Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 Statistics Related to the Quality of Science and Mathematics Teaching As noted in previous chapters, the supply and demand for teachers of mathematics is brought into equilibrium in the short term by adaptations in the selection criteria for teacher or teaching quality. Thus a school system unable to hire science and mathematics teachers at a preferred quality level will have to lower its minimum quality requirements. Conversely, school systems facing a supply of teachers of acceptable quality in excess of the number they need will be able to choose those at the top of their quality scale, thus ending up hiring teachers of higher quality than suggested by their minimum criteria. While this comprises a generally accurate description of school system hiring practices, it does not tell us anything at all about what factors go into quality teachers or quality teaching. It is to that topic that we now turn. It should be recognized from the beginning that we do not have very precise notions about what constitutes teacher or teaching quality, and thus we cannot provide definitive prescriptions as to types of data that need to be obtained in order to monitor either the level of teacher quality that exists or changes over time in quality. The problem is that assessment of quality is an extraordinarily difficult enterprise, and existing research does not go very far in identifying the factors that determine quality. It is the panel's view that the right dimensions of teacher or teaching quality are factors that produce a positive influence on student outcomes that is, higher quality in our view should be defined to mean better student outcomes, given the influence of other forces besides teachers or school system factors that influence student outcomes. Perhaps the best way to summarize the current state of knowledge on this topic is to note two sets of facts that come from existing studies of teacher quality. 116
STATISTICS RELATED TO QUALITY 117 1. Teacher quality matters a good deal to student outcomes, in the sense that it is possible to identify teachers who have produced well below average outcomes. In this context, i~lentib simply means that specific teachers can be shown to produce relatively good outcomes, and other specific teachers can be shown to produce relatively poor outcomes (Contra and Potter, 1980~. 2. If one tries to describe what factors are associated with teachers who produce good outcomes or bad outcomes, one finds very little associa- tion between particular characteristics of teachers and the resulting student outcomes. That is, better formal credentials, better preparation in terms of course work more years of teaching experience, better scores on standard tests of teacher qualifications, etc., do not generally show up as teacher characteristics that are strongly related to better or worse outcomes (Druva and Anderson, 1983; Hanushek, 1986, 1989~. It has often been found that teacher verbal ability is positively related to better student outcomes, but the relationship is not exceptionally strong; most other factors do not show up at all (Darling-Hammond and Hudson, 1986~. In sum, we know that there must be characteristics of teachers or of classroom situations that produce better student outcomes, and qualities or characteristics that produce worse student outcomes, but we do not know what these characteristics or qualities are with any degree of assurance. Although it may be surprising to some readers that so little is known about what factors are related to teacher or teaching quality, a little re- flection suggests that it is not so unusual that the state of knowledge is so limited. If one were to ask whether some people are more effective social workers and others less effective, whether some people turn out to be very successful business executives and others less so, or whether some people are very successful at doing survey research interviews and others are less successful, the answer in all these cases will surely be that there are very large differences in the degree to which people are successful or unsuccessful in particular kinds of professional activities. If one goes further to ask what factors are associated with success in being a social worker, a business executive, or a survey research interviewer, the answer will commonly be that very little is known about why some people succeed and others fail. The probable reasons are that the factors making for success are complicated, that personal characteristics and characteristics of the particular environment interact and may be idiosyncratic to particular situations or types of work environments, and that success has a lot to do with motivation, energy, striving for success, interpersonal skills, and
118 PRECOLLEGE SCIENCE AD ~THE~TICS TRACHEA myriad other factors that come together in subtle ways to produce better or worse outcomes.] Given this state of knowledge, what should be done about the collection of data that relate to teacher or teaching quality? It is the panel's view that, although little is known about what factors are importantly related to quality, something is known about the kinds of factors that probably play some role in determining quality. We should try to collect the best such set of factors, recognizing that the data collected will not be sufficient to do a satisfactory job of explaining student outcomes. Thus in ' this section we discuss a number of types of data that are probably related to quality, although they have not been convincingly shown to be either strongly or systematically reliable indicators of quality. These results may be caused by systematic errors: for example, the better teachers teach higher-order skills, but tests measure primarily lower-order skills, so the quality difference in teaching is not measured. The reader will note that we have talked about quality both in terms of teacher quality and teaching quality. The two are not synonymous. By teacher quality we mean those personal characteristics of individuals that enable them to be more effective in classroom settings: education level, subject matter knowledge, interpersonal skills in working with students, degree of inservice training, formal credentials, etc. By teaching quality we have in mind a somewhat broader notion that encompasses not only teacher characteristics but also the school setting in which classroom teaching takes place. Thus teaching quality includes factors that are beyond the control of the 'individual teacher: disciplinary norms of the school system or of the building principal, support given by principals to teachers, the presence or absence of inse~vice training opportunities or opportunities for interaction among teachers, types of textbooks that are selected for use in the school systems, amount of time allocated to each subject, number of classroom hours taught, and so on. Thus, teaching quality encompasses factors that 1The nature of the problem is illustrated by the example of survey research interviewing. This subject has been studied for many decades, and what we know with certainty are only a few relevant facts, none of which is sufficient to design a test to predict success at survey research interviewing. There are enormous differences in degree of success. Some interviewers achieve close to a 100 percent cooperation rate and have virtually no refusals, collect consistently high- quality data, and do so with relatively few hours expended in the interviewing task and thus have lower costs. Other interviewers have extremely high refusal rates, do not collect consistently high-quality data, and take a great many hours to produce relatively mediocre results. Although we know that these differences exist, it has not been possible to identify personal characteristics that would enable survey research organizations to predict who will be a good interviewer and who will not. Conventional demographic characteristics (educational level, experience, age, etc.) are of virtually no use in explaining success. Although a few personality characteristics seem to have some association with success, the state of knowledge is still relatively crude, despite a great deal of methodological work.
STATISTICS BELA TED TO QUALITY 119 are not within the control of individual teachers, while teacher quality includes only those factors that relate to the personal characteristics of individual teachers. In examining the quality of mathematics and science teachers, we have in mind a broader notion than assessing the quality of teachers who specialize in mathematics or science. Although some districts employ teachers who specialize in science or mathematics as early as the fourth grade, most teaching in mathematics in grades K-8 is done by teachers in either elementary or middle school who may not be classified as science or mathematics teachers, but rather as teachers who teach science and mathematics. The distinction is important: we are interested in assessing the quality of mathematics and science teaching on the part of teachers who teach those subjects, and many of them probably most- are not specialized in the teaching of either science or mathematics. Moreover, we are also interested in those dimensions of quality that relate to preferences of the school systems for the types of teachers they wish to hire. It is clear enough from our case studies, as well as from extensive discussions with the personnel directors of large city school systems, that mathematics or science teachers are not hired solely for the perceived quality of their mathematics or science teaching. Many school systems have other dimensions of teacher performance in mind when they hire teachers. In some school systems, the ability to fit in with the community is important; in some, the ability to teach other subjects or to direct extracurricular activities is important; in some, the ability to work with the types of students in the school system is perceived to be extremely important. The basic point is simple enough: school systems do not hire teachers to teach science and mathematics solely because of their perceived ability to be effective in classroom settings. Rather, hiring decisions are influenced by a great many other factors, some of which will necessarily result in hiring people who are likely to be less effective in teaching science and mathematics than teachers who were not hired because they lacked other skills or characteristics. In the remainder of this chapter, we attempt to sort out the major ingredients of teaching and teacher quality that call for further data. We look first at school system policies and practices and the school-level condi- tions that can affect teaching quality. Next we look at the qualifications of incoming teachers their college and professional preparation, their level of achievement in science and mathematics, their cognitive abilities, and so on. Finally, we examine other factors that also influence student outcomes but do- not fall neatly under either school system policies and practices or teacher qualifications and characteristics: curriculum and textbook selection issues, time-on-task issues, and issues relating to the home environments of students. All of these do or may influence student outcomes to a substantial
