Donna S. Sabers
University of Arizona

Katherine S. Cushing
Harrison School District, Colorado

David C. Berliner
Arizona State University

DONNA S. SABERS is a doctoral student at the College of Education, University of Arizona, Tucson, AZ 85721.

KATHERINE S. CUSHING is the Coordinator of Research and Evaluation at the Harrison School District Two, 1060 Harrison Rd., Colorado Springs, CO 809063586. She specializes in program evaluation and research on teaching.

DAVID C. BERLINER is a professor of curriculum and instruction at Arizona State University He specializes in the psychology of teaching and learning.

Expert, beginning, and novice teachers viewed three television monitors, each focusing on a work group of a junior high science class, simultaneously. Participants expressed their thoughts as they viewed the monitors, indicated the monitor to which they were referring, among the groups were found in their perceptions, monitoring, and understanding of classroom events characterized by simultaneity, multidimensionality, and immediacy. This stud illustrates bow more than content knowledge is required for successful teaching, and that learning to teach requires a great deal of time. Findings from this study have implications for the development of preservice and inservice training programs, which may require redesign to facilitate the development of pedagogical expertise.

Studies of classroom processes indicate that teachers must continually monitor a large number of events, many of which occur at the same time. The term used to describe this characteristic of classrooms is simultaneity. As Doyle (1986) notes:

While helping an individual student during seatwork, a teacher must monitor the rest of the class, acknowledge other requests for assistance, and keep track of time. During a discussion, a teacher must listen to student answers, watch other students for signs of comprehension or confusion, formulate the next question, and scan the class for possible misbehavior. At the same time, the teacher must attend to the pace of the discussion, the sequence of selecting students to answer, the relevance and quality of answers, and the logical development of the content. When the class is divided into small groups, the number of simultaneous events increases, and the teacher must monitor and regulate several activities at once. (P. 394)

Doyle (1986) also writes of two other properties of classrooms: multidimensionality, which refers to "the large quantity of events and tasks in classrooms" (p. 394); and immediacy, which refers to the rapid pace of classroom events.

Researchers have estimated that elementary teachers have many hundreds of distinct interactions with individual students in a single day (Gump, 1967; Jackson, 1968) and, further, that classroom order depends to a large degree on a teacher's ability to maintain momentum and a consistent flow of classroom events (Kounin, 1970). In Kounin's (1970) study, teachers whose students had high levels of work involvement were high on "withitness" and "overlapping." In other words, they were aware of what was happening in the classroom, communicated this awareness to students, and were able to attend to two or more events simultaneously. Thus, Kounin concluded that teacher attention and monitoring are central components of the skills needed for classroom management.

Clark and Peterson (1986) have studied these issues as well, concluding that teachers apparently think constantly about what is happening in their classrooms and make many nontrivial decisions about what to do at rates of about one decision every two minutes. It appears, however, that teachers assess only certain classroom behaviors and events ' namely, those needing immediate teacher attention. Other perceived behaviors and events are rapidly assessed as being less critical, resulting in a decision by the teacher to delay action or to take no action at all.

Erickson (1984) has also studied simultaneity, focusing on how teachers observe and make sense out of what happens in their classrooms. He and his colleagues found that experienced teachers seem to know what to look at and what to ignore. In Erickson's study, experienced teachers rarely noticed things in isolation, demonstrated the capacity to note very specific details, and attended most to situations and events that called for some decision and action. In contrast, teacher education students focused on the most unruly students in the room. Consequently, these prospective teachers appeared to be more concerned with management than with instructional strategies or students learning. Fine details of subject matter instruction took second place. To them, keeping order was the primary concern.

Copeland (1987) developed a computer-administered test battery to measure student teachers' multiple-attention and vigilance skills as they relate to successful classroom management. He concluded that teachers, when monitoring classroom behavior, focus on different events, called cues, for decision making. The successful teacher processes and interprets multiple cues simultaneously. Once the cues are internalized, the teacher makes appropriate decisions to ignore cues that are determined not to require action and to respond to those cues that are determined to require action.

The research in this area, then, points out the complexity of classroom processes, the skillfulness needed by teachers to deal with such complexity, and the ways teachers, particularly teachers with different levels of experience, focus on different classroom behaviors and events. Following Erickson's work, we sought to assess differences between experienced and inexperienced secondary science teachers in their perception, monitoring, and understanding of classroom events characterized by simultaneity, multidimensionality, and immediacy. This study is part of a larger research project to identify differences in pedagogical knowledge among experienced (expert) and less experienced teachers (novices and advanced beginners).

The decision to use expert and beginning teachers was designed to maximize variation in experience. The inclusion of a group of novices--content matter specialists with no formal classroom experience-was for two reasons. First, when novices are compared with beginners we are helped in understanding the effects of teacher education and student teaching. Furthermore, the novices resembled those who enter teaching by alternative certification routes and who possess only content knowledge. When compared with the two other groups of participants, their pedagogical knowledge could be assessed.

Data from the studies in this series are congruent with a general model of the development of expertise across many fields proposed by Dreyfus and Dreyfus (1986; see also Benner, 1984). Five stages of development have been described. In the novice stage the common places of a situation are learned; context free rules of behavior are followed; behavior is relatively inflexible; experience must be gained. The advanced beginner melds book knowledge with on-the-job experience, building up case or episodic knowledge to go along with semantic knowledge. The beginner, however, is still not very flexible in behavior. The competent performer of a skill is rational, flexible, can articulate goals, and can employ sensible means to achieve them. The proficient performer has a heightened sense of pattern recognition, has a holistic understanding of the processes involved, and recognizes similarities between apparently quite different situations. The experts not only have a holistic, integrated view of the situation, but they also respond effortlessly, fluidly, and appropriately to the demands of the situation with which they are confronted. The heuristic value of this stage theory for thinking about the development of competence and expertise when learning to teach, as well as the policy implications of such a model, have been described by Berliner (1989).

