Final Report to the U.S. Department of Defense on the Defense Reinvestment Initiative (1999)

By Martin Haberman

University of Wisconsin, Milwaukee (February, 1991)

Between 1958 and 1961 as part of my work supervising student teachers in a fifth year, masters level program offered by Teachers College, Columbia in the Harlem schools of New York City, I reached the following conclusions:

1. Teacher education is not a generic process. The preparation of teachers for urban, multicultural schools is a distinctive enterprise.

2. Mature college graduates who are adults with work and life experiences are more likely to be successful in urban schools than traditional students in undergraduate programs.

3. Selection is significantly more important than training. It is easier and wiser to select people with attributes that will enable them to succeed in metropolitan schools than it is to expect that individuals who might be sexist, racist, uncreative, uninterested in the world of ideas, rigid, moralistic, humorless, or fearful will be transformed by virtue of completing a traditional teacher education program.

4. Urban teacher preparation actually occurs in schools, with children, while functioning in the role of teacher with the help of a coach or mentor and not as an undergraduate in a generic teacher education program.

5. Since exceedingly few college faculty have ever taught in urban, multicultural schools, the best teacher educators are practicing classroom teachers who are effective with urban children and youth.

During this period I thoroughly reviewed the literature related to the MMPI (Minnesota Multiphasic Personality Inventory) as well as other personality tests. Two of my mentors at the time were Irving Lorge and Robert Thorndike who helped me see that written tests of personality could not predict who would be an effective teacher and that it was conceptually not reasonable to expect personality dimensions to remain constant across different school situations. Previously, A. S. Barr and his doctoral students at the University of Wisconsin Madison in the 1930s and 1940s had also attempted to identify universal teacher attributes that would predict success but they could not, except in general terms,

relate teacher traits to pupil outcomes. By 1960 David G. Ryans of the University of Texas had captured the imagination of teacher educators with his exhaustive analysis of what makes a good teacher. Ryans ’ conclusions were stated as personal characteristics such as humor, enthusiasm, creativity, etc. which not only reflected personality dimensions but also attributes which began to relate to the work of the teacher. But the issue remained: Are these teacher characteristics which can be reasonably taught to large numbers of undergraduates in teacher education programs, or is it more reasonable to select and prepare individuals already predisposed to manifest these characteristics?

At this same time Robert K. Merton presented a sociological analysis of professions in which he put forward the concept of mid-range functions. In Figure 1, I have applied this concept to the task of predicting teacher success. At one extreme (left) are psychological approaches which purport to identify personality traits that individuals can be expected to demonstrate regardless of the situations in which they are placed. At the other extreme (right) are the situational demands of teaching as it is practiced in a specific school or institution. Merton argued that both extremes were dysfunctional. It is not possible to generalize from personality dimensions to how individuals will behave across the range of school situations they may encounter. We all know “ shy,” “aggressive,” and other types of teachers who might be successful in some schools but not others. At the other extreme, it is also useless to attempt to specify the precise behaviors required of effective teachers in a given situation. The numbers of such behaviors become too large for purposes of either teacher training or evaluation. For example, the State of Wisconsin has officially recognized 227 effective teacher behaviors and many more might be generated if specific situations were fully examined.

To negotiate between these extremes Merton advocated that each profession develop mid-range functions, that is, clusters, chunks, or groups of behaviors that particular practitioners must demonstrate in order to be effective. These are relatively small in number and while they manifest an individual’s personality, they are sufficiently behavioral so that observers can identify what effective teachers do.

Figure 1 BASES FOR PREDICTING TEACHER SUCCESS

Observing 124 student teachers over a three year period in New York City schools (1958-1961) enabled me to identify 18 “Stars” and 14 “Failures.” “Stars” were those individuals defined by all supervisors and cooperating teachers as equal to or better in performance than satisfactory, experienced teachers after only the first few weeks of their teaching. “Failures” were those who all others (cooperating principals, teachers, and college supervisors) agreed should not be teaching and who screened themselves out of teaching. By comparing these extremes, eight mid-range functions were identified: organizational skills, stamina, creativity, human relations, planning, discipline, teaming, and self analysis. Comparing the performance of these functions seemed to account for the difference between “Stars” and “Failures.”

