EDUCATIONAL PSYCHO LOGIST. 24~4). 325-344
Copyright 1990, Lawrence Erlbaum Associates, Inc.

Process-product research on teaching, an adaptation of the criteria-of-effectiveness paradigm, has its roots in functionalist psychology. Functionalism, is a pragmatic and electric approach to psychology that is concerned with the fitting of persons to environments. Functionalists have always accepted cognitive behavior as a legitimate object of study. Thus, a perspective from functionalist psychology may be useful for designing studies of teachers' thinking. From this perspective, the majority of contemporary studies of teachers' thinking are found wanting because there has been no concern for criteria of effectiveness. Recommendations for the improvement of research on teachers' thinking are offered.

It is important that those who develop the research agenda for studying teacher thinking, a contemporary program of research on teaching, should profit from the lessons learned while engaged in an older style of research on teaching, often called process-product research. The older research program has often been linked with the behaviorist tradition in psychology, whereas the newer research program is associated with cognitive psychology in particular and with the exciting new field of cognitive science in general. To discuss the lessons that have been learned, however, I must first share with you my own idiosyncratic view of the two apparently disparate research traditions or programs. A personal view of these issues is offered in the hope that it will contribute to the dialog-the "grand conversation" that Shulman (1986) described-about the future course of research an teacher thinking.

The studies of teachers in previous decades that were called process-product research should not have been automatically classified as traditional Skinnerian behaviorism. Also, contemporary research studies do not automatically fit within the broad research program called cognitive psychology. Much of the old and new work is, I believe, in the finest tradition of a much older research program in psychology called functionalism, which has its roots in Darwin, in the philosophy of pragmatism, and is generally conceded to have been started by John Dewey (and others) around the turn of the 20th century. The "Chicago school of psychology," as functionalism was sometimes called, was a loose confederation of people and ideas. The functionalists, therefore, were eclectic. Boring (1957) once labeled functional psychology as "what a psychologist does when free of systematic compulsions" (p. 559). They borrowed ideas and methods from all their contemporaries - Pavlov, Watson, Thorndike, and the Gestaltists. As applied scientists they were, as Boring (1957) stated, "pragmatists; they take what works and accept those systems which provide convenient terminology for new general principles" (p. 570). Thus, the contemporary, research, programs of both behaviorism and cognitivism, can supply the functionalists with concepts and methods for thinking about how to do the particular kind of work they want to do. The kind of work functionalists want to do has remained unchanged for about a century. That work is the establishment of relations between important characteristics of the environment and the thoughts and actions of human beings. What should be noted, however, is that one can borrow concepts and methods without taking on the ideology of the program, of research that does the lending.

Functionalism was given its most articulate formulation by James Rowland Angell in his presidential speech to the American Psychological Association in the early years of this century (Angell, 1907). It came 6 years before Watson's famous paper that announced behaviorism and rejected the study of mental life (Watson, 1913). Watson, it should be noted, was Angell's student and a functionalist in outlook. His contribution to the evolution of scientific psychology, however, was to jettison the mind as a necessary concept in the determination of functional relationships between an organism and its environment. Watson narrowed the new functionalism that Angell had announced only a few years earlier and, by so doing, led psychology down a much narrower path than it needed to go. One consequence of the ascendence of the Watsonian view of psychology was that it paved the way for Skinner and the radical behaviorism associated with that research program.

Like Watson, Angell insisted on using the new indigenous American form of experimental psychology, and recommended avoiding, the methods employed by structuralists such as Wundt and Titzhener. Nevertheless, Angell was quite clear about including the study of mental life in the functionalist research program. The functionalists were interested in the how and why of mental operations (their functioning) more than the what of mental life (the content of conscious experience). The latter was the concern of the structuralists.

Because environments clearly change, habits that once were successful become inappropriate. Thus, the problem-solving skills and other mental activities used to direct new, more-relevant behavior were clearly important for functionalists to study. It was the functionalists who fostered the enormous number of verbal learning studies conducted from the 1930s to the 1960s (see McGeoch & Irion, 1952). It was the functionalists who fostered the work on transfer and mediation in learning (see Osgood, 1953). These areas of psychology, though perhaps not studied in ways we now would consider fruitful, were not following the same research programs that were characteristic of the more well-known behaviorists of that time, such as Hull, Spence, Skinner, or even Tolman.

Many functionalists preferred human studies to animal studies. They were concerned with observable verbal behavior, including meaningful verbal behavior, and cared much less about movement or act psychology. They were linked to the British associationist tradition that was both a simple cognitive psychology and a simple learning theory. These functionalists saw reinforcement as merely another variable that affects performance, like massed or distributed practice, rather than as a cornerstone for a theory of learning. Thus, the functionalists were inescapably cognitivists but of a different sort than those of today. Although they looked for relations between input and output, between stimulus and response, they never denied that what was in between the two was important! Furthermore, the functionalists always believed that the higher cognitive processes were amenable to study. They believed that, through correlational and experimental work, reliable relations (including expressions of mathematical functions) would be established between thoughts and actions, on the one hand, and measures of adjustment to the environment on the other. Their emphasis on adjustment to environments led them to propagate applied psychology. In this way, they could study the adjustment of people in special environments. Thus, it was the functionalists who, at the start of this century, promoted educational psychology, industrial psychology, personnel psychology, and the mental health professions. In each area of human concern, such as the school or the workplace, the psychologists' job was seen to be the same, namely, to find out what thoughts and actions went with measures of adjustment to the environment.