120 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS degree, and none is likely to be under the control of either the teacher or the school principal. SCHOOL SYSTEM POLICIES AND PRACTICES The assignment of a teacher to courses and pupils appropriate to the individual's educational background, certification status, and experience is crucial to quality instruction in precollege mathematics and science. But district personnel policies, budget constraints, and other external factors can impede the ability to achieve the most effective match. A policy maker with the specific goal of higher~uality instruction often finds that it is difficult to change many of the policy variables that affect the quality of instruction. District policies exist in a complex web of competing goals and pressures. Even if the central goal is quality teaching, the policy maker must also consider school system policies and union contract provisions regarding recruitment, initial assignment, and transfer and retention of teachers. A given set of policy guidelines can have quite different effects depending on whether enrollment is stable, growing, or declining. For example, seniority rules for assignment or transfer have different effects in environments in which enrollments are rising or declining. Personnel policies are also affected by the enrollment size of a particular school system, the enrollment size of a high school, the extent to which the curriculum is taught by specialists, and the match among educational background, teaching assignment, and teacher and student cultures. Recruitment and Hiring Practices Certain policies set by the school district, teacher organization, or state school finance plan can have deleterious effects on the ability to hire the most talented teachers. The examples given here apply not only to science and mathematics teachers but probably also to teachers in general. Discussions with personnel offices of large school systems suggested that recruitment of new teachers by large districts with diverse student pop- ulations was often hindered by the fact that recruiters could not specie the school to which the applicant would be assigned. Many persons would find such a school system desirable only if they could teach in a given section of the school system or in a specified school. Since recruiters could not make such commitments, or could not make those commitments early enough in the recruitment period, candidates were lost to the school system. This problem stemmed from district policies related to the timing of hiring, in- te~viewing, and specific placement. District policy in some systems requires
STATISTICS RELATED TO QUALITY 121 the applicant to be interviewed 'only by the district administrator; subse- quent assignment is a central office decision. Other district administrators screen applications and refer promising candidates directly to principals, who conduct the interviews. The uncertainty of initial assignment also seemed to be exacerbated by seniority rules of internal transfer. In one medium-sized school district in a western state that participated in our case study analysis, internal transfer rules took months to implement. With a tendency for junior high science and mathematics teachers to request high school positions, and for elementary teachers to request junior high positions, the process of considering all transfer applications and then determining which positions were actually vacant continued well into the summer. Job offers could not be made until August. Since other districts could make job offers in March and April, this district was left with candidates who had not obtained positions elsewhere. In some circumstances, the problems stemming from seniority rules become especially severe when combined with rehiring rights after teachers have been laid off due to enrollment decline or financial constraints. In such circumstances, district rules, regulations, and practices rather than pro- fessional judgment often seemed to determine the match between teacher and classroom assignment. For example, seniority rules may restrict new hires to the least desirable schools in the district. These rules may drive teachers not only from the school system but also from the profession. Se- niority rights may also prevail when teachers are transferred among schools. When vacancies occur, the teacher with the greatest longevity in the school system may have first choice. When enrollment declines, teachers with higher longevity in the school system, the school, or a teaching field may have rights to bump less senior teachers. The length of the waiting period before opening vacancies to outside applicants greatly affects the district's ability to sign on talented applicants. Many officials said they lose good annlicants to cipher districts whose rules or budgets allowed them to hire err -^ -^ sooner. Enrollment size and composition also influence district policies. The hiring restrictions of one large urban school system in the West contrasted starkly with the innovative practices for meeting future needs employed by a small suburban school system in the same region. The suburban superintendent, in conjunction with a nearby college, recruited well-trained graduates to fill projected vacancies. The smaller enrollment size and relative wealth of the suburban school system, as well as the homogeneity of the student population, accounted for the differences in practices between the suburban and the urban systems. In another suburban school system in the East, a teacher who attracted high school students to advanced science classes had been allowed to develop his own teaching assignment. Such
122 PRECOLLEGE SCIENCE AND 1~4THEMi4TICS TEACHERS flexibility is less likely in a larger school system concerned with uniform course offerings among schools. One of the reasons for more rules, and sometimes less flexible ones, in larger school systems is the need to adhere to goals of equity among staff members in conditions of employment. Factors external to the school district can also affect local hiring prac- tices. Increases in state-mandated graduation requirements for mathematics or science can cause the district to fill vacancies in those fields with teach- ers not yet certified in the particular subjects, in order to meet the state requirement. As noted in Chapter 2, 42 states have added requirements in science or mathematics since 1983. The Center for Policy Research in Education (CPRE), which has surveyed the states' graduation requirements, has found that in schools affected, about 27 percent of students are taking an extra mathematics course and 34 percent an extra science course (CPRE 1989:33~. Many of these student are middle- to low-achieving, the CPRE study relates (p. 35~. CPRE inquired as to the nature or level of the additional courses. In many instances the added courses were remedial or lowerlevel science and mathematics courses (p. 35-36~. The increased requirements undoubtedly have changed schools' staffing patterns and course assignments and have probably affected hiring practices for science and mathematics teachers. State-mandated minimum competency test scores and state school- finance formula constraints on local funds for laboratory equipment and supplies, computers, teacher aides, or teacher salaries are other external factors that local personnel officials must take into consideration in hiring teachers. An unintended consequence of decisions made under these conditions may be a loss in teacher or teaching quality. Of course, not all rules act to restrict supply or make the task of matching persons and assignments more difficult; certain rules may benefit some school systems. When there is a potential for future growth in high school enrollments, teachers in a school system may pursue advanced study so that they can move from elementary school or junior high to high school. Other teachers may be attracted to begin their career in the district with a thought toward future advancement. Without seniority rules, there would be no such encouragement, as new hires might occupy newly created positions in high schools. Data are needed to better describe the incidence of these and other policies and practices that affect the ability to hire and place the most promising candidates to assure instruction of high quality. The Schools and Staffing Survey (SASS) does not yet provide data related to most of these areas. In-depth conferences with a sample of SASS districts on a regular basis are recommended (see Chapter 6) to gain more accurate insights into the use of such policies and practices.
STATISTICS RELATED TO QUALITY 123 Misassignment of Teachers Teacher assignment is critical to quality instruction in all subjects, es- pecially so for science and mathematics. Misassignment of science teachers can occur when a vacancy in a science specials is filled with a certified science teacher who is unfamiliar with that particular field. High schools may be too small to have a full-time chemistry or physics teacher or even a full-time biology teacher.2 In 1986-87, only 13 percent of teachers who taught physics in secondary schools had teaching assignments in physics alone. Almost two-thirds of the teachers who taught physics had their primary concentration of classes in chemistry, mathematics, or general and physical science (American Institute of Physics, 1988:17~. There may be a need for one but not two science teachers. The same type of misassign- ment can occur in mathematics, when a teacher is trained to teach areas of mathematics other than that assigned or some other subject altogether. In many states, it is legal to assign a teacher to teach part time in an area in which the teacher is not certified, under a practice called out-of-field teaching as opposed to "misassignment" (Robinson, 1985~. Estimates of the prevalence of misassignment based on data from the early 1980s collected by the National Center for Education Statistics (NCES) and the National Education Association (NEA) vary considerably. In a preliminary report on indicators of precollege education in science and mathematics, the National Research Council (NRC) notes the erosion of the quality of the existing teaching pool by misassignment of newly certified teachers. This report cites NCES findings that, among bachelor's degree recipients in 1979-80 who were teaching elementary or secondary 2 In one of the case studies, the employment of a full-time chemistry teacher by a school system was mentioned. This condition was treated as rare for the school systems studied. Such employ- ment can be seen as unusual for the United States by examining some necessary conditions. If one assumes that a teacher teaches 5 classes and that a class has between 25 and 30 students, then to teach a single subject at the same grade level requires 125 to 150 students per grade level. For a 4-year high school this means a school enrollment size of 500 to 600. For a 3-year high school, it means an enrollment size of 375 to 400. In 1982-83 9.5 percent of secondary students attended schools below the latter size criterion. An additional 10.5 percent of secondary students met the former criterion. If only half of the students take a chemistry course, then slightly more than half of the students, 53.3 percent, attend such secondary schools. If only a third of the students take a chemistry course, then only slightly more than 10 percent of secondary students (13.4 percent) attend schools of that enrollment size (ACES, 1986:68~. That only a third of secondary students are likely to take a chemistry course can be garnered from the fact that 65.4 percent of public secondary school students take natural science (p. 41), and the average number of Carnegie units (a standard of measurement that represents one credit for the completion of a one-year course) in natural science is 1.9 (p. 44~. Expanding the ranges of possible courses in natural science to include two courses in chemistry or chemistry and physics would indicate that only 3.9 percent of schools, that is, the schools with larger enrollments that enroll 13.4 percent of the students, would be able to hire a full-time chemistry or physics teacher.