Participants

For this study, a subset of participants was drawn from a larger pool of expert, advanced beginner, and novice teachers initially identified for the research project. Descriptions of the goals of the research project may be found in Berliner (1986). Experts consisted of junior and senior high school science teachers who were first identified by nominations of school superintendents and principals. Fifty-five nominated teachers were observed by project personnel, all of whom were knowledgeable about research on teaching and who had either classroom teaching experience or who were trained in classroom observation techniques. The performance of each teacher was discussed by the research team, and those who received the highest ratings from the first observer were visited by a second observer. Once again discussion ensued, and seventeen teachers were identified as experts. A third independent observation was made of these teachers, resulting in seven of those teachers being designated as the most expert in our pool of participants. Experts for a particular study in this series were first recruited from the pool of seven, and, if needed, additional participants were recruited from among the other ten teachers in the pool. These teachers all had over five years of experience and taught a wide range of courses and grade levels within the secondary science curriculum.

Advanced beginners making up the pool of participants for this series of studies were selected from among secondary science student teachers or first year classroom teachers who volunteered to participate in the studies. The student teaching records of all these participants were reviewed to ensure that each of the advanced beginners selected for the study was competent, capable, and possessed the potential to develop into an "excellent" teacher.

Novices were individuals who expressed an interest in teaching but had no training or experience in public school teaching. Novices are viewed as those who might request alternative certification for teaching. These individuals were employed in business or industry and had experience and expertise in fields such as chemistry, engineering, or computer technology. While many of these individuals had experience working with youth or church groups, and many expressed a desire to teach, they had received no formal pedagogical training. Novices were first contacted about their willingness to participate in the research through their personnel office. After interviews in which the nature of the study was explained and their seriousness about teaching was explored, a pool of about twenty individuals was identified as a source for potential participants in this series of studies.

For this study the sample consisted of seven experts, four advanced beginners, and five novices selected from each of their respective pools. All of the expert and advanced beginners taught secondary science while the novices worked in fields in which scientific knowledge was needed. An in-depth study of the qualitative differences in the perception and monitoring of complex classroom events by individuals in these three groups requires the use of small samples. Nevertheless, the number of participants involved in this study is comparable to the number involved in other studies of experts and novices (cf. Mazzeo, 1985; Chi, Glaser, & Farr, 1988).

One class period of a junior high science classroom was videotaped for an entire week. The teacher and students had been requested to act as they typically would and to ignore the video equipment in the room. No attempts were made by the experimenters to change classroom conditions. The intent was to represent classroom activities as they would occur naturally. Success in this endeavor was evidenced by the teacher's comment that the students' behavior was "pretty typical" of what she would expect.

A videotape from one of the classroom periods was selected and edited into three tapes, each showing a different view of the classroom. The selection was based on the following criteria: interaction among the teacher and students; presentation of scientific content; and involvement of the students in the lesson. Editing was done so that each monitor would show the students in a proximately one-third of the classroom throughout the viewing period. The nature of the action on any monitor represented what was happening in that third of the classroom at a given time. Thus, the viewer needed to be simultaneously aware of all screens in order to experience the activity of the classroom at that moment. Although using three video cameras, each focusing on a different part of the room, might be considered more desirable than editing one tape into three, the former was not feasible because of limited classroom space for setting up equipment and the obtrusive nature of three cameras.

The goal of the editing was to create a videotape that displayed simultaneity, multidimensionality, and immediacy, in a natural way. There were two criteria by which this editing was judged to be successful. First, the comments of the experienced teachers indicated that the simultaneous monitors in this experiment present a realistic situation of a classroom task. Second, the amount of data collected from viewers' comments indicated that the televised situation provided sufficient stimuli.

Each tape focused on different student groups and thus provided different stimuli for the participants. The lengths of the three tapes vary according to how well the edited portions could be fit together without the participants noticing any editing effects. The video for the middle monitor lasted approximately 25 minutes and was used to begin and end he viewing. The first three minutes of the video showed the classroom environment without students. Participants saw shelves of books and reference materials, tables of science equipment, a small greenhouse, gerbils in cages, an aquarium, and instructional equipment including a computer and monitor and an overhead projector. The video then showed students entering the classroom and finally focused on two groups and a single student. The girl, seated by herself, had limited interaction with the teacher and with the other students. A group of four boys was somewhat loud, frequently interacted with the teacher and other students, and generally seemed to be enjoying life. The boys in another group were quieter and more reserved, but generally followed the teacher's directions and classroom instruction. They also interacted with the teacher. Because the teacher tended to remain in the middle of the room, this monitor also focused on her instructional strategies and interactions with the students. The last five minutes of this video showed students measuring their corn plants and preparing to leave when the bell rang.

Approximately three minutes after the middle video had begun, the left monitor started. This tape lasted approximately 12 minutes, initially showing students entering the classroom and then focusing on two groups of students. One group of four girls worked the entire period. Their books were open on their desks, and they were cutting out materials associated with a previous experiment, filling out worksheets, and measuring their corn plants. However, these four girls infrequently interacted with the teacher; they appeared to be self-directed and "on task" but not necessarily on the same task as the teacher. The second group of students seen on the left monitor appeared interested, cooperative, and attentive to the teacher's directions. These students interacted with one another and the instructor.

Five minutes after the first tape started, the third monitor began. This monitor, on the participant's right, lasted approximately 14 minutes and focused on two groups of students. Two girls appeared to be working intently on something, but it was difficult to tell whether it was the science lesson. They frequently sprawled over their science tables, watched the clock, and interacted with students in other parts of the room. Another group of four girls appeared to be on task. They raised their hands to ask and answer questions and appeared to be following the classroom instruction and completing the worksheet.