In 1962 I began to further refine the process of identifying mid-range teacher functions and trying to translate them into questions that might be used in selection interviews. If we could identify potential success in urban teaching by using an interview, much time and effort might be saved and much inconvenience might be avoided. Individuals with high potential for urban teaching and with satisfactory undergraduate GPAs might be screened into teacher education while many others, also with satisfactory GPAs, might be screened out. Most of all, children and youth would not have their time wasted. In the traditional situation of utilizing inappropriate selection criteria, children and youth are, in effect, used as the screening mechanism. Once an individual has an appropriate skills level (approximately 8th grade reading and math levels are required in most states) and a satisfactory GPA in university study, there is no further screening until s/he actually fails with children and youth. Fifty percent of beginning teachers fail or quit in their first 3-5 years of urban teaching. As a result of this non-system of selection, it is children and youth who serve as the actual selection process. Because GPA, skills tests, grades in student teaching, and personal references are the typical selection criteria currently in use and because these criteria do not predict effectiveness in urban schools, we have a revolving door for teachers in urban districts. Children and youth most in need of greater stability in their lives are most likely to experience the most teacher turnover.

The first selection interviews were conducted in Milwaukee in 1962. I had been asked to develop a fifth year intern program for liberal arts graduates to become teachers in the Milwaukee Public Schools. The model we developed became the National Teacher Corps in 1965. In these 3 years (1962-1965) 108 interns were admitted and followed through their first year of teacher practice. In this period we refined interview questions to more accurately reflect mid-range functions that distinguished between outstanding teachers and failures. We evaluated the interview instrument by checking our initial predictions against how interns actually performed in their subsequent teaching. In this three year period the interview was a research tool for evaluating predictions and not actually used to screen people in or out of the program. The 108 admitted (on traditional criteria) were then ranked by two interviewers in terms of how well they did on our interview. The following scheme was used to categorize our predictions.

Figure 2

The comparison of “before” assessments based on our interview against “after” evaluations based on actual teaching performance can be seen in Figure 3.

Figure 3 COMPARISON OF ORIGINAL INTERN INTERVIEWS WITH ACTUAL PERFORMANCE

Our operational definition of an error was the following:

1. Assessing any individual as any category other than Failure who subsequently fails.

2. Assessing any individual as a Failure who subsequently achieves any higher level in practice.

3. Misplacing a candidate by more than one category. This means that a person assessed as a Star who proved to be High would not be considered an error unless he proved to be Average and vice versa.

Defining these misassignments as mistakes, we made 3 errors on the first 108 candidates interviewed.

“Performance” was operationally defined as the sum of the evaluations of those who had actually observed and supervised the interns: the cooperating teachers, principals, and college supervisors. The self evaluations of interns were also considered.

Beginning in 1966 the Urban Teacher Selection Interview was used to select college graduates for the Milwaukee Intern Teaching Program. At the time of the program’s termination in 1973 approximately 1,500 interns had completed the interview and the program.

Interviews were always conducted by two professionals: one faculty member representing the School of Education of UW-Milwaukee and one representative of the Milwaukee Public Schools, usually a central office supervisor. During this period several classroom teachers who served as cooperating teachers were also trained to conduct the interview and to serve on the two-person interview teams.

In the eleven years the program was operational there were five different Directors who utilized different pairs of individuals to conduct these selection interviews.In every year Directors reported less than a 3 percent error between prediction and performance. As training of interviewers has become systematic, this error has dropped to approximately 1 percent.

In 1966, the Great Cities Research Council was comprised of fifteen great cities each represented by its Superintendent of Schools and the Dean of the School of Education of a major teacher preparation institution in each of the cities. Superintendents frequently invited their deputies to represent them at these meetings and Deans were sometimes represented by a designated faculty member. At that time, Milwaukee was the thirteenth largest school system and I attended these meetings as a representative of the UWM School of Education. Evelyn Carlson was the Associate Superintendent of the Chicago Public Schools and was also a frequent attendee. As is common during many periods, the Chicago Public Schools was short of regularly prepared, certified teachers. In 1966, approximately 1,000 college graduates without teacher preparation were hired and appointed on probationary licenses. With the help of Ms. Carlson, I followed-up this population. In May of 1967 there were only 167 of these individuals still teaching. We interviewed this population and discovered that they had not all survived their first year on the job (with no special help or supervision) because they were particularly effective. Indeed, most of them were “strong, insensitives.” Comparing our interview against the ratings of supervisors, we found less than a 5 percent error.