Functionalism and the Criteria-of-Effectiveness Paradigm

In education and business, the functionalists' goals were very often transformed into what was called the criteria-of-effectiveness paradigm (Gage, 1963), which encouraged researchers to find a criterion of interest, one representing a good adjustment of the person to the environment (e.g., a measure of successful sales in insurance, telephone operating, piloting of an aircraft, or classroom teaching) and to find predictors of that criterion of effectiveness. The relations between predictors and criteria of effectiveness could then be used to select people who would fit that environment better than would a sample of individuals chosen at random. The empirical relations could also be used to develop a curriculum for training people to work effectively in that environment, or the information about relations could be used to select and train individuals to succeed in a particular environment. In education, the criteria-of-effectiveness paradigm, an integral part of the functionalist psychology program of research, became known as the process-product approach. This is really its roots -a special case of the functionalist psychology of an earlier era.

Unfortunately, the process-product approach to research on teaching became linked to the mechanistic view of humanity put forth by the Skinnerian style of behaviorism. It was also linked to the anticognitive views of Skinner and most other behaviorists. With a little better luck, a slightly different emphasis, a few different key writers or studies to guide it, process-product research would have been linked to a more purposive view of humans as adjusting organisms, in the best functionalist tradition. In that tradition, cognition was already seen as a factor in adjustment, and it was seen also as a legitimate area of study. If those working in the contemporary program of research on teacher thinking could put aside the negative feelings they have developed for process-product research and look instead at what could be learned from the criteria-of-effectiveness research paradigm emanating from the functionalist research program, they might find the effort expended to be worthwhile.

Functionalism is my research tradition. It always has been, and I see no reason to change. That is what I meant when I said the goal of my research, across the years, has remained the same. I may be the last living functionalist. I am an unabashedly applied psychologist interested in the selection and training of teachers to be successful in a particular kind of environment - the public school classroom. With this goal, using the criteria-of-effectiveness approach seems undeniably sensible. In fact, it should be noted that the, criteria-of-effectiveness approach to selection and training was considered so successful by Cronbach that, in the first edition of the Essentials of Psychological Testing (see Cronbach, 1949), he called it one of the outstanding achievements of social science. The criteria-of-effectiveness approach to research-this outstanding, achievement of functionalist psychology-can be used as a framework for examining the contemporary research on teacher thinking (e.g., teacher planning, teachers' interactive thoughts and decisions, and teachers' theories and beliefs). Reviews by Clark and Peterson (1986) and by Clark and Yinger (1987) are the primary sources of literature on these topics. Recommendations that are applicable to the contemporary program of research are discussed later on in this article.

Functionalism and the Study of Teacher Planning

Planning is a central topic of research on teacher thinking. Clark and Yinger (1987) defined it as "a basic psychological process in which a person visualizes the future, inventories means and ends, and constructs a framework to guide his or her future actions-what Yinger calls thinking in the future tense" (p. 345). Both the focus of teacher-planning studies and the methodology used in these studies would be abhorrent to a radical behaviorist, but they are perfectly acceptable to a functionalist. The functionalist, however, would quickly, ask: For what purposes and to what ends are planning activities directed? Thus, the descriptive studies of the types of planning that are done by teachers are not valued in and of themselves but gain their value when the functions of the types of planning activities can be understood. Therefore, learning that teacher planning early in the year focuses on the physical environment and social system of the classroom, or that much planning is not recorded on paper, or that teachers depend on published teachers' guides for planning instruction, or that the "activity" is the basic unit and starting point for planning is only potentially important information to those who study teaching from the functionalist tradition.

In the functionalist research tradition, one is bound to ask a very practical set of questions once the revelations about the types of planning activities are revealed. For example, why is the activity the unit for planning? What is its function? Compared to teachers who plan with the activity as their focus, how would people who plan by using content, objectives, or the individual student as their focus fare in the classroom environment? Using the criteria-of-effectiveness approach, a functionalist would ask: What do different foci of planning yield in terms of different measures of effectiveness, such as student satisfaction, teacher satisfaction, student engagement, smoothness of classroom processes, student achievement, and other easily named and important criteria of effectiveness? For other revelations about the types of planning used by teachers, we would ask, quite early in the development of a research program, questions about what function is served by, for example, a reliance on textbooks? Why are plans often not written down? Why do teachers focus on the physical environment and social system of the class at the beginning of the school year? These are only some of the questions a functionalist would be inclined to ask.