For each experimental session, a single participant from one of the three groups (expert, advanced beginner, and novice) was shown all three video tapes at the same time and asked to monitor all three screens simultaneously. First the tape on the middle monitor began. About three minutes later the tape began playing on the left monitor, and two minutes later the tape on the right monitor began. For approximately ten minutes all three screens were presenting information about that classroom simultaneously. Then the picture on the left monitor disappeared, then on the right, and finally on the middle.

The Experimental Task

The experimental task that the participants engaged in consisted of four segments. During the first segment, the participants viewed the three monitors with the sound coming from just the middle monitor. After the viewing, two requests were made of the participants:

1) Describe the instructional techniques or strategies you observed being used in this classroom.

2) Describe the management techniques or strategies you observed being used in this classroom.

For the second segment, the participants again viewed the same three monitors with the sound coming from all three monitors. During the second viewing, the participants were asked to "talk aloud," telling what they were thinking and seeing, and commenting about what concerned them and pleased them. Beyond this set of instructions, no training or practice in "talk aloud" procedures was provided. During this segment the participants also indicated what monitor they were talking about by pressing a number on a keyboard ("I" for the left monitor; " 5 " for the middle monitor; and "0" for the right monitor). A computer recorded the number and time between participant comments and provided a record of the pattern of monitoring as well.

During the third segment of the task, the participants responded to nine questions concerning routines, content, motivation, learning environment, students' attitudes, teacher's expectations, teacher's roles, critical thinking skills, and the relationship that existed between the teacher and the students.

For the fourth segment, a measure of memory for specific details was administered. The participants responded "yes" or "no" to questions about what they had seen on each monitor. For example, the participants were asked: "On the left monitor did you see a girl asking a question?"

The data were analyzed using several methods. Descriptive and inferential statistical analyses were used when appropriate (e.g., the percentage of each group's comments for each category presented in proposition one) and are discussed at length in the "Results and Discussion" section.

Given the unique nature of the task presented in this experiment, qualitative data analysis procedures were also employed. Participants' "talk alouds" and responses to questions were audiotaped, and protocols were transcribed for analysis. All responses were analyzed through a multi-step, iterative process designed to determine patterns, trends, and differences in both kind and quality of responses. Two researchers coded all the data independently, each using two coding systems.

One coding system focused on the content of comments according to the following six categories: student behaviors, teacher behaviors, instructional techniques, management techniques, time use, and classroom environment. The second coding system focused on the nature of the comments and included five categories: description, interpretation, evaluation, conclusion, and suggestion. If a comment merely mentioned what was happening, it was coded as a description: "The kids walk in." A comment using words such as "seemed" or "appeared" was coded as an interpretation: "The students seem relaxed enough." If a comment indicated a judgment was being made, it was coded as an evaluation: "It's fairly efficient at this point because they're used to the form they are using." If the comment indicated a conclusion had been drawn, it was coded as a conclusion: "This is about the time I decided they must be an accelerated group." If a participant suggested a technique or strategy, it was coded as a suggestion: "It would have been a good idea to start out the class with measuring the height of these plants." All coded data were also analyzed according to what was showing on the three monitors when the comments were made.

After initial coding, tentative hypotheses were formed and researchers met to share coding schemas and hypothesized findings. Coding procedures were refined and, where appropriate, categories were combined or split to more accurately reflect the data. Propositions 1, 2, 4, 5, and 6 were identified independently by both coders who had selected many of the same instantiations. A further reading of the protocols was undertaken to find the best examples to support or refute the propositions. The other three propositions were based on the analysis of scanning patterns and memory capacity.

The decision to include representative instantiations; for findings was made because of the richness of the detail in participant comments and their usefulness when searching for differences between the groups of subjects. Obviously there is subjectivity involved in the decision regarding which instantiations to include and, as Peshkin (1988) has pointed out, it does little good to acknowledge subjectivity without attending to it in a meaningful way. The propositions we present, below, were documented by the independent findings of two researchers, but Ziman (1978) would label that agreement as intersubjectivity rather than objectivity. Nevertheless, numbers and tables cannot do justice to the richness of the data that were collected, and agreement in assigning numbers is no less subjective than the agreement of researchers about propositions that faithfully describe the data from this experiment.

Results and Discussion

The findings from this study have been summarized in the form of three major conclusions, each supported by two or three propositions. Each proposition is followed by documentation in the form of data summaries or excerpts from individual protocols which instantiate the findings. Because the intent of the study is to learn more about the differences in classroom performance among teachers at different levels of experience and expertise, each of the propositions is phrased as if it referred to performance in an actual classroom setting. It should be remembered, however, that the data were collected from a laboratory study where videotapes of classroom events were the stimulus materials. This laboratory simulation involving a small nonrandom sample severely limits the generalizability of the propositions presented to other populations and settings.

Conclusion A: Experts are able to monitor, understand, and interpret events in more detail and with more insight than either novices or advanced beginners. The two propositions that support this conclusion were based on data analyzed from the first and second segments of the experimental task.

