents, as well as interviews and other observations).
Research on the link between preparation and teachers’ knowledge of content and research-based instructional practices and frameworks and between preparation and teachers’ skills and performance in classrooms would also be valuable. Some of this research would also examine the contexts in which teachers from various programs and pathways are more or less able to use the knowledge and research-based skills they develop during preparation and the conditions that support or constrain their capacity to use what they know.
There is currently almost no nation that is not concerned about teacher quality and teacher preparation. The conviction is widely shared that the economic health of a nation depends on the quality of its education system, which in turn depends directly on how teachers are selected, prepared, sup-
ported, and evaluated. The U.S. Congress has asked for answers to important questions about teacher candidates and the nature and quality of the preparation they receive. We offer two recommendations for building an empirical base to provide more complete answers to these questions:
Recommendation 9-1: The U.S. Department of Education should take the lead in coordinating existing data collection efforts and encouraging new ones, with the goal of developing a national education data network that incorporates comprehensive data related to teacher education.
Such a network would provide both baseline monitoring of the status of teacher preparation (and improved opportunities to link that information with other aspects of the public education system) and a common foundation on which to build research efforts that investigate important aspects of teacher preparation.
Ultimately, the kind of network we are recommending would include
systems that provide integrated data within states using common definitions across states;
a short-term national indicator system to monitor the status of teacher education; and
a longitudinal, nationally representative study of teachers’ career pathways beginning with their undergraduate education.
Recommendation 9-2: Researchers and those who fund research related to teacher preparation should focus on topics that have the highest potential effects on outcomes for students, specifically, research that explores the benefits of particular kinds of teacher knowledge and clinical experiences and the factors that affect the quality of entering teacher candidates.
Teacher preparation is a key element in the K-12 education system, not an isolated enterprise. It is affected by and affects every other element in the system. The logic of systemic standards-based reform of public education is very clear in calling for each element of the system to be aligned to consistent state standards. Data collection and accountability at the state level are critical to this alignment as well. Teacher preparation has not yet been brought into this alignment at the state level, but high expectations for teachers and for teacher preparation programs are a critical aspect of an aligned system.
The quality of the nation’s teachers has been the subject of blistering critiques, as have the institutions that prepare teachers. Moreover, the preparation offered to aspiring teachers has long been characterized by inequity in both resources and opportunities. This report begins by highlighting how much teacher preparation matters, both to the long-term success of efforts to improve public education and to immediate outcomes for students. Policy makers, educational researchers, and scholars in relevant fields have shown a growing awareness of its importance and of the gaps in the knowledge base.
The critical questions about teacher preparation cannot be answered without the kind of nationwide coordination we call for. Clearer understanding of the content and character of effective teacher preparation is critical to improving it and to ensuring that the same critiques and questions are not being repeated 10 years from now.