Using the criteria of effectiveness as a guide for generating practical research, we propose a correlational and experimental research agenda that inquires whether there is any relation between higher and lower reliance on textbooks as guides for planning, on the one hand, and such things as measures of teacher confidence, student attitude toward the subject, student test performance, and other measures of effectiveness, on the other hand. In a similar manner, a correlational and experimental research agenda would be built around questions about the relation between plans that vary, say, from unwritten to fully specified, on the one hand, to criteria of effectiveness on the other hand. The criterion of interest might be the degree of alignment between the curriculum and the tests used as outcome measures, or teacher confidence, student knowledge of goals, or teacher spontaneity. The latter issue was dealt with, in a way in Zahorik's (1970) study. He apparently demonstrated the possibility of a reduction in teacher spontaneity as a function of well-specified planning activities. This study received a great many citations. Perhaps this was so, in part, because it was a study in the functionalist tradition, using the criteria-of-effectiveness approach, from which we derive information with patently obvious training implications. Uncovering findings that have training implications for a particular field is what applied psychology is about.

The knowledge about the planning activities of the people in a field take on a special significance when that planning can be related to some criteria of effectiveness. For example, using a different context, we may cite a well reviewed book by Halberstam (1986), The Reckoning, in which he discussed the decline in world markets of the American automobile industry. The major, and once unthinkable decline in sales (a criterion of effectiveness), was traced to planning activities of a particular type. The American planners pursued short-term goals and profitability for the maintenance of stock prices. The Japanese, Halberstarn noted, did long-term planning and picked consumer rather than stockholder satisfaction as their goal. The type of planning (i.e., long- or short-term planning dealing with either consumers or stockholders) done by these business leaders is certainly worth knowing about, as are the types of planning activities used by teachers. The function of a type of planning activity and the relation of a type of planning activity to a criterion of effectiveness however, gives it special meaning to policy makers and to those who select and train business executives or educational personnel.

Some research on teacher planning has been conducted in this way. These studies provided evidence that such an approach can be useful and would have an appreciative audience. Carnahan (1980), for one, showed that many of the criteria of effectiveness concerned with classroom processes (e.g., calling on a broad range of students or using motivational statements) probably are unrelated to planning activities. Planning does not usually relate to such interactive behavior, rather, it is related to structural characteristics of the actual lesson such as the content, teaching activities, grouping patterns, materials to be used, and the like. When we put together our understanding of the functions to be served by planning with this research finding, we have enough of a theory and enough of a relation to be useful. With this knowledge, selection and training could be improved, and that could lead to better fits between persons and environments. That is the goal of applied psychologists and of functionalists.

So, in education, we should all delight in the new and creative taxonomic work occurring in the study of teachers' planning behavior. However, the research agenda that has special significance, which should follow closely on the heels of that taxonomic work, is, I believe, one that analyzes the function of this planning characteristic and its relation to measures of effectiveness that we value.

Another body of literature forming a part of the program of research on teacher thinking and teachers' interactive thoughts and decisions may also, be analyzed by a functionalist, using a criterion-of-effectiveness approach. Thus far, we have learned from such studies that teachers' thoughts during interactive teaching are directed primarily toward students. Teachers' concerns about content, materials, objectives, and the like often fade or are redirected into concerns about students' learning. The students, as well, provide the major source of cues that give rise to interactive decisions. In this vital research area, Clark and Peterson (1986) suggested the development of a matrix to help guide future research. They suggested crossing certain kinds of cognitive processes that have been found to be used regularly (e.g., perceiving, interpreting, reflecting, and so forth) with content categories found in teachers' interactive thoughts (e.g., students, objectives, materials, and so forth). Such a matrix could be used to direct researchers to more fully describe the nature of each process-by-content cell A functionalist, however, would immediately ask a set of questions about that matrix that others might not ask right away. What is the perceiving for? What does it accomplish? Why are some things perceived and not others? What function is served by the overwhelming concern for students shown by teachers during interactive teaching?

Long before the matrix is fleshed out, functionalist concerns, which are practical concerns, would come creeping into this research program. For example, we are learning from a number of sources that expert and experienced teachers perceive anomalies, or atypical events, quite a bit better than do less expert, less experienced teachers (e.g., Berliner & Carter, 1986). The expert teachers we studied simply did not perceive ordinariness at all. It was, apparently, not functional to do so. It did not seem to add anything to the teachers' ability to predict the course of events in the classroom-a very important skill-and, therefore, was of little value to them. Ordinariness, sameness, or ongoingness in classroom life requires neither attention nor the making of a decision (except perhaps to stop or change activities). Only the perception of atypicalness, irregularity, or out-of-order events concerning students requires attention and, perhaps, some decision. These studies allow us to develop a sensible little theory concerning the function of the perception of cues about student behavior, explaining why perception is the primary mental process used during interactive teaching and why students are the primary content of thinking during interactive teaching. This theory, however, is laden with ideas whose origins stem from information processing and cognitive psychology. These ideas may be kept in the back of our minds as we cross over to a different research tradition, one that relies on classroom observation, and enter the heart of the process-product research program in order to bring to the forefront some work by Kounin (1970).