124 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS school full time in May 1981, only 45 percent of science teachers and 42 percent of mathematics teachers were certified or eligible to be certified in the field in which they were teaching (NRC, 1985:52~. More recently, Darling-Hammond and Hudson (1987a:21) reported "estimates that vary depending on who is asked to estimate the degree of misassignment (school administrators versus teachers) and on how misassignment is defined." They reported (1987a:21~: · Not certified in area of primary assignment: 9-11 percent by teacher report, 3.4 percent from central office administrators' estimates (NEA, 1982; NCES, 1985a). Not certified for some classes taught: 16 percent by teacher report (NEA, 1982). Less than a college minor in area of primary assignment: 17 percent by secondary school teacher report (Carroll, 1985). The 1985 National Survey of Science and Mathematics Education found higher proportions for science and mathematics 18 percent of grade 7-9 mathematics teachers and 14 percent of grade 10-12 mathematics teachers teach courses for which they are uncertified. For science teachers, the percentages are 25 for grades 7-9 and 20 for grades 10-12 (Weiss, 1987:77-88~. Transfer policies can sometimes lead to a misassignment and thwart a teacher's potential for advancement. In our contacts with school district administrators, a tendency was reported for principals to transfer teachers from subject fields of surplus to subject fields of need. Often, these trans- fers moved the teachers from their primary subject fields to different areas. Transfers of this nature took place due to changes in student demand for subjects under stable enrollments as well as in times of changing enroll- ments. Such transfers also occurred because principals sought teachers able or willing to handle extracurricular tasks such as athletics, the school paper, the yearbook, or student clubs. The extent to which misassignment occurs today in science and math- ematics may be greater than for other subjects. Data on the extent of misassignment for all fields at the school district level will be obtainable from the SASS Teacher Demand and Shortage questionnaire. It will also be possible to estimate misassignment by field by using the SASS teacher questionnaire. This questionnaire obtains courses currently taught by each departmental teacher, the teacher's arrears) of certification, and college major and minor. Estimates of misassignment by field as defined by certi- fication status can be made using these data. Since certification standards vary so much across states, the fact that one was not certified in the field in which one is teaching does not necessarily mean misassignment. To obtain a more complete picture of
STATISTICS RELATED TO QUALITY 125 misassignment, information on inservice training and actual course-taking preparation should also be analyzed, as Darling-Hammond and Hudson suggest (1987a: 21-22~. The SASS teacher questionnaire represents a promising step forward. It requests data not only on certification status (as above), but also on degrees earned and major and minor fields of study, amount of course work in primary and secondary teaching assignment fields, and, for teachers who teach any science or mathematics courses, the number of graduate and undergraduate courses taken in various categories. These are rich data to examine misassignment and out-of-field teaching. Information from SASS should be analyzed together with state certifi- cation data on the number of emergency certificates issued in science and mathematics; 46 states allow emergency certification. Of these, 30 require university course work in order to renew and work toward full certification (McKibbin, 1988:32~. Supplementary data would include state rules on the extent to which out-of-field assignment is legal. Such information from various sources, when analyzed jointly, will help monitor the extent and trends of misassignment in science and mathematics teaching. Providing for Inservice and Continuing Education Some of the most important district and school practices that affect the quality of instruction are those directed to teachers already in place. maintain quality instruction throughout their careers, teachers require professional support from their schools and districts. This support includes working conditions, facilities such as laboratories, materials and supplies, collegial and administrative support, resources for continuing education, and opportunities to influence decision making (Darling-Hammond and Hudson, 1987a:27-37~. District practices regarding inservice and continuing higher education for teachers in place affect teacher quality directly and can make it more or less attractive for a teacher to continue in a district. School districts have been the primary sponsors of inservice programs, but such programs are highly vulnerable to district budget cuts. Decisions as to what kinds of inservice education to fund with a limited budget affect teaching quality in ways that data alone may be unable to illustrate. In one large, suburban, low-wealth district we studied, much of the staff development budget was geared to weaker teachers. Teachers had little release time during the school year 17 days allotted for each high school. Only about 20 percent of staff development was used for college-level course work. A national commitment to teachers' continuing education appears to be missing. The federal government does support inservice education through the Title II program of the Education for Economic Security
126 PRECOLLEGE SCIENCE AND M'4THEAL4 TICS TEACHERS Act of the Department of Education and through the National Science Foundation (NSF) Teacher Enhancement Program (Office of Technology Assessment, 1988:69), but funding for both activities is severely limited. Appropriations for the Title II program have been uneven, dropping from $100 million in fiscal year 1985 to $42 million in 1986, then $80 million, $120 million, and $127 million in 1987, 1988, and 1989, respectively (OTA, 1988:123; U.S. Department of Education, 1989~. These are small amounts when viewed on a per-pupil or per-teacher basis. The Office of Technology Assessment notes by comparison that a $40 million education program equates to a spending of $1 per pupil or $20 per teacher (1988:123~. NSF's Teacher Enhancement Program funds a small program of teacher institutes emphasizing teaching techniques in science and mathematics. The institute program is much smaller than it was in the past. Between 1954 and 1974 NSF spent over $500 million on teacher training institutes that at their peak involved 40,000 teachers (OTA, 1988:119-120~. The Teacher Enhancement Program has been revived somewhat since 1982, when it was virtually nonexistent. According to Charles Hudnall of the NSF staff, from 1983, when $11 million were appropriated, it has grown steadily to $43 million in 1989. There is little national information available on the extent to which inservice programs other important professional resources are used. Most of the existing data on this topic were collected from teachers, through self-reporting, in 1985-86 and reported in Weiss (1987~. The SASS local education agency questionnaire asks whether the district reimburses teachers' tuition and course fees. It also asks whether free retraining is available for teachers for shortage areas, and what those shortage areas are. The school questionnaire for the 1990 follow-up of NELS:88 asks principals (primarily of middle or junior-high schools) whether teachers are rewarded with time off for professional workshops, extra materials, choice of classes, etc. Teachers in NELS:88 are asked about the number of hours spent on noncollege inservice education. The NEA Survey of the American Public School Teacher (described in Appendix B) includes three fairly detailed items concerning inservice of various types over the past three years, including how much of the teacher's own money was spent on college credit programs. More data on policies related to inservice and other professional programs are needed from school districts. Among useful measures to obtain on inservice program use would be the number of hours of inse~vice training in mathematics, science, and related pedagogy accumulated in the last 12 months. Graduate courses should be distinguished from refresher workshops. Substantial inservice work in the form of graduate courses in one's primary field may indicate a high level of quality and professionalism or the intent to move from middle school to high school. The SASS teacher
STATISTICS RELATED TO QUALITY 127 questionnaire asks whether in the past two years the teacher took any inservice or college courses requiring 30 or more hours of classroom study, the subject field, and a choice among several purposes for this continuing education. NEA:s Survey of the American Public School Teacher, conducted every five years, also asks in considerable detail about kinds of inservice and college courses taken in various subjects. Beyond survey data, in-depth interviews with personnel officers from a sample of SASS school districts, held on a regular basis In a conference format, are recommended. These conferences could yield information and context on inservice programs and the incentives behind them that no formal data collection can achieve. In this vein, improvements in inse~vice data are called for, perhaps using the NEA Survey of the American Public School Teacher as a starting point. Other Practices That Affect Teaching Quality This section discusses information on other practices affecting teaching quality that is relatively easy to obtain (and in some cases available in a national data set, as Appendix B shows). Time allotted during the day for actual science and mathematics in- struction. Darling-Hammond and Hudson (1987a:30) note some possible indicators of time use: (1) amount of time within the school day allocated to classroom instruction, preparation, nonteaching duties (bus duty, hall duty, etc.), meetings with colleagues, conferences with parents and stu- dents and (2) amount of time outside the school day teachers spend on planning and preparation, grading classroom assignments, contacting par- ents, working with students, completing administrative paperwork, reading professional journals, and participating in other professional development activities. With regard to the former category time use during the school day the National Research Council's Committee on Indicators of Precol- lege Science and Mathematics Education found that "the amount of time given to the study of a subject is consistently correlated with student perfor- mance as measured by achievement tests . . ." (National Research Council, 1985:106~. Although it is possible to estimate instructional time through course enrollment at the secondary level, teachers at the elementary level have considerable latitude in the amount of time allocated to science and mathematics. Because of concern about the small amount of time spent in science instruction (Weiss, 1978), the committee recommended that time spent on science and mathematics instruction in elementary school be tracked on a sample basis at the national, state, and local levels (National Research Council, 1985:106-7~.
