The Education and Evaluation of Effective Teaching: The Continuing Challenge for Teacher Educators and Schools of Education

As science teacher educators we are constantly challenged to provide our preservice and in-service teachers with the best possible knowledge of teaching that will in turn result in maximizing the learning of kindergarten–Grade 12 students. An essential component of any program purporting to accomplish this goal is the development of valid and reliable assessments that provide data that can serve for the revision and improvement of our efforts. Try as we may, our expertise in the area of science teacher education is continually called into question by those who are peripheral to or outside of our professional field. These challenges, for political reasons, often manifest as state-adopted assessments and requirements to which our students and our professional community are held accountable.

The history of the knowledge base for effective teaching can be conveniently divided into six phases of empirical research extending back to the 1920s. The first phase assumed effectiveness to be a consequence of personality traits or characteristics of the teacher, the second phase focused on teaching methods, the third related teacher behaviors to student learning, the fourth focused on the mastering of a repertoire of competencies, and the fifth focused on teachers’ abilities to use competencies appropriately. This focus is labeled professional decision making (Medley, 1979). The sixth phase, which characterizes the current wisdom, has focused on the importance of the interaction of a set of knowledge domains resulting in clearly delineated subject-specific instructional knowledge and skills. This current conceptualization of effective teaching is commonly known as pedagogical content knowledge (PCK) and was first introduced by Lee Shulman (1986a, 1986b).

Presage variables

The earliest research on effective teaching set out to describe the characteristics that differentiated more effective teachers from less effective teachers. Techniques commonly used today to measure mental abilities, personality traits, attitudes, and similar factors were unavailable in the early 1900s, so researchers simply asked students to describe effective teachers they had known (Hart, 1934; Kratz, 1896). The results were various lists of traits attributed to teachers regarded as effective. Some of the most common characteristics included makes greater demands of students, has more teaching skill, has more knowledge of subject matter, and has better discipline. This approach to establishing the knowledge domain for effective teaching eventually moved from using the feedback of students to using the opinions of experts on teaching. During the 1920s the top six attributes of an effective teacher were considered to be good judgment, self-control, considerateness, enthusiasm, magnetism, and adaptability. By 1930, the list of common attributes had been revised to include cooperation (helpfulness, loyalty), personal magnetism, personal appearance, breadth and intensity, considerateness, and leadership. It is important to note that all of these lists characterized teachers perceived as effective. In no instance was any evidence gathered to document that teachers possessing the listed characteristics were actually more effective at helping students achieve educational goals. Consequently, there was a call for more sound research from which a knowledge base for teacher education programs could be derived.

Methods of teaching

The intuitive recognition that what a teacher does is far more important than what a teacher is gave direction to those wishing to establish a reliable knowledge base for teacher education. It was agreed that what a student in a preservice program needs to learn is not what he or she should be but what he or she must do in order to be effective. This notion led to a series of what were known as methods experiments. In a typical experiment two or more classes were taught using different methods, and the mean gains in knowledge of the classes were compared to establish the relative effectiveness of the teaching methods. The results of this type of research were mixed and inconclusive. In retrospect, it appears that most of these investigations suffered from a particularly fatal flaw. That is, they were designed to use the student as the unit of analysis as opposed to the teacher. Consequently, no valid generalizations to teachers other than those actually participating in the investigation could be made (Medley, 1979).

Process–product research

By the 1960s it had become clear that sound research on teacher effectiveness must focus on both teacher behavior and pupil learning. The focus of this research was not on teacher characteristics but on stable behavior patterns that have been referred to as teaching styles or dimensions of classroom climate. The assumed importance of stability was critical to this approach of establishing a knowledge base for effective teaching. Teachers’ behaviors were observed on random occasions, and the researchers searched for patterns of behavior that were stable across observations. These patterns were then correlated with student achievement as measured by a variety of standardized assessments. In addition to educators being able to establish, for the first time, a link between teacher behavior and student learning, it was also noted that many of the behavior patterns could be taught to teachers and/or modified in certain ways. The patterns of teacher behavior and classroom climate initially identified as distinguishing effective and ineffective teachers included clarity, variability, enthusiasm, task-oriented or businesslike, criticism, teacher indirectness, student opportunity to learn criterion materials, and use of structuring comments (Gage, 1963).

Although process–product research can produce, and has produced, reliable information for use in the development of teacher education programs, it is not without some significant problems. First, although this approach to research was based on naturally occurring correlations, effective teaching was determined through an act of synthesis by the investigator in which the individual behaviors associated with learning were aggregated (across time and context) into a new composite. There was little evidence that any observed teacher had ever performed the collective pattern of behaviors specified in the composite. Second, and probably more important, process–product research was openly empirical and atheoretical. It simply focused on what worked rather than why it worked. It was often described as an engineering perspective to research on teaching as opposed to a scientific approach.