Kounin gave us the amusingly named concept of "withitness," defined, half seriously, as the teacher's ability to perceive events before they happen. That is, to catch events that could blow up into problems before they mushroom out of control. Kounin related the teachers measured withitness to student engagement and the classroom's freedom from major behavior problems. Other researchers went on and related withitness to student academic achievement (e.g., Borg & Ascione, 1982; see also Brophy, 1983). This quintessential process-product finding and the research program of which it was a part has been the driving force behind dozens of studies and has fostered many programs, for selecting, identifying, and training teachers (see Emmer, Evertson, Sanford, Clements, & Worsham, 1984; Evertson, Emmer, Clements, Sanford, & Worsham, 1984). Kounin's research is popular because it falls squarely in the functionalist camp, using the criteria-of-effectiveness approach, for which there is a large and eager market. Practitioners in applied fields want knowledge about relations between the things people do and criteria of effectiveness so that they can learn to do what they do better. Kounin's finding supplies this knowledge. Although it comes from the heart of process-product research, it is, in fact, a kind of a cognitive concept. It is the name for the kind of mental activity of teachers we found in the matrix when we crossed the mental process of perceiving with the content category of interactive thought called students. The function, then, of that kind of thinking is to be "withit," to perceive anomalies in classroom life early enough to stop them from getting out of hand. An understanding of the functions of such thoughts, when melded with replicable relations between that kind of thinking and highly valued criteria of effectiveness, gives the relationships found special significance.

Those who do cognitive research in the program of research concerned with teacher thinking may take credit for their part in uncovering the fact that a good deal of teachers' thinking is concerned with the perception of behavioral information emitted by students; those in a behaviorist, process-product tradition may take pride in their replications and elaborations of Kounin's finding of a relationship. It is the functionalist, however, who takes pride in seeing that the two apparently disparate strands of research are a whole, providing all that is necessary for designing better matches of people and environments. We have enough theory to satisfactorily understand the relations we find and, therefore, have enough knowledge to develop ways to help teachers make better adjustments to the complex and dynamic environment of the classroom. Furthermore, the information can help us design environments in which teachers can function more effectively.

There is no doubt that continued descriptive and taxonomic work to foster better understanding of the teachers' interactive thoughts and decisions is highly desirable. This line of research is new, and the body of knowledge is small and fragile; thus, the scholarly community should promote ways to help this line of research flourish. It is argued here, however, that the establishment of relations between interactive thoughts and decision making, on the one hand, and criteria of effectiveness, on the other, are what will lead to the greatest growth of knowledge in this field. Functionalism, which concerns itself with practicality and applied research programs, has a certain valued place in a field of inquiry like education. Functionalist concerns can, and often should, drive the research agenda. The almost inescapable demand for this kind of research was noted by Clark and Peterson (1986), who turned to these topics near the end of their review of teachers' interactive thoughts and decisions. They discussed a criteria-of-effectiveness approach in the review, a functionalist way of thinking, and, perhaps, could not do otherwise; it is demanded by the field. When building a research agenda in this area it might be desirable to make more conscious efforts to take these demands into account.

The third area of research on teacher thinking, that of beliefs and attributions, has had a built-in kind of concern for the usefulness of the data that have accumulated. Attribution training began right on the heels of the discovery of different types of attributions. The Teacher Expectancy and Student Achievement program has been used in hundreds of school districts, with general satisfaction. It is loosely based on the teacher expectancy literature. The criteria-of-effectiveness approach was evident in the original Rosenthal and Jacobson (1968) study of expectancy and has permeated that field of inquiry. The function of an expectancy and its relation to behavior is now well understood (see Cooper & Good, 1983). This is precisely the kind of information that people, particularly school people, want. In this area of research, people now can find an acceptable theory accounting for scores of reliable findings. This is the kind of research program expected of applied researchers; it is also in the mainstream of the functionalist research program. This may be one reason why expectancy research is so popular.

The flavor of a functionalist critique of research on teacher thinking should now be clear. The critique provides the background for considering recommendations that might prove useful in building, the research agenda for the study of teacher thinking across a wide variety of areas. Three of many possible recommendations are presented next. As might be expected from a functionalist, all three are designed to bring the process-product research program and the research program on teacher thinking closer together.