Figure 4 PREDICTING THE SUCCESS OF CHICAGO FIRST-YEAR TEACHERS WITHOUT TEACHER PREPARATION (1967)

At this point it would be useful to review the nature of the interview questions and how they were transformed and refined as a result of the numerous trials. First, it must be recognized that two of the most critical Mid-Range Functions have never been utilized in the interview because we have never been able to develop questions which adequately assess or predict their manifestation in subsequent practices: these are, organizational skills and stamina.

Organizational skills refer to the myriad of ways in which Stars systematize their classrooms and the rules they set up to govern pupils living together on a daily basis. These organizational skills refer to how materials are stored and used, how equipment is used and cared for, seating arrangements, the allocation of time, procedures for passing in work, pupil movement within and outside of classrooms and the numerous other procedures which successful teachers organize effectively and which unsuccessful teachers find to be a continuing source of problems and disruptions.

Stamina refers to both the physical energy required of those who teach all day, every day and the emotional strength required to withstand the unending shocks to the payche and emotions as one deals with pupils of boundless energy who are frequently hungry, without adequate medical or dental attention and frequently uncared for or mistreated. For educators to strive to do their best and not succumb to helpless pity or be overwhelmed by their pupils’ unmet needs requires much physical strength and strength of character.

In over thirty years of comparing Stars and Failures, two truths have remained constant. First, there are no Mid-Range Functions more critical to the success of an urban teacher than organizational skills and emotional/physical strength. Second, there are no ways we have ever devised which will enable us to interview individuals and predict their

organizational skills and stamina. These functions seem to be beyond our ability to create interview questions. Fortunately, there are seven other Mid-Range Functions for which interview questions have been developed and refined and which do predict subsequent success in urban teaching. Following is a brief review of how the original Mid-Range Functions were clarified and refined as a consequence of the iterative process:

Creativity was originally used to describe the first Mid-Range Function. We soon realized that successful teachers demonstrated more than alternative-seeking behavior. They were manifesting problem solving skills; the ability to define and evaluate as well as the ability to generate options. After a few years we realized that we were also seeing teachers demonstrate these functions with tenacity and commitment. In recent years we have used the term “Persistence” to denote both the creative and problem solving functions and to emphasize that the work of the teacher requires their continuous application. As used in our interview, therefore, “Persistence” refers to the total process of continually seeking solutions to never-ending problems.

The second Mid-Range Function was originally identified as human relations skills and referred to the effective teacher’s ability to get along with other adults, colleagues, and administrators in the school setting. Through subsequent trials this skill was refined further since substantial numbers of Failures were sometimes liked by other adults. The critical dimension here is not popularity or “getting along” but how effective teachers protect their pupils’ right to learn in situations which may be contrary to school rules or norms. This Mid-Range Function is particularly relevant in urban schools which are highly bureaucratic and overly organized. To better pinpoint the function the term became clarified as “Response to Authority ” and refers to the effective teacher’s willingness to support student learning in the face of or even against school policy.

The third Mid-Range Function refers to the effective teacher’s ability to apply generalizations about teaching, learning, and development of his/her particular classroom. Originally, this function was termed “planning” but by the iterative process of trial and retrial it soon became clear that what was meant was beyond traditional notions of teacher planning. Successful teachers were consistently distinguishable from failures by their ability to apply principles to practice, orto generalize from their practice about the principles they were demonstrating.

The fourth Mid-Range Function evolved into two parts. Originally labeled discipline, it soon became evident that what was meant was the effective teacher’s ability to not blame the victim and accept accountability for teaching all children. At the same time effective urban teachers realize they cannot love every pupil but are still responsible for teaching even the less lovable and especially the unlovable.

The fifth function was originally termed “teaming”, then “bureaucracy”, and finally, “burnout.” This evolution refers to the continuing distinction between Stars and Failures to explain and act on their perception of the causes and cures of low teacher morale.

The final function was originally termed self-analysis. This has evolved into fallibility to more adequately account for both the ability to study one’s behavior and motives and to recognize and accept oneself and others as human beings.

In sum, the actual Mid-Range Functions have not changed. The original distinctions between Stars and Failures have remained constant. What has been refined through the numerous trials and validations of the interview over the years have been the terms used to identify the Mid-Range Functions and the questions used in the interview to get at these functions. Figure 5 summarizes the original and current terms used to denote the seven components of the interview. The interview document itself and its accompanying procedures of administration further clarify this development.