128 PRECOLLEGE SCIENCE AD ~THE~TICS TRACHEA Among current national data sets, the SASS and NELS:88 teacher questionnaires and the NSF survey of science and mathematics education appear to provide the most detailed data on time use. Class size and teaching load bear on the teacher's ability to be effective. Darling-Hammond and Hudson (1987a:30-31) note some evidence that smaller class sizes are related to higher achievement, and they believe that "the relation between class size and teacher satisfaction and commitment has been, apparently, too obvious to warrant much study" (p.30~. Such in- dicators should be monitored regularly, with particular attention to changes over time. The SASS, NELS:88, the NEA teacher questionnaires, and the NSF survey of science and mathematics education collect such data. Opportuniizes for collaboration and decision making, when encouraged, lead to more discussion of teaching, more use of new ideas, more involve- ment in solving teaching problems, and stronger commitment to teaching (Darling-Hammond and Hudson, 1987a:32~. Collaboration and participa- tion in decision making also seem to reduce absenteeism and turnover. Schools vary widely in opportunities for collaboration. In a medium-sized urban district studied by the panel, new teachers received little support beyond being handed the syllabus for the course. One new science teacher said that she was sure that help was available but she had no time for discussions; she spent every free minute setting up or taking down labs. One school included in another district case study pairs each new sci- ence teacher with an experienced teacher, and interaction is frequent. The SASS and NAEP teacher questionnaires and the NSF survey of science and mathematics education collect general, self-reported opinions by teachers on these aspects of quality. While it would be difficult to measure the effects of these practices with survey instruments, they are noted because researchers have found them to be related to teaching quality. Salaries do seem to affect individuals' decisions not to enter teaching, and low salaries influence existing teachers' propensities to take second jobs (Darling-Hammond and Hudson, 1987a:33~. But research has shown little about the effects of salaries on teacher performance. Among the current national data sources, the National Education Association provides data on starting salary of teachers and the SASS teacher questionnaire inquires about salary, including income from nonschool employment and total family income. This questionnaire also requests opinions of various pay incentives. A wide variety of data collection and research concerning school system policies and practices that affect teaching quality have been proposed in this section. We set priorities on these data needs in Chapter 6. It is critical, however, to build a foundation of data about school and district practices relating to quality and to embed the data in a context obtained
STATISTICS RELATED TO QUAL17Y 129 by interaction with school districts (such as those recommended at the end of Chapter 6~. MEASURING TEACHER QUALIFICATIONS A good information base on the quality of the teaching force would permit descriptive profiles of teachers and would allow for the measurement of change in teacher characteristics over time. Such information may be particularly important for understanding the quality adjustments that bring supply and demand into equilibrium. In the future, it may be possible to introduce quality information focusing on teachers, as well as on school and district practices, into models of teacher supply and demand. Although certification is available as the baseline measure of teacher quality, it is a most imprecise indicator. More comprehensive information would be available from a teacher's transcript showing all courses taken. A higher standard of quality still would be approval by a professional board of standards in science or mathematics. Thus, one presumes ~ higher level of teacher knowledge as more of the standards are met and, therefore, a better quality of instruction. It is clear that the number of teachers meeting the standards declines as one moves from state certification to those of professional associations, and to the qualitative rather than the quantitative dimension of the professional standards. We discuss these standards of quality in order of difficulty of attainment. Certification as the Basic Proxy for Teacher Quality When measuring quality of the supply of teachers of science and mathematics, certification is the obvious first standard. Despite differences amens states in certification rules and the level of preparation implied by the different standards, as shown in Appendix Table 5.1 (the tables in this chapter appear at the end of the chapter), certification is easily monitored. Certification does suggest some minimal level of knowledge and training. Recently, alternative certification programs have been established in 21 states, in response primarily to shortages in particular subject areas, but also in response to dissatisfaction with the quality of traditional university programs (McKibbin, 1988~. What do we know about the quality of teach- ers certificated through these programs? In a survey of these alternative programs, McKibbin concluded: "In most cases, the entry requirements were equal to or greater than the requirements for entry into university teacher education" (p. 34~. But the weaknesses of the alternative programs (which supply a very small percentage of all new hires in the states that permit them) are similar to the weaknesses of traditional programs, McK- ibbin added: "In the larger programs the training resembles the offerings
130 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS in university certification programs" (p. 35). He concluded that, although alternative programs thus far do meet specific subject-area needs, they do not seem to be superior in quality and they are not likely to replace conventional routes to certification" (p. 35~. -rig certification, whether obtained through traditional or alternative certi- fication programs, is generally a poor indicator of quality, even for teachers of science and mathematics. Requirements are often nonspecific with re- spect to required breadth and depth of subject knowledge. For instance, in some states only a minimal number of hours in mathematics is spec- ified with no level indicated (e.g., no requirement that the courses must contain work in calculus or beyond). Certification is also possible in a minor area, or as an endorsement on some other area; in these cases, even fewer hours of a given subject are required. Nonetheless, certification is the obvious first cut at a quality dimension of both the teaching force and the supply pool. Hiring uncertified teachers often means a diminution of quality. Conversely, increasing standards for certification can be expected to improve quality. Monitoring changes in the number of teachers with traditional, emergency, and alternative certificates, by subject areas, would provide useful information on quality at a baseline level. Course Preparation and Transcript Data What is known about the actual course preparation of teachers of mathematics and science? Ho studies related to the question are (1) the analysis of mathematics and science preparation in the major teacher training institutions of the southern states reported by Galambos (1985) and (2) the study of a nationally representative sample of mathematics and science teachers, which includes their course backgrounds, reported by Weiss (1987~. The research indicates that education majors tend to have less course work in mathematics and physical sciences work than do arts and science majors, although they tend to have more course work in biology and geology (Galambos, 1985~. Education majors also have completed less college-level mathematics course work than have arts and science majors. The science preparation of education majors and arts and science majors is quite similar. Both groups take about the same number of science courses and accumulate about the same amount of laboratory experience. The groups differ in the relative amount of biology: and geology versus chemistry and physics that they take: two-thirds of education majors take no chemistry or physics at all. Arts and science majors take almost twice as much chemistry and physics as teachers do. Course preparation required to earn certification to teach physical sciences or life sciences can affect teacher quality in a rather unexpected
STATISTICS RELATED TO QUAIdIY 131 way: by classifying geology and earth sciences (which are relatively de- scriptive courses) with physics and chemistry (which are quantitative), as Is commonly done, a teaching candidate can obtain certification to teach physics/chemistry/geology by taking many geology/earth sciences courses and few physics and chemistry courses. The certificant can then be as- signed to teach physics and chemistry classes with only limited knowledge of the subject matter. D~saggregation of geology from physics and chem- istry in the science-teaching certification process should result in a closer fit between teachers' course preparation and their certification-to teach certain courses. Courses typically taken by those preparing to be elementary teachers and secondary teachers are described below in more detail. Elementary Teachers According to a recent RAND report (Darling-Hammond and Hudson, 1987b:32), the typical preparation program for elementary teachers includes four science courses and two and a half mathematics courses. A large proportion of elementary teacher preparation is in social science or general areas rather than in courses specifically related to subjects taught. Weiss indicates that elementary teachers are most likely to have taken a biological science course, a physical science course, but not chemistry or physics. These teachers are likely to have content-specific methods courses in both science and mathematics.3 Mathematical content courses designed specifically for elementary teachers help explain the increase among these teachers in their confi- dence to teach mathematics. Either they feel that they are well qualified, or their perception of elementary school mathematics is limited to arith- metic computation, which they feel comfortable teaching (Weiss, 1987~. No such improvement in their confidence to teach science is noted. Ei- ther elementary teachers take too little science course work in college, or college-level science courses are not relevant for elementary teachers, or both. The importance of high-quality instruction in science and mathematics at the critical elementary level cannot be overemphasized. Science in the elementary grades can become language arts that is, vocabulary and not 3Mathematics for elementary school teachers in the Weiss data are courses developed in the 197Os and designed specifically for teachers. In many places they have solid course content that is more appropriate for elementary teachers than precalculus or introductory calculus courses. A consensus exists in the mathematics community on the Concept of these specifically designed courses, which are considered to be a significant gain in elementary teacher preparation. Since these courses differ from the below-college-level courses in the Galambos' study, the two types of offerings should be distinguished from one another.
132 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS conceptual or experiential science, and the same can be said of mathe- matics. The question concerning the preparation of elementary teachers for mathematics and science gives rise to two contrasting answers. One answer calls for a better preparation of all elementary teachers in science and mathematics. The other answer calls for specialists to teach these subjects beginning at the fourth grade. One relevant issue is the youngest age at which children can comfortably tolerate varied teaching styles during the course of a day. While further research is needed toward raising the quality of elementary science and mathematics teaching, for the short run it is important to determine, perhaps through a future edition of SASS, the educational background of teachers at the elementary level. What science or mathematics courses or majors have they pursued? Secondary Science and Mathematics Teachers In the southern states included in the Galambos study, secondary teachers complete a major in their content areas. The Galambos study indicates, however, that they take fewer courses and fewer upper-division courses in their majors than do their counterparts in arts and sciences. Mathematics majors tend to resemble each other more closely than do science majors. Secondary teachers, like elementary teachers, tend to prepare more for biology than for chemistry or physics. As indicated above, this fact may result from widespread offerings or requirements for high school biology and more restricted offerings of chemistry and physics as electives. The cause and result are not clear. Availability of biology teachers may lead to a variety of high school course offerings. The Weiss data show that secondary science teachers do tend to concentrate in life sciences. The Galambos and Weiss methodologies have limitations, and in some aspects their results are not comparable. The Weiss data on course back- ground were self-reported. Galambos collected transcripts, but one does not know whether all the teachers trained in the Galambos study actually took teaching positions. However, the Galambos and Weiss studies do indicate the importance of course background data to gain a clear picture of actual preparation programs. Data on courses taken and transcript data would seem the most con- crete measure of qualifications. Differences in course titles among institu- tions and in course content could be assessed. Transcripts could be used to examine the teachers' majors while in college or graduate school and to examine the teacher's academic preparation in terms of specific courses taken. The transcripts could also be used to identify the teachers who failed, withdrew from, or repeated required courses in mathematics or science.