Proposition 1. Experts, advanced beginners, and novices differ with respect to their abilities to monitor and interpret simultaneous, multiple events within a classroom. During the second segment of the experimental task, in which participants were asked to "talk-aloud," expert teachers were not only the most adept at monitoring all three screens simultaneously, but they were able to make sense out of events that puzzled advanced beginners and baffled novices. As reported earlier, all comments made by each participant during this segment of the task were coded into five categories: descriptions, interpretations, evaluations, conclusions, and suggestions. The number of comments in each category was changed to a percentage for each individual because for these participants, just as those studied by Peterson and Comeaux (1987), the experts gave more elaborated answers. Table I presents the percentage of each group's comments for each category. These data were analyzed with a 3 (groups) x 5 (types of comment) analysis of variance. The F test for the variable "types of comment" was significant (F = 32.92, df = 4, p < .001), as was the interaction term (F = 7.09, df = 8, p < . 00 1). With percentages for each participant summing to unity across types of comment, it could not be expected that differences between groups would be significant. The significant interaction term, indicating a group's x type of comment interaction, is the relevant term to examine to inquire whether significant group differences were present. An arc sine transformation of these data was also completed. This procedure is recommended when percentages are the dependent variable and there is some question about homogeneity of variance. An analysis of variance with the transformed data revealed almost the same pattern as presented above.

Experts, more than the others, attempted to interpret the events and behaviors they described as well as to draw conclusions about their meaning and to make evaluative judgments about them. Some illustrations of these kinds of evaluative comments are the following:

Expert 6: On the left monitor, the students' note taking indicates that they have seen sheets like this and have had presentations like this before; it's fairly efficient at this point because they're used to the format they are using.

Expert 7: I don't understand why the students can't be finding out this information on their own rather than listening to someone tell them because if you watch the faces of most of them, they start out for about the first two or three minutes sort of paying attention to what's going on and then just drift off.

Examples of interpretative comments are the following:

Expert 2:Left monitor again ... I haven't heard a bell, but the students are already at their desks and seem to be doing purposeful activity, and this is about the time that I decided they must be an accelerated group because they came into the room and started something rather than just sitting down or socializing.

Expert 4:Again, viewing the middle monitor, I think there is an indication here of the type of structure of this classroom. It's pretty loose. The kids come in and go out without checking with the teacher.

In contrast, advanced beginners and novices gave a step-by-step account of what was happening as though they were announcing what they were viewing to someone who could not see the screen. in this respect they were reminiscent of radio announcers reporting an athletic event. Some examples of these descriptive comments are the following:

Advanced Beginner 1: The kids walk in. She doesn't say hello to any of them. They're sort of wandering in.

Advanced Beginner 2: Right monitor; it looks like the girl just finished her note and has put it in her book.

Advanced Beginner 3: In the right monitor, we have the teacher lecturing, students taking notes.

Novice 1: On the left, they umm ... I can't tell what they are doing. They're getting ready for class, but I can't tell what they are doing.
Novice 2: They seem to be more studious on the left here, if they are doing... I can't tell; I'm not close enough to tell if they are doing the actual work, but they are not participating in the lecture.
Novice 3: She's trying to communicate with them there about something, but I sure couldn't tell what it was.
Novice 4: I think it looks like the class is breaking up. Nope, not breaking up. They went to get lab dishes. So they've got something in paper cups.

Proposition 2. Experts, advanced beginners, and novices differ in their interpretations of what instructional strategies or techniques were used by the teacher. When queried in the first segment of the experimental task about the instructional techniques or strategies employed by the teacher in the science lesson that they observed, all seven of the experts and three of the four advanced beginners agreed that the teacher's primary instructional technique for this lesson was lecture. In addition to their characterization of the lesson as primarily a "lecture," four of the seven experts elaborated on their responses by explaining what strategies the teacher was using. Furthermore, expert protocols often appeared to be qualitatively more analytical when discussing the teacher's instructional practices, even though the experts did not always agree about what those practices were. In contrast, the attempts made by the advanced beginners to explain or elaborate on their answers conveyed little evidence of analytical thinking. Protocols of advanced beginners seemed simply to be statements of what they perceived the instructional practices to be.

Representative excerpts of protocols from the two groups follow. First are those of the experts:

Expert 2: The strategy was activity oriented. The teacher's method was activity oriented. The teacher's method, associated with this activity was, I guess you would call it Socratic, with the question and answer sort of approach.

Expert 4: There was some formal lecture, and there was a formal activity. I think the technique that she used was very low key, perhaps a process type approach to teaching science rather than a very structured approach.

Advanced Beginner 1:It looks ... well, mostly lecture. She had some activities for the kids to do. Some use of media. She used the overhead a little bit.

Advanced Beginner 2: There were some kids working on their own. The teacher used the overhead projector, and the students had a handout which matched the overhead projector. The teacher lectured a little bit and talked to the students. They had an activity where they were measuring corn or something.

Advanced Beginner 4: She did lecturing near the beginning with the overhead projector. She was using some audiovisual things.

Novice protocols were more similar to the protocols of the advanced beginners in the respect that three of the five seemed to focus on the materials that the teacher used for instruction. For example,

Novice 1: Well, she gave them a diagram, and then she explained it with an overhead projector. It was also a lesson in taking notes.
Novice 3: She began the class by handing back some homework and trying to get some people to put a better response to the homework. She was trying to tie some of the material, some of the technical terms like epiglottis and things like that, down to just chewing your food and not choking on it.
Novice 4: There was no demonstration. They weren't given any equipment. It all seemed kind of disjointed, no evidence of any continuity.

Conclusion B: Experts, advanced beginners, and novices differ in how they attend to the multidimensional nature of the classroom. The three propositions that support this conclusion were based on data analyzed from the first three segments of the experimental task.

Proposition 3. Experts, advanced beginners, and novices differ in their scanning patterns as they monitor multiple events in a classroom. During the second segment of the experimental task, participants were seated in front of a computer and were instructed to strike different keys to indicate the television monitor upon which they were focusing their attention. To accomplish this task, participants needed to scan three monitors, looking and listening for events that stimulated some type of thought or action, to strike the appropriate computer key, and then to talk about what they were thinking. All of this was done with the knowledge that they would also be asked general and specific questions about what they saw and heard.