If one desires, a bit of the criteria-of-effectiveness approach can be incorporated, quite easily, into building the agenda for research on teacher thinking. For example, when doing research, a great deal is learned by studying teachers that are known to differ in systematic ways. One approach to developing contrasts between teachers is to find expert and novice teachers, another is to use experienced and inexperienced teachers, and still another is to use effective and ineffective teachers. After the identification of such groups, one can study the planning of these contrasting groups; their interactive thoughts and decisions; their beliefs, attributions, knowledge of students; and so forth. Research on teacher thinking, using stimulated recall and think aloud techniques, is almost always done with small samples because it takes so much time to work with a single subject and to analyze one subject's protocol. Even with small samples, however, contrasts can be useful, as already seen in about 10% of the studies of teacher thinking completed thus far (Clark & Peterson, 1986). In process-product research, some of the more fruitful studies have used this design. For example, the use of contrasting samples worked well in classroom studies of management, whereby the behavior of a small number of teachers who had well-managed classes was contrasted with the behavior of a few teachers whose classes were chaotic (e.g., Evertson & Emmer, 1982). A small number of teachers who were more or less effective in producing gain scores on achievement tests were studied in the Beginning Teacher Evaluation Study with great success. That is, relations of interest were found even with small samples (Berliner & Tikunoff, 1976). The literature describing effective schools, ranging from Edmonds (1979) to Rutter, Maughan, Mortimore, Ousten, and Smith (1979), also shows a reliance on the method of contrasting groups. In fact, the entire research program had its origins in a few contrasts between more and less successful schools in which similar kinds of students were enrolled. If contrasts are built into the design of a study, when differences are found, one immediately has a hint of a relation. For airplane pilot selection, this simple approach has, resulted in their savings of millions of dollars. For example, the, responses of experienced pilots with high reputations among their peers differ enormously from those of novice pilots to questions such as: "What do you do when the landing gear won't come down?" or "Tell me what you think about landings?" The experienced pilot may say, "When the landing gear won't work, I open the emergency procedure book to the appropriate page and follow instructions." The experienced pilot may, also say that landings are fun. The novice often says things like, "When the landing gear won't work, I shut off the pump" or "I call for help" or "I prepare for a crash." The novice reports that landings are "the scariest part of a flight," or in the words of someone I would rather not fly with, "landings are controlled crashes."

Some novices, however, think more like some experts. It is those novices that have proven to be the best candidates for pilot instruction among the big commercial airlines. Because instruction in that field costs hundreds of thousands of dollars per pupil, enormous amounts of money are saved by a simple, valid, contrast approach that allows a personnel director to hire those who most think like the criterion group. Somehow it seems "forced" to talk of this kind of applied industrial/personnel psychology, concerned with pilot selection and training, as process-product research; it is hard to attribute to it all the negative connotations associated with such research (behaviorist, mechanistic, anticognitive, etc.). Perhaps it is simply an intelligent use of the criteria-of-effectiveness approach for finding people likely to succeed in a stressful, highly demanding working environment. The study of teaching could also be well served by this approach, when it is possible.

Research on teacher thinking can profit from this contrasting group approach as well as other effective measures of making contrasts. A basic tenet of process-product research (and, for that matter, all of experimental psychology) is that you should disaggregate data if you expect differences between groups. Gender, IQ, and social class are common ways to separate people because we have found that such distinctions affect performance in a wide variety of tasks. Researchers in the area of teacher thinking should probably also consider this same factor. We already have considerable evidence that a whole set of beliefs about instruction, students, outcomes, and so forth are associated with the different kinds of preparation for teaching that occur in different subject matters. Science teachers, language teachers, and teachers of the humanities all show discernable differences in the categories of things they think about. They hold different beliefs about the purposes of schooling and the roles of students (Yaakobi & Sharan, 1985). To combine data from these different groups, when studying teacher thinking, would definitely lead to errors. It would be better, even with small samples, to contrast these groups. The experimental psychology literature discusses this as a concern for the identification of moderator variables (see Smith & Glass, 1987). Those who engage in research on teaching are likely to learn more by contrasting, groups than by continuing to use an approach consisting of mixed samples.

This recommendation, like the first, is concerned with who should be in the small samples used to study teacher thinking. The question to ask now is: Why not abandon research on the thinking of ordinary teachers and devote more effort to the study of people who are more likely to be worth studying? It is a bit embarrassing to admit, but the thinking of ordinary people usually bores me. Furthermore, it may be misleading if the data garnered from them are used in selection or for training. Very few people I know want to read any of the thousands of biographies of minor officials of, say, the British expeditionary forces in the Sudan in the late 1800s. The Victorian era spawned a great number of such biographies, as each minor player in history put forth, for posterity, his or her views of the heathens and the inevitable wars that were fought. These biographies are, as a group, boring. Worse, yet, they are really uneducative. Instructive biographies, boring or not, may occur when the person is important, interesting, or unique. The biographies of Ghandi, Roosevelt, Khadafy, Hitler, and Helen Keller all have a chance to educate. The biography of for example, my mother is much more limited in what it can teach us. So it is, I think, with small studies of teachers' thinking.

The study by Elbaz (1981) comes to mind because it is often cited in the field of research on teacher thinking. We need not be overly worried about it being a study of a single subject, because it was a lengthy and intimate probing of the depths of thinking engaged in by that one person. Some of the insights obtained are worth serious consideration, though I believe those insights to be the result of a literary rather than a scientific investigation. Neither sample size nor the scholarly tradition in which the study is placed makes me reject Elbaz's view of teachers' thinking as having 75 facets. (Elbaz presented 75 cells for us to think about. She found a way to classify teacher thinking as the result of crossing five content areas of teachers practical knowledge with five orientations to practical knowledge and three structural forms of practical knowledge.) What scares me most about such studies is that a reader obtains no assurances that the one and only high school teacher who was studied extensively was anything but ordinary. In fact, the subject of the study quit classroom teaching at the end of the investigation. Should her thoughts be used to launch a dozen confirmatory research studies? In a more general context we might ask: Is it a wise expenditure of our minimal resources to do studies of the thinking of ordinary people? Such studies can so easily mislead us.