STATISTICS RELATED TO QUAW7Y 133 Working from such a disaggregated data base, one could determine, when a teacher does not meet certification requirements, whether it is a case of (1) lack of subject background, (2) lack of student teaching, (3) lack of content-specific methods courses, or (4) lack of general education courses. When a person meets certification requirements, one could also determine the degree to which the four factors considered above are sat- isfied. Itanscripts are especially useful for assessing the backgrounds of recent graduates. They are also valuable for monitoring the changes taking place in teacher training programs. Teachers educated 10 years- ago, for example, are unlikely to have had a course in computer science. Itachers trained today would be more likely to have that exposure. Transcripts would document that sort of change. Transcript data thus permit measure- ment of change over time in preparation programs, and they can suggest the prevalence of the four factors mentioned above among noncertified teachers. Collecting transcript data for national purposes, however, could prove too daunting a task except for a relatively small sample. NSF has con- tracted for a transcript study for the science and mathematics teachers who responded to the teacher questionnaire in NELS:88. This appears to be an appropriate group of teachers to study since NELS:88 will also have student outcome data and data on teaching practices that can be analyzed in conjunction with the teacher background information. When teachers move from elementary to secondary positions, do they take more science or mathematics courses to strengthen their content background? The SASS teacher questionnaire contains gaps in this area of information. Respondents note the number of courses they have taken in specific disciplines, but not when they were taken. One does not know whether they were taken before or after the teacher moved from elementary to secondary teaching, and therefore whether they were taken to strengthen background. Professional Standards as a Quality Dimension A measure of higher quality of the teacher supply would be the number of teachers meeting the professional standards of mathematics and science teacher associations for preparation programs. A slightly higher standard still would be teachers who also meet the inservice education standards of these associations. These professional association standards indicate whether teachers have a subject background in science or mathematics that includes a sense of how the discipline should be taught with regard to content and student background (Richardson-Koehler, 1987~. The recommendations of professional associations of mathematics and science teachers call for more than certification for measuring the quality
134 PRECOLLEGE SCIENCE AD ~THE~TICS TRACHEA of the teaching force. The abbreviated standards issued by the National Science Teachers Association (NSTA) and the National Council of lbach- ers of Mathematics (NCTM) indicate both the quantity and the pattern of preparation (Appendix Table 5.2~. The standards cover both elementary and secondary teachers. Both the Galambos and Weiss studies suggest that teachers often meet the quantitative standards but may nevertheless fail to measure up to these criteria for quality: teachers take too many low-level courses, and they devote too little time to the quantitative physical sciences in favor of biological and descriptive sciences. For science teachers, the fre- quent mismatch between preparation and assignment leads to instructional situations in which the published professional standards are not met. For secondary mathematics teachers, these standards are more likely to be breached in terms of the pattern and quality of preparation, not in terms of the total number of courses taken in mathematics, or the fact that one majored in mathematics. A common yardstick of these and other standards Is that a secondary mathematics teacher's-preparation should encompass more than introductory calculus. More importantly, the preparation program should sample areas within mathematics and culminate in an overall sense of the discipline. Most state certification requirements for mathematics include a content- specific methods course. Some of the more recent alternative routes to cer- tification do not have this requirement. Professional preparation specific to the teaching of mathematics, including understanding of mathematics learning, is an important dimension of quality. Both the NCTM and the Mathematical Association of America (MAA) guidelines include specific recommendations for mathematics education courses. Data on the extent to which these and inservice standards of professional associations are fol- lowed should be collected and monitored over time, building on monitoring activities conducted by the professional associations themselves. Testing for Subject-Matter Knowledge Most states now require that teachers pass a competency test as a prerequisite for certification (Appendix Table 5.3~. By the fall of 1987, 45 states had enacted competency testing programs as part of the certifica- tion process. And in 31 states, rules also required that students take an examination for admission into a teacher education program. One subject of debate concerns what competency tests should cover. No nationally accepted test exists, so some states use commercially devel- oped tests, and others design their own. They cover a combination of basic skills, subject matter knowledge, and pedagogy. Appendix liable 5.3, which shows the states mandating competency testing of teachers, indicates the
STATISTICS REL-A TED TO QUALITY 135 variety of tests employed. These range from low-level tests for screening entrance to teacher education programs to exit tests from these programs or for hiring, to more sophisticated measures such as the National Teachers Examinations (NTE) area tests for teachers within a field. The NTE tests for both subject matter and content-specific method. From test results, it does not appear that the ability of prospective teachers qualified in chem- istry, physics, and general science has been declining from 1980 to 1984. However, as a measurement of instructional quality, the major limitation of the NTE test is the weak relationship of test performance to teacher performance (National Research Council, 1988:96~. Other limitations of the NTE are the fact that not all states require the test and the fact that some test takers may not have taken teaching positions (National Science Foundation, 1985:127~. The NRC Committee on Indicators of Precollege Science and Mathe- matics Education recommended, as a key indicator of quality, "that samples of current teachers be selected to take tests that probe the same content and skills that their students are expected to master" (NRC, 1988:9-10~. More specifically, the committee recommended that tests be given every four years to a sample of all teachers and every two years to a sample of newly hired secondary school science and mathematics teachers. The Holmes and Carnegie Recommended Standards Within the past three years, two groups of education experts have proposed more sophisticated measures of the quality of teachers' profes- sional preparation. These groups call for placing greater requirements on teachers in the preservice stages ensuring higher quality through more rigorous preparation, certification and selection and ultimately for more professional autonomy once in the classroom. The two groups are (1) the Carnegie Disk Force on Teaching as a Profession, which has given a grant to Stanford University to develop measures of teacher quality that may be used by its proposed National Board for Professional Teaching Standards (Carnegie Task Force on Teaching as a Profession, 1986) and (2) the Holmes Group, composed of 96 deans of education from universities nationwide, which aims to develop higher standards for teacher education at their institutions (Holmes Group, 1986~. The Carnegie group has proposed a three-stage voluntary assessment process covering subject matter mastery, education courses taken, and actual teaching performance, all under the aegis of a National Board for Professional Caching Standards. Researchers at Stanford have classroom- tested measures of teacher quality for elementary school mathematics and high school history teachers. Both of these classroom-based studies of
136 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS measures of quality seek to incorporate an evaluation of actual teaching effectiveness with regard to subject matter, content-specific teaching, and student characteristics. The development of these procedures could provide better measures of quality than those now available. Both the Carnegie and Holmes recommendations are similar in call- ing for completion of a subject-matter major before initiation of training for teaching. Both permit an introduction to education, but only at the undergraduate level. Both are similar In calling for a restructuring of the teacher corps. (Appendix Able 5.4 summarizes their major recommenda- tions.) The Holmes Group categorizes teachers as "career professionals," "professional teachers," and "instructors." The Carnegie Forum distin- guishes between "licensure" and `'certification." Licensure would be what is now called state certification. Beyond that, Carnegie's proposed National Board for Professional Caching Standards would give board "certification." The assessment technique planned for board certification would go beyond knowledge and preparation of teachers to an assessment of their mastery of teaching techniques in the classroom. The ultimate measure of the quality of the teaching force would be the number of teachers that were board certified or that were categorized as "career professionals" under the Holmes definition. The purpose of these measures Is to enhance the general profession- alism of the field and thereby attract and retain higher-quality personnel. As these approaches are refined and implemented more widely, more so- phisticated measurements or data on aspects of teacher quality should emerge. The changes called for by these groups, to the extent that they are adopted, will revamp many existing practices and give rise to new questions some of which center on the supply response to changes in quality require- ments. If teachers are required to study for five years, what adjustments are required in compensation to attract teachers? If teachers are required first to have a subject-matter major, will fewer or more persons continue on for a teaching degree? Will states and localities fund the change? Will the changes be willingly embraced or grudgingly made by prospective teachers, school administrators, school boards, and taxpayers? The Presidential Awards for Science and Mathematics Teachers The Presidential Award for Excellence program for recognition of ex- cellence in teaching, sponsored by the National Science Foundation, was begun in 1983. Each year an award is given to one science teacher and one mathematics teacher in each state. Actually 54 jurisdictions are covered,
STATISTICS RENA TED TO QUALITY 137 consisting of the 50 states, the District of Columbia, Puerto Rico, the U.S. Must Territories, and the Department of Defense dependency schools. Eligibility is restricted to teachers in junior high, middle, or high schools with a minimum of five years of experience who devote at least half-time to classroom teaching. An examination of descriptions of winners reveals a profile of highly visible teachers. Many have published articles in professional journals, and all are involved heavily in after-school curricular activities, such as workshops and student projects. A few have higher degrees in their fields. Each winner receives a monetary award of $5,000 given to the school. The candidate indicates how the money should be used. The range of uses includes travel expenses to attend courses, stipends for outside speakers, computer hardware and software, and science equipment. In a number of instances the money has been designated for materials that one would think the school budget should normally provide. There has been no follow-up on how the money is actually spent. We recommend that there be a follow-up study at schools of previous winners to determine "quality" effects of the awards. In how many schools were the monies used for basic materials that would normally belong in the school or district budget? 1b what extent could the findings from follow-up analysis of the awards and recipients yield information about quality of instruction and the qualifications of the teachers? The award winners constitute an interesting group for research. As an example of possible research, one study of 34 winners of the 1983 Presidential Award for Excellence in Teaching Mathematics was conducted by Yamashita (1987) to compare their level of professional development with that of a comparison group who were members of NCTM. A list of 21 professional development activities was given to all the participants to rate for importance to their own professional development. Awardees rated the most important activities as attending conferences and institutes, reading and writing for journals, developing curriculum beyond that for their immediate courses, advising student math activities, and teaching inseIvice courses. The comparison group rated writing for publication and consulting as of primary importance; the other activities mentioned above were rated less important to them than to the awardees. Awardees participated in more activities than did the comparison group. Yamashita concluded: "It may well be that the most distinguishing difference between the awardees and the comparison teachers in this study is the number of activities in which they engage and the higher energy level manifested therein" (p. 66~.