The participants, therefore, were required to do a very complex task, consisting of several parts, just as teachers are required to do when they monitor multiple events in their classrooms.

Based on what expert teachers are observed doing in their own classrooms as they continually monitor events characterized by simultaneity, multidimensionality, and immediacy, the results are not surprising. Expert teachers scanned all three screens while subjects from the other two groups tended to focus most of their attention on the middle monitor. The computer-generated records of scanning behavior during the ten-minute period of simultaneous viewing revealed differences in the distribution of key strokes for the three groups. The data were analyzed by the percentage of each viewer's key strokes indicating a monitor position; for example, Expert 2 indicated the left, center, and right screens with 27%, 39%, and 33 % of the strokes. These data are presented in Table 2. A 3 (groups) x 3 (monitor position) analysis of variance was computed using these data.

Experts not only used all three monitors to view the classroom activities, but by all outward appearances they seemed more at ease in completing the task. They gave the overall impression of enjoying the experiment and participated enthusiastically. To them, the task of scanning a classroom is natural because it is a task that is familiar to them and one they engage in daily. On the other hand, the task of scanning a classroom is not familiar to novices. Perhaps at no time during their work day are they required to use a comparable skill. Therefore, as might be expected, novices were observed to be much less proficient in performing this task than were the experts. Even thought they were informed that there were three screens to watch, novices paid considerably less attention to the two side screens than they paid to the middle monitor. Advanced beginners, like the novices, had difficulty in scanning all three screens. Most of their attention was focused on the middle monitor, too, with less attention given to the side monitors. One plausible explanation of this occurrence is that both groups focused their attention on the teacher, who appeared more frequently in the middle monitor than in the other two. Perhaps, to them, what the teacher does is perceived as critical for understanding classroom events. The teacher is viewed as the main actor with the students being the supporting cast. In contrast, experts realize that all members of the cast are important if a production (lesson) is to be well received. Thus, they viewed all three screens. Most advanced beginners probably have had more experience in using classroom scanning skills than novices, but such experience was not evident in their performance on this task.

Proposition 4. Experts, advanced beginners, and novices differ in their ability to monitor classroom sounds and to use language to assist them in understanding, interpreting, and evaluating classroom events. Experts attended more to sound and to the instructional language used in the classroom than did members of the other two groups. This monitoring of language by experts provided them the opportunity to support, amplify, or question their visual perceptions of classroom events. The ability to monitor the language used in instruction was not as apparent in the protocols of the novices or advanced beginners. They tended to use visual cues to describe what they observed, and their comments about sounds, or language, tended also to be descriptive in nature. Novices and advanced beginners were, therefore, less likely to make meaningful associations between auditory and visual clues in their attempts to understand and explain what they observed on the classroom videos.

Experts' monitoring of language included references to specific comments about instructional content and classroom management. Based on what they saw and heard, experts were able to explain the management techniques (or lack thereof) as well as the content of the lesson in some detail. The following examples from the talk-aloud protocols support this contention:

Expert 3: In the middle monitor they're discussing notes that were taken on the reproductive system. On the right hand monitor she's asking specific individuals what the different parts of the digestive system are ... she was very broad in just talking about the digestive system, rather than being specific and talking about parts or function and it seems like those are two separate objectives.

Expert 4: 1 don't see how this ties in with genotype or the other things that she was mentioning earlier in the class. The teacher seems to be basically just doing her thing, talking to kids individually.

Expert 5: 1 generally feel from the teacher's comments in the center monitor that she's not really interested in the subject, especially when she said that this is the best that she could do and [she makes] references to the materials as junk and stuff.

Expert 6.- In the middle monitor, she's talking about how the energy is released in the cell. Unless I missed it, she seemed to skip the lower digestive system, the two intestines, and how sugar and glucose get from the digestive system to the cells.

Expert 7: She's talking about using the upper and lower case letters to represent genetic crosses, and yet the way that the information is conveyed, is that she's being critical of what they've written down before she gives them some correction as to how it ought to be.

In contrast, the protocols of novices and advanced beginners made few references to the language of instruction, indicating, perhaps, that they were unable to understand the details of the lesson content. In general, they commented on the teacher talking as a means of describing what it was they were observing. Their comments lacked interpretive value in terms of instructional content or classroom management. For example,

Advanced Beginner 1: She's telling them things she wants them to do. Advanced Beginner 3: They fill in the diagram as it's being explained by the teacher simultaneously... there seems to be more of a discussion going on between the teacher and students.

Advanced Beginner 4: The teacher is talking a lot.

Advanced Beginner 5: In the center she seems to be good at asking questions and getting the students to talk to her.

Novice 1: She's explaining. . . she seems to be very serious about her explanations but nobody seems to be really listening ... she just asked a question.

Novice 3: She keeps making a lot of references to the computer that they're using. . . she's trying to communicate with them there about something, but I sure couldn't tell what it was.

Novice 5: In the right monitor it seems that the teacher is going over something; there doesn't seem to be much interest from the students ... the teacher is still trying to get instructions across and it's hard to believe that everyone is comprehending what she's saying, or even hearing what she's saying.

One interpretation of these data is that expertise requires integration of information about instruction. Experts, with many more years of experience, use both the language of instruction and the observation of behaviors in the classroom, as well as extensive pedagogical content knowledge to provide meaningful interpretations of classroom life. Advanced beginners and novices may approach the experts' level in terms of their content expertise, but, perhaps due to their limited experiences in teaching, they have not yet developed the ability to interpret the meaning and use of language in classroom settings nor to integrate that information into their overall conceptions of classroom life.