I think of my mother, again, and how she would be described if someone had studied her thoughts about planning to cook and the thoughts and decisions she made while cooking. She spent enormous amounts of time in and around the kitchen, shopped carefully and often, talked about recipes with neighbors, carefully watched pots simmer, and boil, and served up large portions of food to a hungry family whose individual preferences were always kept in mind. Because she was very articulate, she could describe every aspect of the cooking process in detail and with panache. However, we had a problem with my mother. She was a terrible cook! She messed up the vegetables daily, and she ruined hamburgers and chicken once each week. A record of her thoughts would surely be misleading!

In industry, the study of the "best" executives or "top" corporations is in the functionalist tradition of trying to use a criterion of effectiveness to ferret out relations with the criteria that might exist. The incredible success of this approach, in the eyes of the business community and the lay public, was noted a few years ago from sales of the book In Search of Excellence: Lessons from America's Best-Run Companies by Peters and Waterman (1982). This book was a hardcover, best seller for over a year and became a paperback best seller as well. It has influenced the management of schools and industry worldwide, yet it is nothing more than our old, sensible, functionalist, criteria-of-effectiveness paradigm, brought in to study an applied phenomena of interest. The useful educational version of this approach to research is seen in studies of commonalities among effective schools (where no contrasts with ineffective schools are made) and studies of common characteristics of effective teachers, without necessarily comparing them to less effective teachers. The selection of the best teachers, on whatever criterion one values, is a practice worth considering when preparing to study teacher thinking.

This recommendation calls for a reverse in direction. Instead of looking at how thought and action are related, perhaps we might look more frequently at how action and thought are related.Process-product research, our indigenous form of the criteria-of-effectiveness approach in research on teaching, has provided, a shopping list full of teaching characteristics thought to be important within that research tradition. Time management, for example, is, generally, believed, to account, for at least 10% of the variance in achievement test scores in the public schools, and that is granted even by those who consider time to be a trivial variable in the overall order of things (see Fisher & Berliner, 1985; Karweit, 1985). How do teachers think about allocated and engaged time? How do they decide to start or stop activities? In what ways do the clock, the schedule, and student attention, influence interactive decisions about what to do and for how long? Studying the thinking of teachers with regard to issues of apparent importance appears to be a worthwhile endeavor.

Another example is easily found. We know that mastery programs, in research study after research study, when compared to traditional instruction, show enormous positive effects on achievement and student attitude (e.g., Block & Bums, 1976). We hear as well that, in practice, mastery programs often fall apart in Grades K through 12 in the public schools. My hypothesis, though as yet untested, is that teachers' beliefs about their roles enter in here. The teachers' conceptions about what is "proper" student behavior is challenged in mastery approaches (e.g., ability is a time variable not a genetic quality of the individual), and their capacity to manage heterogeneity - 30 individuals proceeding at their own pace--is also strained. In short, I think the research from process-product investigations of mastery, which is so positive at first glance, needs to be understood not in terms of its effect on student's performance but, rather, in terms of its effect on teachers' thinking! It is teachers' thoughts about mastery, and its associated characteristics, that probably need to be investigated more thoroughly than the outcomes or products of that method of instruction. We do not need another experimental study demonstrating that mastery is effective. However, we do need more knowledge of teachers' thoughts while doing mastery programs, and apparently, also while doing them in. It would be helpful for teachers' educators to have this kind of information.

There is an example of what happens when the recommendation to study things that we already know are important is followed. A team at Michigan State University (Schwille et al., 1983) studied teachers' content choices, a cognitive concern, related to the well-established process-product finding that content coverage is a consistent predictor of achievement (e.g., Berliner & Rosenshine, 1977). This research is noteworthy because it is one of the few examples that blends the two research traditions so well.

Teachers are often given the opportunity to learn apparently important and useful skills, methods, and concepts but make personal decisions to use or not to use them. Why? We do not yet know. Why is it, for example, that some 16 years after Rowe (1974) revealed the fascinating findings associated with "wait time" during recitation, that we have no knowledge of why teachers ignore the simplest of all process-product findings? Dozens of replications have occurred (e.g., Tobin, 1987), and still there is scant evidence that practitioners will implement this reliable finding. The point is that many, process-product findings, are not implemented when learned. They apparently are not functional, or, one would assume, that they would be used more often. A set of studies, therefore, of great utility, could be designed to work back from process-product relations to teachers' thoughts about the context and other variables associated with the empirical relations that were found. This would satisfy the functionalist who needs to know how certain behavior is interpreted by the performers of the behavior, in order to better teach (or abandon teaching) those behaviors.

Some of my research on instructional time has been criticized as, trivial, because time is an empty vessel. Such critics say that it is what fills time that is important to study (see Berliner, in press). The same partially silly, partially true criticism can be leveled at those who study teacher thinking. Thinking must be about something. It is like an empty vessel if there is no object of thought. Thus, it is recommended that one of the things to study is teachers' thinking about phenomena that have been found to be important, at least found to be important from the perspective of the process-product research program.