For a more complete explanation of the Mid-Range Functions, see “Interview Questions to Accompany the Urban Teacher Selection Interview Continua” and the “Urban Teacher Selection Interview” itself.

Figure 5 ORIGINAL AND CURRENT TERMS FOR IDENTIFYING MID-RANGE FUNCTIONS

The interview is not scored by assigning points. Each interviewer places an X on each of the seven continua. At the completion of no more than three interviews both interviewers discuss and agree upon a ranking of the three candidates. The candidates are not only ranked but also assigned a category (Star, High, Average, Failure, as in Figure 1). Each subsequent set of three interviews requires the interviewers to fit the additional 3 candidates into the total rank order and to place each candidate in a category. Both the ranking and categorization of candidates is done by discussion and mutual agreement

between the two interviewers. If the interviewers agree that a candidate goes off the extreme left on any continuum on any one of the seven functions, the candidate is considered to have failed. The seven functions are divided into two subparts yielding a total of 14 continua. Each candidate must, therefore, avoid being rated on the extreme left end of any continuum on all fourteen subparts. Again, the basis of this procedure is the teaching behavior of successful Stars as that behavior is distinguishable from Failures. It is easy (and appealing) for readers to simply “feel” or “believe” that they agree or disagree with these functions and overlook the basis of their derivation.

Urban school districts, particularly those in states with alternative certification, have used the interview to help select college graduates without teacher training as beginning teachers. This is the population and purpose for which the interview was developed and its most appropriate use. In some cities candidates selected as teachers also have coaches or mentors to help them and in some places there is also supporting coursework. In many cities both on-site coaches and some classes are offered to beginners. So long as the individual being interviewed is a college graduate and the teaching position subsequently performed is in an urban school, the interview is being used with the population and for the purpose it was developed.

Teacher educators in “regular” university teacher education programs have long requested me to apply this instrument to the prediction of success in student teaching in urban schools with undergraduates. I resisted this pressure for many years because student teaching seems to me to be an inadequate criterion of effectiveness. It is possible for a poor, even inadequate student teacher to appear better than s/he really is because the classroom is actually set up and managed by someone else–a coperating teacher. It is, in effect, not a fair test of the student teacher and, therefore, an inadequate criterion for judging the interview ’s ability to predict success. A second reason for my reluctance is the relative ease of the student teaching experience when compared to the extraordinary pressure on a beginning urban teacher without any teacher training. If I have an instrument which predicts who can swim across Lake Michigan, I would not consider it a fair test of the instrument to use it to predict who can swim two laps of an indoor, heated pool. Major differences in degree create a new order of activity. Swimming is not swimming. Teaching is not teaching.

I first attempted to try the instrument “against” subsequent success in student teaching in Fall 1985 with a group taught by Jack Stillman in the School of Education, UWM. At this point I began using Dr. Linda Post as my interviewing partner. We interviewed, categorized, and ranked 14 of them. Professor Stillman ’s personal judgment of how well the student teachers achieved was the criterion of “success.” He ranked and categorized his students at the end of their student teaching. The results were as follows: we concurred on the single Failure; and disagreed on the number of Stars (2 versus 4). Our rankings were essentially

identical with his categorization being somewhat higher; that is Stars, including Rank 4 and Highs, extended several ranks deeper into the list. Given the ways errors in categorization are calculated, this trial was 100% accurate. (See Figure 3). I then undertook several subsequent trials with undergraduates in regular programs against the criterion of college supervisors ’ ratings of success in student teaching.

We also conducted two trials with student teachers in Exceptional Education. One group had 22 and a second 26. With the exception of one disagreement by interviewers, there was 100 percent prediction against the final evaluations of student teacher supervisors.

Continuing with Dr. Linda Post as my interviewing partner, we began a summer program in human relations training for undergraduates in University of Wisconsin System institutions. These students were required to take classes but, more importantly, to work in the summer schools of the Milwaukee Public Schools for six weeks. These were all undergraduates, mostly sophomores, having their first direct experiences as well as their first urban, multicultural experiences.

The criterion of success was the judgment of college supervisors regarding students’ interactions with children during this six week period. In Summer 1989, we ran a trial of 22 students. In 1990, we interviewed sixteen. In both cases there were minor differences between our initial ratings and subsequent practice. These differences merely refer to how deep into the rank order a particular category (i.e., Star, High, Average) went into the list. In effect, the supervisors ’ ratings were somewhat higher than the initial interviewers’. In both years there was again 100% predictability between the initial interview and subsequent practice. This means, as errors in categorization are calculated, there were no Failures overlooked and no individuals whose categorization was miscalculated by more than one category.