Proposition 5. Experts, advanced beginners, and novices differ in terms of where they place their primary focus when monitoring multiple events in a classroom. Subjects across all three groups appeared to be concerned about student interest, participation, and utilization of time. However, without exception, novices commented most and focused more on student behavior, especially behavior that they felt to be inappropriate. Issues of managing student behavior and maintaining classroom control, naturally enough, caught their attention and consequently became the greatest source of their interest and concern. Novices did not seem to try to determine what might be causing the students to act the way in which they did. Nor did novices seem to offer possible solution strategies for the problems they noted. They just seemed to want to express their disapproval of students who wasted time, who paid little attention to what the teacher said or did, and who displayed a general disinterest in what was happening. For example,

Novice 1: On the left, I don't think they're watching the, oh, they've got the diagram. It doesn't seem like they are watching the overhead though.
Novice 4: There seemed to be a number of people who weren't really doing much, taking notes or anything.
Novice 5: On all three monitors, it seems to be about the time when she has lost control over the group.

Novice 2: The kids are not disciplined enough in that classroom to know that when the bell rings, they had better be organized, better be sitting down.

Novice 3: In the center there, the one girl is almost laying down on the table. I could've sworn someone would have told me not to do that when I was in this sort of class.

Expert 1: She [the teacher] seems like she's finished now, and she's not really going around to help students and to monitor how they're doing. There's time left in the class, and she's just kind of finished her work. She's not going around to check on how students are doing.
Similarly, another expert suggests how the teacher might make herself more available to students;
Expert 6.. On the left hand monitor, the students seem to be working on an independent desk work type project. Perhaps utilizing the aide at this point to pass out papers that need to be passed out would enable the teacher to go around the classroom and answer questions and provide assistance.

Expert 5: Looking at the center monitor, she's talking towards the center of the room, and they're the ones that are paying attention.
Another expert offered this explanation for the inattentiveness of the students:
Expert 4: 1 doubt if that student is listening particularly well because she seems to be filling in blanks that are on the worksheet which appears to be part of this class. You're given worksheets; you're given things to fill out. You're given perhaps short little laboratories to perform. There is not really the necessity to listen to what the teacher is saying.

Expert 3: in the monitor in the middle again, if she could get people more involved instead of just sitting and listening, it would probably help that group quite a bit.
Expert T I don't understand why the students can't be finding out this information on their own rather than listening to someone tell it to them. Because if you watch the faces of most of them, they start out for about the first two or three minutes sort of paying attention to what's going on, and then they just drift off.

Most advanced beginners in this study tried to attach meaning to why students behaved the way in which they did. In this respect, they were similar to experts. They, too, attempted to offer explanations and to make suggestions. The comments of the advanced beginners, however, did not always appear to be as insightful as those of the experts, but they certainly did exhibit understanding of classroom life. Three representative examples from the narratives follow:

Advanced Beginner 1: There were some kids who didn't look like they were paying too much attention, and she probably could have called on them more.
Advanced Beginner 2: A lot of the students seem to be asking if they're going up to the computer center. It might be a good idea to write on the board that they'll be working in the lab today, or maybe to have told them the day before that they would be working in the lab today.
Advanced Beginner 4: The kids seem to get distracted easily. Probably the subject matter is too easy for them. They don't have to do much to learn it. Learn it in two minutes and the other forty-eight minutes is just for having fun and looking for things to do.
Conclusion C Experts, advanced beginners, and novices performed similarly on tasks requiring judgments of content selection and on tasks requiring memory for nonmeaningful details. The three propositions that support this conclusion were based on data analyzed from the third and fourth segments of the experimental task.

When participants were asked to identify any evidence they saw of the scientific accuracy or inaccuracy of the content, almost all of them qualified their responses to this question, emphasizing that they were making this determination based on their knowledge or personal experience, not on any advanced study or training.

Perhaps this finding occurred because of the decontextualized nature of this task. Since the participants were not actually teaching, and did not know either the students or the history of the class, they were hesitant to make judgments about the appropriateness and accuracy of what they were observing. It is interesting to note that at no other point in this experimental task was additional information requested by members of any group. Perhaps this speaks to the limitations of the use of simulations for this type of research as well as to the characteristics of these participants.

Proposition 7. Experts, advanced beginners, and novices showed little difference in memory capacity for questions requiring specific memory for details. In the fourth segment of the experimental task, participants were asked to respond "yes" or "no" to questions about incidents they may or may not have viewed on each monitor. Examples of some of these questions include the following: On the left monitor did you see (1) four girls at a table, (2) a girl turned sideways, (3) a girl reading a magazine; on the middle monitor did you see (1) a boy with a paper airplane, (2) a boy pretending to play a piano, (3) a boy listening to the alarm on his watch; on the right monitor did you see (1) a girl folding a note, (2) a girl watering plants at the sink, (3) a girl sitting on top of a desk. Few differences were noted in the average correct scores for experts, advanced beginners, and novices. These data are presented in Table 3. A 3 (groups) x 3 (monitor position) analysis of variance using these data revealed no significant differences among groups. The F test for monitor position was significant (F = 17.29, df = 2, p < .001), though not of great interest. It may only indicate that the questions asked about events on the right monitor were easier to answer than the questions asked about events on the other two monitors.

This finding is not altogether unexpected. The events about which participants were asked to remember were isolated events not meaningfully tied to instruction. Also, participants had a 50% probability of answering each question correctly by chance. In earlier studies of expert and novice teachers (Carteret al., 1987; 1988) as well as in a finding reported previously in this article, the responses of experts have been interpretive in nature, whereas novices and advanced beginners gave descriptive responses. The set of questions used in this study did not provide the opportunity to interpret what was observed; rather, there was only an opportunity to indicate whether an observation had been made. In terms of ability to observe particular events in classrooms, it appears that experts, advanced beginners, and novices are similarly proficient.