The source of the information, the process-product research program, should not cause a researcher concerned with teachers' thinking to reject the information about a relationship between teachers' actions and some criterion of effectiveness. Dewey (1916, p. 120), in Democracy and Education pointed out that to act purposefully, to act with an aim, is the very essence of intelligence. He said that:

A man is stupid or, blind or unintelligent -lacking in mind,- just in the degree in which, in any activity, he does not know what he, is about, namely, the probable consequences of his acts. A man is imperfectly intelligent when he contents himself with looser guesses about the outcome that is needful.

Process-product research, even with its defects, provides some information of value about how things might turn out. It is, as Dewey noted, a sign of imperfect intelligence to merely guess about the outcomes of events when reasonably reliable information about relations is available. The relations from the process-product research program can be used to inform a planner so that a reasonable estimate can be made of the probable consequences of Actions A or B or C. The functionalist, armed with the criteria of-effectiveness approach to the world, tries to provide the relations that decision makers can use, so that mere guesses about the outcomes of certain actions are not always needed when attempting to act purposefully. Learning how people think about the relations that are uncovered changes those relations from mere empirical findings to findings embedded in webs of meaning. It is in this way that we know so much more about them.

Process-product research is the way that the criteria-of-effectiveness approach is carried out in research on teaching. Its roots are in functionalism and should not automatically be considered a radical behaviorist research program or inherently anticognitivist. What process-product research reveals, with all its shortcomings, is worth thinking about. In fact, it is not really sensible to maintain the split between the two research traditions based on some partially true conceptions of what the other is about.

I was told of a wonderful interchange that illustrates this point (L. S. Shulman, personal communication, April 1, 1996). It is said that N. L. Gage, a key figure in the process-product research tradition, was talking to Lee Shulman, a key figure in the program of research on teacher thinking. Gage worried that Shulman's teachers were "too often lost in thought." It took some time for Shulman to reply, but eventually he noted that "it is better for his teachers to be lost in thought than missing in action." I would enter this disagreement between my two dear friends and colleagues by pointing out that thought and action are so inextricably linked, that it is really a fiction to maintain their separateness.

Functionalism, concerned as it is with the reasons why behavior occurs, has recognized the inseparability of thought and action from the time it began as a psychological system. Dewey's (1896) famous reflex arc article pointed out the indivisible nature of the, stimulus (which. could be a thought, or an action) and the response (which also could be a thought or an action). In a complex serial process, extending over minutes or months, What is stimulus? What is response? Does one ever know what thoughts stimulate action, what actions stimulate thoughts? How can only really wrench apart ongoing behavior? Thus, functionalism, as a research tradition, attempts to unify research in teacher thinking with research on teacher action, because from its inception it could not distinguish between the stimulus and response, seeing behavior, instead, as ongoing and purposeful.

That teachers' thinking and actions are inseparable is also recognized by the title of recent meetings of the International Study Association on Teacher Thinking. The first few meetings of the association were called conferences on teachers' thinking. More recently these have been titled conferences on Teacher Thinking and Professional Action. As a functionalist, my response to this change is to note that this was appropriate behavior, I think.

An earlier version of this article was presented as an invited address at the third conference on Teacher Thinking and Professional Action, sponsored by the International Study Association on Teacher Thinking, Leuven University, Leuven, Belgium, October 1986.

Angell, J. R. (1907).The province of functional psychology. Psychological Review, 14, 61-91. Berliner, D. C. (1969). Microteaching and the technical skills approach to teacher training (Tech. Rep. No. 8). Stanford, CA: Stanford University, Stanford Center for Research and Development in Teaching.

Berliner, D. C. (1986). In pursuit of the expert pedagogue. Educational Researcher, 15(7), 5-13.

Berliner, D. C. (1989). Implications of studies of expertise in pedagogy for teacher education and evaluation. In Educational Testing Service (Ed.), New directions for teacher assessment. Princeton, NJ: Educational Testing Service.

Berliner, D. C. (in press). What's all the fuss about instructional time? In M. Ben-Peretz & R. Bromme (Eds.), Time in school. An international perspective. New York: Teachers College Press.

Berliner, D. C., & Carter, K. J. (1986, October). Differences in processing classroom information by expert and novice teachers. Paper presented at the meeting of the International Study Association on Teacher Thinking, Leuven, Belgium.

Berliner, D. C., & Rosenshine, B. (1977). The acquisition of knowledge in the classroom. In R. J. Anderson, R. J. Spiro., & W. E. Montegue (Eds.). Schooling and the acquisition of knowledge (pp. 375-396). Hillsdale, NJ: Lawrence Erlbaurn Associates, Inc.

Berliner, D. C., & Tikunoff W. J. (1976). The California beginning teacher evaluation study: Overview of the ethnographic study. Journal of Teacher Education, 27, 24-30.

Block. J. H., & Burns, R. B. (1976). Mastery learning. In L. Shulman (Ed.), Review of research in education, 4 (pp. 3-49). Itasca, IL: F. E. Peacock.