138 PRECOLLEGE SCIENCE AND MATHEM'4 TICS TEACHERS TEACHER QUALIFICATIONS AND STUDENT OUTCOMES Evidence The literature to date does not indicate strong relationships of measur- able teacher qualifications and such educational outcomes such as student performance on standardized tests. In a meta-analysis of 65 studies that had sought relationships between science teachers' characteristics and teaching effectiveness or student outcomes, Druva and Anderson (1983) find gener- ally weak correlations, and many of these correlations were based on only one study. However, certain positive correlations are identified, on the basis of more than one study, that warrant statements of results: teaching effectiveness is positively correlated with the number of education courses taken, the student teaching grade, and length of teaching experience. Stu- dent outcomes are positively correlated with teachers' science training and general educational preparation. And this correlation between teachers' science training and cognitive student outcome is progressively higher in higher-level science courses. From studies summarized in a comprehensive literature review by Darling-Hammond and Hudson (1986:24-32), it appears that certain teacher characteristics exhibit some positive relationship (often weak) to student performance: verbal ability; number of mathematics credits (for mathemat- ics teachers); educational background in science, particularly for science teachers in higher grades; recent continuing educational experience; in- volvement in professional organizations; years of teaching experience; and positive attitudes toward teaching, flexibility, and enthusiasm. Other mea- sures, such as IQ, National Teacher Examination (NTE) scores, and various measures of subject knowledge, have not shown any relationships to out- comes. In a review of the literature, Blank and Wizen (1986) note that the failure of any research to establish a strong relation between teacher characteristics and student outcomes may be explained by a number of problems with the research to date on teacher effectiveness: · The degree of variation in the independent variable, e.g., NTE scores, is often so small that no effect on outcomes would be measurable. · Many studies have not included teachers with emergency certificates or low levels of training in the field in which they were teaching, so that, again, one would not expect to find strong relationships of such measures as extent of subject preparation and outcomes. · Many studies have used student achievement tests as the sole measure of outcomes. The tests themselves may not relate to the goals of the students' courses; moreover, other measures such as attitudes toward science or math might show different results.
STATISTICS RELATED TO QUALITY 139 In sum, the presumption of a relationship between higher teacher qualifi- cations and improved instruction is in need of testing and should not be discarded. Implications for Data and Research The panel reaffirms its earlier recommendations that relate to mea- suring teacher qualifications and their relationship to student outcomes (National Research Council, 1987c:8~. Both the case studies and the meet- ing with large district personnel officers confirmed the usefulness of these recommended data collection efforts. As stated in the interim report: 1. We recommend that the National Center for Education Statistics surveys of teachers regularly include: · Measures of general intellectual ability and of academic prepa- ration to teach mathematics and science fields, particularly for new entrants, in order to provide time series for monitoring and analysis. These measures should be obtained to the extent possible from transcript records rather than through survey questions. · For experienced teachers, measures of recent inservice prepa- ration and participation in professional activities in mathematics and sci- ence fields. These surveys should also obtain measures of years of teaching mathematics and science distinct from total teaching experience. · Measures of certification (type and subject fields). We also recommend that the NCES obtain and disseminate available information on state certification policies and practices; we note that NCES has since published such information (NCES, 1988b, p. 123~. 2. We recommend that further research be conducted on the re- lationship of measurable characteristics of teachers of mathematics and science to educational outcomes of students in these fields. In order to permit comprehensive and methodologically appropriate research on this issue, the National Educational Longitudinal Study of 1988 should in- clude appropriate measures of student outcomes together with a rich set of teacher characteristics and characteristics of schools and districts. (We note that NCES includes such data items in NELS-88.) Research relating teacher qualifications and student outcomes may be pursued using student and teacher questionnaire data from the 1985-86 NAEP assessment for science and mathematics as a starting point. NELS:88 is another useful source of information, especially for longitudinal research. Of all the national data sets highlighted in Appendix B. the Schools and Staffing Survey asks for the greatest level of detail regarding teachers' qualifications. SASS asks teachers in detail about their past teaching expe- rience, breaks in service, and previous occupation. It asks for the teacher's
140 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS major and minor at every postsecondary level completed, the year each degree was completed, and the name of the undergraduate college. The respondent further notes in what field he or she teaches the most classes and the second-most classes. The respondent then is to provide the num- ber of courses taken in these fields. For teachers who teach any science or mathematics in grades 7-12, there are data on number of courses in science and mathematics areas. SASS further asks detailed certification questions regarding fielders) and type (full, probationary, or emergency) for each field. The teacher then describes the amount-and purpose of inservice or col- lege courses taken in the last two years. Whether these courses were taken in the teacher's primary assignment field is also discernible through the questionnaire. Thus, with its ability to single out science and mathematics, the SASS teacher questionnaire will advance the level of statistical infor- mation on teacher qualifications far beyond the mere presence or absence of certification. From the district's perspective, SASS asks district respondents which screening devices they use~r require for hiring: full state certification; emergency certification; graduation from an approved teacher education program; college major or minor in the field to be taught; passing of a district test; passing of a state test of basic skills; passing of a state test of subject knowledge; passing of the NTE. Screening devices used by districts, which constitute standards of qual- ification, and their changes over time should be monitored through SASS and by continued collection and dissemination of certification data from states. However, the collection and use of more statistics related to teacher quality must be tempered as their limitations are recognized. There are some statistics not included in SASS, such as NTE scores, grade point averages, and other information that transcripts would provide. The NSF-sponsored teacher transcript study being carried out in conjunc- tion with NELS:88 will provide the opportunity to explore the potential of transcripts as measures of academic background. In addition to transcript data, monitoring changes in admission standards for teacher education programs, by publishing data collected by the American Association of Colleges for Teacher Education (AACTE), is also recommended. Finally, research is needed on the supply response to changes in certi- fication requirements. By itself, more stringent qualification requirements will tend to reduce the supply of new teachers unless it is offset by salary prospects, greater prestige, or better working conditions. Absent any of these offsets, tougher qualification requirements are likely to shift supply between school districts and states, not produce a more qualified supply pool.