Studies of experts and novices in fields other than teaching have produced similar findings. De Groot (1966), for example, found that when chess pieces were placed randomly on the board, master players did no better than weaker players in reconstructing the chess boards from memory. DeGroot concluded that chess masters' superior performance with meaningful positions lies in their ability to interpret and encode that meaningfulness.

In terms of observation skills it appears that experts, advanced beginners, and novices may be quite similar. In terms of making sense of what they have observed experts, advanced beginners, and novices differ significantly.

Proposition 8. Experts, advanced beginners, and novices showed little difference in memory for salient or unusual events that they observed. At times events were portrayed that were so specific or unique to the classroom that members of each group, and in some instances each participant in this study, remembered and described the same instance. For example, the wearing of name tags by students was mentioned by all participants though members from each group experienced difficulty in understanding the purpose of the tags. All but two participants (one expert and one novice) mentioned the use of handouts or worksheets during instruction, and all but one member of each group mentioned the use of the overhead projector, or viewgraph as the novices referred to it, for classroom instruction. The finding that some of the events were so salient that many participants in each group observed and commented on them is not unexpected given the similar observational skills of the participants. It is now apparent that only when participants are asked to make meaning from their observations will differences between groups be expected.

The finding that expert and novice teachers differ in the ways they interpret classroom life is similar to the recent findings of other researchers who have investigated differences between expert and novice teachers (Borko & Livingston, 1989; Livingston & Borko, 1990; Peterson & Comeaux, 1987; Nelson, 1988). This finding is reflected in the way experts focus their attention when observing classroom events. Experts scanned all three monitors and were better able to integrate both visual and auditory stimuli than were novices or advanced beginners. Therefore experts, in general, were able to: a) monitor and comprehend the events presented, b) interpret the instructional strategies used, c) hypothesize reasons for behavior seen, and d) offer solution strategies for problems identified. Interestingly, and similar to research on novices and experts in other domains, the differences are not reflected in memory capacity for nonmeaningful events, only for meaningful events.

The protocols of the experts are rich with interpretations of classroom life. Given that the educational background of the experts is not that different from the advanced beginners and their amount of education is not different from the novices, it seems safe to assume that experience is the critical factor in development of competency in this kind of task. The chess experts studied by Chase and Simon (1973) have spent 10,000 to 20,000 hours staring at chess positions. The expert radiologists studied by Lesgold et al. (1988) have examined between 10,000 and 200,000 X-rays. And the expert teachers we studied have spent between 5,000 and 20,000 hours instructing in classrooms. Experience, at least reflected on experience, is a great teacher. We concur with Badre's (1982) conclusion that a " . . . skilled problem solver is able to process larger amounts of problem data than does the novice even though there does not seem to be a difference between the two on memory capacity. The difference in recall is related to the amounts of prior experience with the given problem domain" (p. 497). In the experimental task employed in this study, the problem domain is the interpretation of an active classroom where several events are occurring at one time. Neither the working world of the novice nor the preservice training of the advanced beginner provides adequate experience with this problem domain. Thus, successful performance in this task (ease, confidence, accuracy of perception, complexity of response, etc.) is not likely to occur for novices and advanced beginners for some time after they start teaching. This conclusion is similar to that of Borko and Livingston (1989) and of Livingston and Borko (1990), who point out that many of the differences in the thinking and actions of novice and expert teachers can be accounted for by assuming that novices' cognitive schemata are less elaborate, less interconnected, and less accessible than that of the experts, and that the novices' pedagogical reasoning skills are less well-developed.

The data from our study and from studies with findings similar to ours, have implications for training and our expectations for and support of beginning teachers. For example, special induction programs might be developed for those, like the novices in this study, who seek entrance to the field of teaching by alternative certification rather than through course work and practice. By many standards they appeared ignorant of important aspects of classroom fife. They may not merely be poor teachers at first; they could do harm to students and become discouraged about teaching. Our data lead us to concur with Livingston and Borko (1990), who state that the findings from their study of experts and novices challenge the assumption that the teaching of mathematics, or any other discipline, requires only a strong content background.

Because the performance of the advanced beginners was not equal to the experts in a number of domains, perhaps policymakers need to rethink the content and structure of typical teacher education programs. Perhaps we need to structure experiences for preservice and practicing teachers that will facilitate the development of expertise. This could be done, we believe, by helping preservice and beginning teachers learn to constantly monitor their environment and, more importantly, to develop the ability to interpret what is happening within that environment. Although experience is critical to the development of comprehension and interpretive skills, opportunities to observe classrooms and reflect upon and discuss those observations with "experts" might contribute to the development of those skills.

The reliance upon an experienced and competent other to mediate a complex environment is the basis of learning in Vygotskian thought, the foundation for Reuven Feurstein's Intellectual Enrichment Program, and the essence of apprenticeship programs (see Brown, Collins, & Duguid, Nevertheless, there is probably a limit to what can be education. It takes considerable time to acquire competence in the pedagogical domain, let alone expertise. Teacher education can, at best, start people on the path toward expertise and provide them with the tools and dispositions to better learn from their experience.

in some tasks where training for classroom teaching and not lengthy classroom experience seemed relevant, the advanced beginners performed more similarly to the experts than to the novices. This was evident in the discussion of proposition 5, where advanced beginners gave explanations for student behavior that were more like those of the experts. However, even in this task the extensive classroom experience of the experts provided them with greater insight into student behavior. As Glaser (1986) stated, "Expert representations of problems and situations are qualitatively different than novice representations. In the course of developing expertise, problem representation changes from surface representations to inferred problem descriptions, to principled (and proceduralized) categorizations." (p 926) The advanced beginners in this task, in some ways, appear to be in the process of developing the kinds of skill and cognition shown by the experts, as noted in Glaser's comment about the developmental nature of expertise and as described by Dreyfus and Dreyfus (1986). The fact that advanced beginners in some ways resembled (though they did not match) the experts is encouraging. It is likely that this is an effect of a teacher education program. One should not expect much more from beginners because, as Glaser (1986) has pointed out in studies in many diverse areas, "Expertise is developed over hundreds and thousands of hours of learning and experience, and continues to develop" (p 926).