Borg, W. R., & Ascione, F. R. (1982). Classroom management in elementary mainstreaming classroom. Journal of Educational Psychology. 74, 85-95.

Boring, E. G. (1957). A history of experimental psychology. New York: Appleton-Century-Crofts, Inc.

Brophy, J. E. (1983). Classroom organization and management. Elementary School Journal, 83, 1-13.

Carnahan, R. S. (1980). The effects of teacher planning on classroom processes (Tech. Rep. No. 541). Madison, WI: University of Wisconsin. Wisconsin Research and Development Center.

Carroll, J. B. (1963). A model of school learning. Teachers College Record, 64. 723-733.

Clark. C. M.. & Peterson, P. L. (1986). Teachers' thought processes. In M. C. Wittrock (Ed.), Handbook of research on leaching (3rd ed.. pp. 255-298). New York: Macmillan.

Clark, C. M., & Yinger, R. J. (1987). Teacher planning. In D. C. Berliner & B. Rosenshine (Eds.), Talks to teachers (pp. 342-368). New York: Random House.

Cooper, H. M., & Good, T. L. (1983). Pygmalion grows up. White Plains, NY: Longman.

Cronbach, L. J. (1949). Essentials of psychological testing. New York: Harper & Row.

Denham, C., & Lieberman, A. (Eds.). (1980). Time to learn. Washington, DC: U.S. Department of Education, National Institute of Education.

Dewey, J. (1896). The reflex arc concept in psychology. Psychological Review, 3, 357-370.

Dewey, J. (19 16) Democracy and education. New York Macmillan. .

Edmonds, R. R. (1979). Effective schools for the urban poor. Educational Leadership, 37, 15-27.

Elbaz, F. (1981). The teacher's "practical knowledge": Report of a case study. Curriculum I Inquiry, 11, 43-71.

Emmer, E., Evertson, C., Sanford, I., Clements, B., & Worsham, M. (1984). Classroom management for secondary teachers. Englewood Cliffs, NJ: Prentice-Hall.

Evertson, C., Emmer, E., Clements, B., Sanford. J., & Worsham, M. (19114). Classroom management for elementary teachers. Englewood Cliffs, NJ: Prentice-Hall.

Evertson,, C. M., & Emmer, E. T. (1982). Effective management at the beginning of the school year in junior high school classes. Joumal of Educational Psychology, 74, 485-498.

Fisher, C. W., & Berliner, D. C. (Eds.), (1985). Perspectives on instructional time. New York: Longman.

Gage, N. L. (1%3). Paradigms for research on teaching. In N. L. Gage (Ed.), Handbook of research on teaching (pp. 94-141). Chicago: Rand McNally.

Halberstam, D. (1986). The reckoning. New York: William Morrow & Company.

Karweit, N. (1985). Should we lengthen the school term? Educational Researcher, 14(6), 9-15.

Kounin, J. (1970). Discipline and group management in classrooms. New York: Holt, Rinehart & Winston.

McGeoch, J. A.. & Irion, A. L. (1952). The psychology of human learning (rev. ed.). New York: Longman.

Osgood, C. E. (1953). Method and theory in experimental psychology. New York: Oxford University Press.

Peters, T. J., & Waterman, R. H., Jr. (1982). In search of excellence: Lessons from America's best-run companies. New York: Harper & Row.

Rosenthal, R., & Jacobson, L. (1969). Pygmalion in, the classroorn. New York: Holt, Rinehart & Winston.

Rowe. M. B. (1974). Wait-time and rewards as instructional variables. their influence on language, logic, and fate control: Part 1. Journal of Research in Science Teaching, 11, 81-94.

Rutter, M. Maughan, B. Mortimore, P., Ousten,, J., & Smith, A. (l979). Fifteen thousand hours. Secondary schools and their effects on children. Cambridge, MA: Harvard University Press.

Schwille. J., Porter. A. Belli. G., Floden, R. Freeman, D., Knappen, L., Kuhs, T., & Schmidt, W. (1983). Teachers as policy brokers in the content of elementary school mathematics. In L. S. Shulman & G. Sykes (Eds.), Handbook of teaching and policy (pp. 370-391). New York: Longman.Shulman, L. S. (1986). Paradigms and research programs in the study of teaching: A contemporary perspective. In M. C. Wittrock (Ed.), Handbook of research on teaching (3rd ed., pp. 3-36). New York: Macmillan.

Smith, M. L., & Glass, G. V. (1987). Research and evaluation in education and the social sciences. Englewood Cliffs, NJ: Prentice-Hall.

Tobin, K. (1987). The role of wait time in higher cognitive level learning. Review of Educational Research, 57, 69-95.

Watson, J. B. (1913). Psychology as the behaviorist views it. Psychological Review, 20, 158-177.

Yaakobi, D., & Sharan, S. (1985). Teacher beliefs and practices: The discipline carries the message. Journal of Education for Teaching, 11, 187-199.

Zahorik, J. T. (1970). The effects of planning on teaching. Elementary School Journal, 71, 143-151