STATISTICS RELATED TO QUAI=Y 141 OTHER SCHOOL AND HOME FACTORS THAT AFFECT OUTCOMES Much of the impetus for concern over the quality of precollege science and mathematics teachers arises from the widespread evidence that U.S. student outcomes test scores and general level of literacy in science and mathematics are poor. The possibility must be raised, however, that the problem underlying these low outcomes does not lie solely with the quality of teaching or the qualifications of teachers. Educational outcomes are a complex product of student and family inputs, teaching inputs, and educational curricula. Poor outcomes can be due to factors entirely beyond the quality of the teacher corps. This section addresses some of the most important of these factors. Curriculum Structure The influence of curriculum structure on U.S. students' mathematics test scores is under debate. It is argued that the consequences of a layered curriculum- through which students are introduced to relatively little new material each year through grade 8, and much of the mathematics training in any given year is thus basically review- are boredom and lack of mastery of the key ideas involved in the development of mathematical skills. A related criticism is that mathematics textbook producers, in trying to market their product to as many school systems as possible, end up with a light treatment of many topics rather than intensive treatment of a few topics. Since the basic text is the primary resource used by most precollege mathematics teachers (Weiss, 1987:31, 39), and since the text usually favors breadth and memorization of facts over depth (Office of Technology Assessment, 1988:30-34), the result is that students master few if any of the key concepts. Quality of Textbooks Although most science and mathematics teachers surveyed by Weiss in 1985 seemed to indicate that poor quality of textbooks was not a seri- ous problem (Weiss, l9S7:40-42), many scientists and educators who have reviewed the textbooks criticize their quality and their extensive use in classrooms (Office of Technology Assessment, 1988:30-33~. The Mathemat- ical Sciences Education Board (MSEB) of the National Research Council, as part of an ongoing effort to identify the key elements needed for reform, has determined that, between the second and eighth grades, there is only one year in which more than half the material is new (National Research Council, 1987b). This suggests that the solution to improving the quality of student skills in mathematics does not rest solely with providing better trained teachers, or even with providing more time for the teaching of
142 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS science and mathematics, but rather depends on more fundamental reform of how the mathematics curriculum is organized. In a decentralized school system such as we have in the United States, fundamental reform of this nature is difficult to achieve. Moreover, there are differences of viewpoint among educators as to the validity of this line of criticism. Classroom Time Used for Science and Mathematics The amount of classroom time devoted to science and mathematics is another area of dispute, especially at the elementary level. Elementary school teachers are said to spend relatively little classroom time on science and mathematics topics. Recent studies have compared the amounts of classroom time spent by students on different topics; the evidence is mixed. To begin with, there appears to be a substantial difference in the instructional time allocated to reading and mathematics in the early grades in the United States. One study indicates that about twice as much time is allocated to reading as to mathematics in the fourth grade (Cawelti and Adkisson, 1985~. Weiss (1987:13) also found substantially more time devoted to reading than to mathematics, though not twice as much time. At the grade 4-6 level, teachers in this survey reported spending 63 minutes per day on reading and 52 minutes on mathematics. At the K-3 level, however, reading took up 77 minutes and mathematics 43 minutes, a wider difference in the earlier years of schooling. Forthcoming data from SASS will provide more recent information on classroom time; the SASS teacher questionnaire asks elementary school teachers in self-contained classes for hours per week spent in each of the core subjects, including science and mathematics. Other studies based on careful observation of actual classroom time spent on mathematics in three cities (one each in the United States, Japan, and Taiwan) have found very large differences between the students in the U.S. city and those in the Taiwanese or Japanese city (Stevenson et al., 1986~: U.S. fifth-grade children spent 3.4 hours per week on mathematics, lkiwanese students 11.7 hours per week, and Japanese students 7.8 hours per week. In grade 1, the differences were similar-2.7 hours for U.S. children, 4.0 hours for 1hiwanese children, and 5.8 hours for Japanese students. In addition, U.S. students were less likely to be attending to the teachers than either ~iwanese or Japanese children, largely because individual work is much more common in U.S. classrooms than in Asian classrooms. However, for eighth grade, another study of classroom hours (McKnight et al., 1987) reports that U.S. students in grade 8 spend more time on mathematics instruction than students from Japan or Hong Kong. Comparability problems limit attempts to draw conclusions from these studies. ~ start with, the studies are of students in different grades. In
STATISTICS ROLL TED TO QUALITY 143 addition, the study by Stevenson and his colleagues contains very accurate measurement of classroom hours but covers a very small and possibly unrepresentative sample of schools; the McKnight et al. study is based on a national probability sample of schools but suffered from a high nonresponse rate and used officially scheduled hours and similar data to estimate time spent. Given the known difficulties of getting accurate estimates of time spent on various activities from very generalized methods (How much time is scheduled? How much is spent on average?), the panel is inclined to believe that the data of Stevenson and his colleagues are probably -closer to the truth, and that one source of the difference in mathematics achievement is the gap in time allocated within the classroom. If students' low skills and test scores in science and mathematics were known conclusively to be due simply to the relative amounts of time spent on these subjects, the solution would be relatively simple provided school systems can be encouraged or induced to change the structure of their curricula. But if time spent or curriculum structure are the basic problems, then the issue is not one of teacher or teaching quality, but simply one of relative emphasis within the curriculum. It may be true, of course, that many U.S. elementary school teach- ers are less comfortable teaching science and mathematics than teaching language arts, and therefore spend less time on science and mathematics. This may be explained by observing that in the United States elementary teachers tend not to be subject specialists, whereas the employment of specialist teachers of mathematics is more common in Japan, China, and Taiwan in the early grades. Other Instructional Factors Other possibly important differences between U.S. and Asian science and mathematics instruction have been identified in the ongoing studies being conducted by Stevenson and his colleagues. There are documented differences in the nature of textbooks-American texts explicate mathemat- ics problems much more extensively and lead the students very carefully through exercises and problems; Asian texts are much shorter (about half the length in some of the texts examined) and make much stronger de- mands on the students to find their own way through the problem. There are also documented differences between American and Asian teachers of mathematics in the number of actual teaching hours per day and the amount of time available for planning and preparation; American teach- ers have much less nonteaching time scheduled during the day than their Asian counterparts. And there are documented differences in the degree to which teachers are autonomous in their own classrooms. In American classrooms, it is not uncommon that teachers are basically on their own
144 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS after the first year, while in Asian classrooms younger teachers are typically under the tutelage of a senior teacher for a number of years (Stevenson, 1987; Lee et al., 1987; Stevenson et al., 1988; Stigler et al., 1987; Stevenson and Bartsch, in press). There also appear to be substantial differences in the training given to U.S. and Japanese mathematics teachers, with U.S. teachers spending more time learning mathematical content, and Japanese teachers spending more time learning mathematics pedagogy (teaching of mathematics) (McKnight et al., 1987:65~. Home Environment The home environment also has a significant impact on young chil- dren's learning. The home environment of many children in the United States is not conducive to concentrated thought and learning. The pro- portion of single-parent households and the proportion of households in which both spouses work are much higher now than in past decades. These realities can create problems for children and can have a potentially serious influence on their skill development. 1b understand educational outcomes requires us to understand the contributing effects of these home environment factors. Parental attitudes, as well as demographic differences in home environ- ments, can also influence children's ambition to concentrate on academic learning. The best documented evidence of differences in attitudes and expectations comes from a comparison of American and Asian households. In general, Asian mothers are less satisfied with the school performance of their children than American mothers (despite the fact that their children are generally doing better); they are more likely to attribute success in school to hard work rather than to native ability; and they are less likely to be satisfied with the way the schools are performing than their American counterparts (Lee et al., 1987~. Poor student outcomes are thus not uniquely correlated with inade- quate quantity or quality of teachers, but could easily be due to factors that are largely unrelated to teacher quality. One cannot conclude that poor science and mathematics outcomes on the part of students necessarily reflect inadequacies in the background or ability of their teachers and to try to remedy the problem only by enhancing the numbers or the quality of precollege science and mathematics teachers. Factors such as the structure of the curriculum, the practices of both K-12 school systems and teacher training institutions, the amount of time spent on science and mathematics topics in schools, and the influence of home environments on development outcomes all need to be understood before we can fully understand the problem or devise appropriate remedies. Thus, although the issues raised in this section are beyond the scope of this study, they serve to point up
STATISTICS RELATED TO QUALITY 145 the many other factors beyond teaching and teacher quality that bear upon student outcomes. SUMMARY In the near term, it is through quality adjustments that the supply and demand for precollege mathematics and science teachers reach equilibrium. The quality of instruction is therefore a central focus of our study. Statistics that can furnish indications of quality and trends or events that can be monitored to illuminate quality and changes in quality over time are called for. Areas of data concern relating to quality not only focus on teacher characteristics but also extend to contextual arrangements that affect over- all teaching quality. These contextual variables include teacher policies and practices regarding assignments and teacher background, course offerings and enrollments, recruitment practices, school and school system policies governing both initial placement and transfers, and inservice training pro- vided by schools, school systems, states, professional associations, and the federal government. Also of concern is the distribution of qualified teachers across districts when classified by enrollment size; by racial/ethnic charac- teristics of its students; by geographic characteristics of urban, suburban, rural; and by socioeconomic status. We have identified some district policies and practices that influence teaching quality, and note in particular the importance of information on recruitment practices, seniority- rules, potential for teacher advancement, teacher assignment and misassignment, and continuing professional devel- opment, as well as external factors, primarily state mandates and policies, that affect the quality of the supply pool. There are numerous ways to measure and assess teacher qualifications that influence overall teaching quality. Some are objective and can be counted; some are subjective and not easily quantified. Some are easily quantified but of little use (such as certification); some would be highly useful but would require more examination (such as transcripts). Some indicators are based on existing standards (such as those of the National Science Teachers Association), and some on proposed standards (such as those of the Holmes and Carnegie groups). While it is acknowledged that a thorough knowledge of content is only a necessary and not a sufficient set of characteristics for a successful teacher, certain qualifications are necessary. And data can be collected to indicate the presence and strength of these qualifications. We do recognize, however, the considerable amount of effort and resources that would have to be invested in collecting these data, when such factors as presence of certification, transcript data, and educational background have not yet been demonstrated to be strongly associated with
146 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS teacher quality or student outcome. Thus, it is important that the National Science Foundation fund a program of controlled experiments on factors that do measure teacher or teaching quality. Such research would include identifying the relationship between measurable teacher qualifications and student outcomes. If the Carnegie or Holmes recommendations for higher professional standards are adopted, the consequent changes in the teaching force should be monitored, together with any changes in supply as a result of the more rigorous requirements. Other factors beyond teacher quality-such as textbook use, time commitments, the structure of science and mathematics curricula, and home environment were noted as influences on teaching quality and student outcomes. These factors complicate any attempts to link outcomes with particular teacher qualifications. In conclusion, to understand the crucial role of quality in bringing supply and demand for precollege science and mathematics teachers into equilibrium in the short term, we have acknowledged some rather daunting data needs and research issues. We realize that these needs might not be able to be met completely enough to introduce teacher quality measures into teacher supply models in the near future. But successful collection of more precise data, particularly through SASS and existing state information files, can be expected to contribute to an understanding of teacher quality, and additional research may help identify the characteristics of teachers and teaching that are determinants of student outcomes.