Competencies

In the late 1960s and early 1970s process–product research fell into disfavor relative to a new model for teacher education known as competency-based or performance-based teacher education. This model assumed that the effective teacher differs from the ineffective teacher primarily because he or she has command of a larger repertoire of competencies, skills, abilities, knowledge, and so on, that contribute to effective teaching. The list of such competencies was viewed as being extremely long, and it was believed that no one individual needed to possess them all. The model allowed for a variety of ways to be an effective teacher while still maintaining that there was at least a small set of core competencies that all teachers should possess. Competencies were viewed very differently from the stylistic behaviors of the process–product research. A competency is used only under certain circumstances as opposed to being stable across all contexts. The ability to ask higher order questions is a competency, clarity is not. There are times when higher order questions are inappropriate, but there is no time when clarity is not appropriate. This approach to establishing a knowledge base for effective teaching took into account teacher behavior as well as when and why teachers behave as they do.

Professional decision making

Although it could easily be argued that the model of the effective teacher as a professional decision maker is part of the previously described competency model, there is some use in delineating it as a separate phase in our efforts to establish a knowledge base for effective teaching. Most of the early research within the competency model did not focus to a large degree on teachers’ appropriate use of competencies. Indeed, there are those who would say that the competency approach really was not clearly distinguishable from process–product research. However, during the early 1970s there was a clear movement toward the importance of discerning why a teacher chooses to make the decisions he or she does during the preinstructional, instructional, and postinstructional phases of teaching (Borko, Cone, Russo, & Shavelson, 1979). Teachers make numerous decisions every minute of teaching and may make almost as many while planning instruction and following instruction. According to this perspective the effective teacher was viewed as an expert decision maker, and the focus of research was on trying to find out how teachers made such decisions and whether these decision-making abilities could be developed in beginning and practicing teachers. The effective teacher seemed to be particularly skilled in making decisions regarding the maintenance of the learning environment, use of pupil time, and methods of instruction. This approach to establishing a knowledge base, as well as the five preceding approaches, were clearly problematic to Lee Shulman as he studied the process of medical reasoning in the mid-1970s. One finding of his research was that clinical problem solving was domain specific, even though the prevailing medical lore assumed a general trait of diagnostic acumen. He subsequently used what he had learned during his study of physicians while pondering the development of measures to be used to evaluate teachers during the early 1980s. It became clear to Shulman that in the effort to distinguish between effective and ineffective teachers across all contexts, the subject matter being taught had gradually been ignored as a variable in terms of either the teachers’ knowledge or students’ processing of new ideas.

PCK

Although there have been a variety of models of PCK in the past decade, it was initially described as the missing paradigm when it was unveiled as the focus of Shulman’s Knowledge Growth in Teaching research program at Stanford University. The primary purpose of the PCK construct was to bring disciplinary knowledge back into the equation of what it takes to become an effective teacher (Shulman, 1986a, 1986b). Shulman and his colleagues believed that research on teaching and teacher education programs had become too generic. The idea that a teacher must be well versed in his or her subject matter is not earth shattering. But Shulman was talking about much more. It was readily admitted that there was no such thing as the expert pedagogue, a prevailing idea that if a person could teach one subject he or she could teach any subject. It was also readily admitted that just because one is an expert in subject matter it does not follow that this individual will be an expert teacher of that subject matter. What was seminal in Shulman’s initial formulations was the contention that the expert 10th-grade biology teacher possessed a domain of knowledge that neither the expert in pedagogy nor the expert in biology possessed. This individual had very specialized knowledge that enabled him or her to represent the subject matter in the 10th-grade curriculum in a manner that was readily understood by 10th-grade biology students. This knowledge (the pedagogical aspects of subject matter) was called PCK. Shulman’s original formulation, which totally changed how teacher education is viewed, was a combination of six domains of knowledge: pedagogical knowledge, subject matter knowledge, curricular knowledge, knowledge of students, knowledge of schools, and PCK. PCK was considered to be a separate domain of knowledge that was derived from the interaction of the other five knowledge domains.

In the past, concerns about subject matter typically correlated broad and indirect measures of teachers’ subject matter knowledge (e.g., degrees earned, grade point average, courses taken) with some measure of teaching effectiveness (e.g., students’ scores on standardized achievement tests). The knowledge base being developed under the rubric of PCK took a very different approach. Research now focused on detailed information about teachers’ subject matter goals to be pursued with students, content selection and representation choices made during the development of instructional plans, and the PCK underlying these decisions. The emphasis was no longer limited to what teachers knew about their subject, it was expanded to include what they wanted students to know and how they proposed to facilitate the development of student understanding. Additional research has more recently focused on how teachers’ decisions about what and how to teach are determined by teachers’ beliefs about the subject matter.