Kolodner (1983) has also focused on the role of experience in the development of expertise. She writes about the evolution of memory as one develops from novice to expert:

Two things happen in that evolution. First, knowledge is built up incrementally on the basis of experience. Facts, once unrelated, get integrated through occurrence in the same episodes. Second, reasoning processes are refined, and usefulness and rigidity of rules is learned. Because experience is vital to the evolution from novice to expert, experience is organized in long-term memory, and guides reasoning processes ... When a person has only gone to school and acquired book knowledge, he is considered a novice. After he has experience using the knowledge he has learned, and when he knows how it applies both to common and exceptional cases, he is called an expert... Experience serves to turn unrelated facts into expert knowledge. (p .498)

In those activities that seemed more related to classroom teaching, the advanced beginners in this study more closely approximated the experts than did the novices. If this finding is replicated, there is an important conclusion to be drawn. Despite what some critics may say, it appears that programs of teacher training provide experiences that enable a trained beginner, more so than a raw novice, to perform more like an expert teacher. At least this is so in a simulated teaching task marked by simultaneity, multidimensionality, and immediacy.

References

Badre, A. N. (1982, July-August). Selecting and representing information structures for visual presentation. IEEE transactions on systems, man, and cybernetics, SMC-12(4), 494-504.

Benner, P. (1984). From novice to expert. Reading, MA: Addison-Wesley.

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 New directions for teacher assessment.- Proceedings of the 1988 ETS invitational conference. Princeton, NJ: Educational Testing Service.

Borko, H., & Livingston, C. (1989). Cognition and improvisation: Differences in mathematics instruction by expert and novice teachers. American Educational Research journal, 26, 473-498.

Brown, J. S., Collins, A., & Duguid, R (1989). Situated cognition and the culture of learning. Educational Researcher, 18, 32-42.

Carter, K., Sabers, D., Cushing, K., Pinnegar, S., & Berliner, D. C. (1987). Processing and using information about students: A study of expert, novice, and postulant teachers. Teaching and Teacher Education, 3, 147-157.

Carter, K., Cushing, K., Sabers, D., Stein, R, & Berliner, D. C. (1988). Expert-novice differences in perceiving and processing visual classroom information. journal of Teacher Education, 39(3), 25-31.

Chase, W. G., & Simon, H. A. (1973). Perception in chess. Cognitive Psychology, 4, 55-81.

Chi, M. T H., Glaser, R., & Farr, M. (1988). The nature of expertise Hillsdale, NJ: Erlbaum.

Clark, C. M., & Peterson, R L. (1986). Teachers' thought processes. In M. C. Wittrock (Ed.), Handbook of research on leaching (3rd ed., pp. 255-296). New York: Macmillan.

Copeland, W. D. (1987). Classroom management and student teachers' cognitive abilities: A relationship. American Educational Research Journal, 24(2), 219-2 36.

de Groot, A. D. (1966). Perception and memory vs. thought: Some old ideas and some recent findings. In B.-Kleinmuntz (Ed.), Problem solving (pp. 19-50). New York: Wiley

Doyle, W. (1986). Classroom organization and management. In M. C. Wittrock (Ed.), Handbook of research on Teaching (3rd ed., pp. 392-425). New York: Macmillan.

Dreyfus, H. L., & Dreyfus, S. E. (1986). Mind over machine, New York: Free Press.

Erickson, F. (1984, Summer). What difference does teaching experience make? IRT Communication Quarterly, p. 3.

Glaser, R. (1986). On the nature of expertise. In F. Klix & H. Hagendorf (Eds.), Human memory and cognitive capabilities: Mechanisms and performances (pp. 915-928). Amsterdam, The Netherlands: Elsevier-North Holland Publishers.

Gump, P. V. (1967). The classroom behavior setting: Its nature and relation to student behavior (Final report). Washington, DC: U.S. Office of Education, Bureau of Research. (ERIC Document Reproduction Service No. ED 015 515)
Jackson, P. W. (1968). Life in classrooms. New York: Holt, Rinehart, & Winston.

Kolodner, J. (1983). Towards an understanding of the role of experience in the evolution from novice to expert. International Journal of Man-Machine Studies, 19, 497-518.

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

Lesgold, A., Rubinson, H., Feltovitch, P, Glaser, R., Klopfer, D., & Wang, Y. (1988). In M. T.H. Chi, R. Glaser, & Marshall J. Farr (Eds.), The nature of expertise (pp. 311-342). Hillsdale, NJ: Lawrence Erlbaum.

Livingston, C., & Borko, H. (1990). High school mathematics review lessons: Expert novice distinctions. Journal of Research in Mathematics Education, 21, 372-387.

Mazzeo J. (1985). An evaluation of the level of skill required of operators of a computer-assisted radiologic total lung capacity measurement system. Unpublished doctoral dissertation, University of Arizona, Tucson.

Nelson, K. R. (1988). Thinking processes, management routines and student perceptions of expert and novice physical education teachers. Unpublished doctoral dissertation, Louisiana State University

Peshkin, A. (1988). In search of subjectivity-one's own. Educational Researcher, 17(7), 17-21.

Peterson, P. L., & Comeaux, M. A. (1987). Teachers' schemata for classroom events: The mental scaffolding of teachers' thinking during classroom instruction. Teaching and Teacher Education, 3(4), 319-333.

Ziman, J. (1978). Reliable knowledge. Cambridge, England: Cambridge University Press.