Sporadically, since 1985 there have been individual and a small groups of student teachers (up to five) who we have interviewed and compared against their subsequent student teaching success.

These trials with student teachers are of little interest to me. As noted earlier, the fact that we have been totally accurate in predicting their “success” has in no way convinced me that student teaching is a worthwhile criterion to use since the large number of student teachers who are judged to be “A” and who subsequently fail in urban schools is too large to ignore. The almost total dependence on the cooperating teacher makes this experience a flawed model of preparation for urban teaching.

Unlike instruments thought to be comparable, there is no per-interview cost to the school district or institution that chooses to use this interview. Consultant companies

currently charge as much as $3,000 for a total processing of each candidate for a position.

The only costs involved in using this interview is the training of the interview pairs. This requires 1 day in which the trainees are trained to question and score candidates on each of the mid-range continua. The trainees learn and practice with carefully developed videotapes and coached by authorized trainers in the precise way of asking the questions. Meeting proficiency at the conclusion of the training enables all certified interviewers to have unlimited use of the interview with no additional cost. All persons, wanting to use the interview correctly, must become certified interviewers by successfully completing the authorized training session. Follow-up and up-date training sessions are scheduled and announced as needed.

Several urban school districts are currently involved in using the interview with adult college graduates hired as teachers in states with alternative certification. These trials are ongoing and provide continuous feedback regarding questioning techniques and interviewer training.

I work closely and personally with the Urban Teacher Education Program in the school districts of Gary, East Chicago, and Hammond, Indiana. In these districts the interview is being tried with college graduates in an experimental certification program. These trials will also yield valuable data.

At the University of Wisconsin-Milwaukee we currently have a trial group of 21 college graduates who are all minorities. When these individuals begin teaching we will have another valuable data set.

In sum, the interview has evolved for over thirty-two years spanning four decades of change in urban schools in cities across the United States.

The precise wording of questions used in the interview continue to be refined. The actual Mid-Range Functions performed by successful urban teachers remain constant.

Dr. Martin Haberman is a teacher educator who has been directly or indirectly involved in shaping every major development in American teacher education over the past thirty years. These include developing:

• Some of the original MAT programs for liberal arts graduates;

• The NCATE standards for accrediting schools and colleges of Education;

• The training of teacher trainers (TTT) program;

• Professional Development Centers planned and operated by teachers;

• Various forms of alternative teacher certification programs; and

• Specific programs for recruiting and preparing more blacks and Hispanics.

Martin Haberman grew up in New York City. His formal education includes bachelors and masters degrees in sociology from Brooklyn College and New York University. These were followed with a second masters and a doctorate in Education from Teachers College Columbia University. It is clear that his advisor, Florence B. Stratemeyer, helped shape his commitment to educating a free people.

In addition to an extremely long list of publications (7 books, 38 chapters, 120 articles) and numerous research studies, Haberman has engaged in some divergent contributions. These include his extensive use of media–particularly radio–in reaching teachers, six years as the Editor of the Journal of Teacher Education, and eleven years as a dean in the University of Wisconsin trying to apply the successes of extension in rural America to the problems of life in urban areas.

Professor Haberman has served on eight editorial boards. He holds several awards for his writing, a Standard Oil Award for Excellence in Teaching, a special award from The Corporation for Public Broadcasting and AACTE Medals for offering a Hunt Lecture and the Pomeroy Award (1990). In January, 1989 Rhode Island College awarded him an honorary doctorate. He is a distinguished member of the Association of Teacher Education and a Laureate member of Kappa Delta Pi. The most widely known of his developments was The National Teacher Corps, which was based on his intern program in Milwaukee and which brought hundreds of thousands of college graduates into urban schools all over America. Currently, his developmental efforts are focused on helping to resolve the crises in urban schools serving twelve million at-risk students. By working directly with teacher unions, school districts and university researchers, he is demonstrating new forms of teacher preparation based on practice and coaching rather than on traditional university coursework. If successful, these research-demonstrations will radically alter the nature of teacher education and certification in America.

Professor Haberman’s work represents a continuous effort to make the preparation of teachers not only more effective, but also more relevant to schools in a multicultural society. In effect, his career has been an effort to transform teacher education into an instrumentality with redeeming social significance.