As a consequence of the recognition of the importance of PCK, the teacher education community has generated a knowledge base that clearly identifies the intimate relationship between the nature of subject matter, knowledge and perceptions about subject matter, and the teaching of the subject matter. Numerous investigations have clearly indicated very complex interactions between subject matter and pedagogy. The result has been a proliferation of subject specificity in the education of both beginning and practicing teachers. The emerging research has no longer made tenable the idea that the professional development of teachers is generic. How one learns to teach is directly influenced by the individual’s knowledge of and perceptions of subject matter. Furthermore, the knowledge and perceptions are directly influenced by considerations of how to teach the subject matter as well as the actual act of teaching. This current focus of research on teaching (i.e., PCK and the other five domains of knowledge critical for teachers’ success) has been credited with professionalizing teaching and providing the knowledge base to productively reform how we educate teachers.

It is clear from this historical backdrop that how effective teaching is conceptualized is in a constant state of flux. The act of teaching is as complex as any other human activity. In the past three or more decades a variety of models have claimed to be worthy of guiding teacher education programs and the assessment of such programs. Only three are mentioned here to illustrate an important observation. In 1979, Madeline Hunter claimed to have synthesized the research on effective teaching under the label of instructional theory into practice (ITIP). Perhaps the most well-known part of the ITIP approach was its emphasis on a particular lesson design. The approach highlighted six essential, but generic, components for a successful lesson: objectives, anticipatory set, modeling, guided practice, independent practice, and closure. The approach seemed well suited for basic skills, and it was popular with administrators because knowledge of subject matter (by the evaluator) was not necessary. Although ITIP was presumably based on research on effective teaching, the overall results of instruction following this model were not particularly effective. The individual components of the model may have been research based, but the combination of the components was not.

More recently, Charlotte Danielson’s (2007) Framework for Teaching has gained some traction across the United States. We mention it here because it has had an enormous impact (in terms of adoption) in our home city of Chicago. This model of instruction and the evaluation of instruction contains four domains: planning and preparation, classroom environment, instruction, and professional responsibilities. It is presumed that systemic attention to these four areas will facilitate the development of effective teaching. Although popular in some locations, the framework has not documented significant improvements in student learning. Nevertheless, teachers are evaluated annually using a rubric based on Danielson’s conceptualization of effective teaching.

These previous models of effective teaching pale in comparison with the current wave of acceptance of edTPA (Teacher Performance Assessment). Spearheaded by Pearson Testing, edTPA Secondary Science (Stanford Center for Assessment, Learning and Equity, 2015) provides a nationally available performance-based assessment to measure novice teachers’ readiness to teach science at the secondary level. The assessment is designed with a focus on student learning and principles from research and theory. It is based on findings that successful teachers do the following:

• Develop knowledge of subject matter, content standards, and subject-specific pedagogy

• Develop and apply knowledge of varied students’ needs

• Consider research and theory about how students learn

• Reflect on and analyze evidence of the effects of instruction on students’ learning

As a performance-based assessment, edTPA is designed to engage candidates in demonstrating their understanding of teaching and learning in authentic ways. edTPA has been adopted by 15 states, and another 25 states are participating with edTPA at some level. Thus, it is having a direct impact on teacher licensure, teacher education programs, and teacher evaluation. Adoption of edTPA and models like it has significant impacts on finances and workload. Again, as with other views of effective teaching, efforts have been made to document the validity of the included components of effective teaching. However, the question remains as to whether individuals who are faring well according to edTPA actually are more effective. Do these teachers discernibly increase student learning? In short, does edTPA and other programs like it have predictive validity? Although only three models (for sure there are many more) of effective teaching have been mentioned here for the sake of brevity, the same conclusions can be reached for those that have not been mentioned. To be blunt, after all of the history of research on effective teaching that has been outlined here, do we really have an effective way of developing and assessing effective teaching?

There is an incredible cumulative wealth of expertise concerning the effective preparation of science teachers and their professional development in our organization (i.e., the Association for Science Teacher Education), the professional community, and colleges/schools of education. See, for example, Studying Teacher Education (Cochran-Smith & Zeichner, 2005). Consequently, it is often frustrating to find those with political and financial leverage having a more prominent role in policy concerning the field we know so well. It is our professional responsibility to be proactive and not just accept what comes our way. We need to actively research the effectiveness of programs such as edTPA relative to teaching effectiveness. Again, does alignment with edTPA result in higher performing teachers? We may find that edTPA is spot on, or maybe not. However, at least we will have an answer or—more realistically—be in a position to make more informed decisions concerning the future of our profession.