- CQ ,n - He - o o v a] o v a] so 4~ · - o · - v · - it v L. v it v q) · - a] cs ce · - ~c cS c; u, m z ~ .= ~ ~ ~ .~ ._ L ,1: V ~ ~ ~ X U] ~ ~ a~ ._ _ ~ - E (V .- U CO _ ._ [U ~ ~ cn ._ o ._ ._ v :^ a~ ._ v a~ v .o m v .= v n5 ·- Ll - ~ m ~ a~ V ._ ~ C~ ~ ~fi ~ ~ o z ~o ~i,, o o 4, o o ~ o Z ~ Z ~ Z - _ CQ ~ _ OO ~0 _ ~ O ~ {D O O e~ Z co c ~e~ c ~eo e ~, t_° O ~00 ~0 0 0 o: e~ ~=1 ~ co co e~ Ut ~. r ~° 0 ~ ut Oo ~0 0 0 c~ Z co c~ ~=1 ~ eo co e~ L. C. Z c°o , I "Q1 , , co , t_ ° O -4 e~ Z ~ ~ ~1 ~ co c~ e ~CD r~ .. ~-c =~ ~ ° o ~ ~ ~ ~ = ~ E ¢ ¢ ¢ ~ v v v 147
:~ a - a' L. in - o us En o v, Al · - o · · v In · - c) I] to · - In ~4 ~ o v l m ~ ¢ =} ~ be ·= ~ X U] ~ ~ Do ._ .~ ~ C) ~ V C. al ~ 'D ·o ~ CQ - ._ o ._ v ._ Ld v 00 ._ L. v L' o v - V] O ~ ~ ~ ._ ~ ~ ~ m v .- ~ - .o O V cn m ~ ~ 00 ._ C~ C~ L. 3 ~ ~l o ~ ~Q1 ~ ~o ~ ~ =1 cs, cn co c9 co ~ ~(D w w w o ~ ~ o o Ql Z ~ ~1 ~ Z - V) oo oo W W ~ ~ ~ O ~ ~ ~ ~ Z ~1 ~ CD ~e ~CD ~e~ rQI I I a, e~ QlI cO I o' a, I o1 `O O ~ ~O O ~ C9 o e~ Ql c~ c ~o~ c~ ~Ico ~I e~ cO cO ~I cO O ~CO ~O O ~ CD O e~ ~21 c~ e ~co e~ Qlco c ~I e~ a) a) oI cO u~ e ~oo oo co CD ~ 1 ~ eo 1 e~ =1 ~ 1 _~ CO CO CO L~1 1 O . ~CO ~O O oo ~ ~ O ~1 c~ e ~ e~ =1 co c ~_~ e~ co o~ o1 e~ o, ~ ~, ~._ - o oC ~ ~ 5 ~.= ~ ~ ~ ~.= O ~ ~ ~ ~ ~ ~ 148
~ I ~ ~u: go ° at oO~:21 al CO ~ ~- - - ~ ~ ~ O 0 ~ O ~ ~ O ~ ~ ~ O ~ ~ ~ ~ ~=1 Z ~ Z ~ ~ Z ~1 ~ ~ ~ ~ ~ Ill 2 m° tD =1 ~ ~ tD of 1 m° ~ ~ =1 I,' m° °° °° 'I 0 ~ o ~ ~ o o z go o ~ 1 o1 -~ of Z L. o O ~ co to ~ to Q1 to 0 us en ~to ~ ~ e~ 40 e~ _ ~21 ·~ ~ 4O Vl CO ~ ~ =1 ~ ~ ~ ~Vl -~ O~ Z 0 0 ~ ~O ~ ~O O u, 0 0 e~ ~ ~oO c~ co e~ _ ~Q1 co e~ cO Vl 01 0~ ~P ~ ^1 cO CO ~ ~C~ O1 - 0 0 ~o cs) 0 ~O O ~0 0 0_1 00 00 a, ~ 03 CD ~ CO =1 co e~ ~O1 O1 C9 ~ --Q1 cv o ~oOe~ ~ Vl - ~Z L L O O O CD o ~ ~O O ~O O 00 CS ~0O tr~ ~O ¢O _4 =1 tO e~ e ~Ol Vl CO ~ e~ =1 tO CO _~ {O e~ O1 ~l C~ _ o ~ > 3 ~ 3 ~ o D o ~ ~ ~ = 0 C ~ ~ ~8 ~ 8~ ~ ~ ~ ~ z ~z ~ z ~ z z o o o ~ ~ ~ ~ ~ ~ ~ ~ ~ 149
al Em - o o Id o c) u] s~ · - o . · v s" e · - c) u] ~n . - - .- C53 m ¢ ~ P4 ¢ V] :> ._ ._ · ~ ~ ~ ._ b4 · ·= . ~ ~C ~U .O ~ C) q) C ~ ._ ~ U) "C L4 _ ._ .w ~ ~ U) - ._ o ._ C. C) ._ v J:2 00 ._ v 0G o v 0t ~ -~m O ._ ~ ~ ~C m v .- ~ .O O ~ ~ 4, cn m ~ ~ 03 ._ CC ~ ~2 S~ o U) -~ ~1 u, _ O ~ O Z; CJ1 ~ Z; ~ e~ I o1 co _ e~ Q1 co Vl a, ~P 0 u' Vl u: CO e~ Vl CO e~ ._ bO C bO .= ~ 3 3 ~ 3 150 . ~ ~.k o ,~ ._ ~V C. ~C. E ceE E 11 ~ ° ~o~ _ o ~ ~ "C ~ V . ~ ~ o C) ~o l.4 ·~V Ll Q. o~ ~ b~ O 11 .=~ ~ , C) ~C ~o ~_ ·o o ~ V ~o ·- . ~ ,- ta b ,~ o C C C ~11 _ , 4, o~Ce ~V ~ V ~ o - ~ E ~ ° °° ~, =: ._ ~ o ~ _ _ ~ E ~ y c E ~ ~E ~ > 0 c a c C .E c c, ~ `<, ,, E u c~ o, [= O . ~O ~I Q'1 o1 0
STATISTICS RELATED TO QUALITY APPENDIX TABLE 5.2 Guidelines for Mathematics and Science Teacher Qualifications Specified by the National Council of Teachers of Mathematics (NCTM) and the National Science Teachers Association (NSTA) 151 NCTM Guidelines Early elementary school The following 3, each of which presumes a prerequisite of 2 years of high school algebra and 1 year of geometry: 1. number systems 2. informal geometry 3. mathematics teaching methods Upper elementary and middle school The following 4 courses, each of which presumes a prerequisite of 2 years of high school algebra and 1 year of geometry: 1. number systems informal geometry 3. topics in mathematics (including real number systems, probability and statistics, coordinate geometry, and number theory) 4. mathematics methods Junior high school The following 7 courses, each with a prerequisite of 3 to 4 years of high school mathematics, beginning with algebra and including trigonometry: 1. calculus 2. geometry 3. computer science 4. abstract algebra 5. mathematics applications 6. probability and statistics 7. mathematics methods NSTA Standards Elementary level 1. Minimum 12 semester hours in laboratory- or field-oriented science including courses in biological, physical, and earth sciences. These courses should provide science content that is applicable to elementary classrooms. 2. Minimum of 1 course in elementary science methods (approximately 3 semester hours) to be taken after completion of content courses. 3. Field experience in teaching science to elementary students. Middle/junior high school level 1. Minimum 36 semester hours of science instruction with at least 9 hours in each of biological or earth science, physical science, and earth/space science. Remaining 9 hours should be science electives. 2. Minimum of 9 semester hours in support areas of mathematics and computer science. 3. A science methods course designed for the middle school level. 4. Observation and field experience with early adolescent science classes. Secondary level General standards for all science specialization areas: 1. Minimum 50 semester hours of course work in 1 or more sciences, plus study in related fields of mathematics, statistics, and computer applications. 2. Three- to 5-semester-hour course in science methods and curriculum. 3. Field experiences in secondary science classrooms at more than 1 grade level or more than 1 science area.
152 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS (Appendix Table 5.2, continued) NCTM Guidelines NSTA Standards Senior high school The following 13 courses, which constitute an under- graduate major in mathematics, each presume a prerequisite of 3 to 4 years of high school mathematics, beginning with algebra and including trigomometry: 1-3. 3 semesters of calculus 4. computer science 5-6. linear and abstract algebra 7. geometry 8. probability and statistics 9-12. 1 course each in: mathematics methods, mathematics applications, selected topics, and the history of mathematics 13. at least 1 additional mathematics elective course Specialized standards Specialized standards 1. Biology: minimum 32 semester hours of biology plus 16 semester hours in other sciences. 2. Chemistry: minimum 32 semester hours of chemistry plus 16 semester hours in other sciences. Earth/space science: minimum 32 semester hours of earth/space science, specializing in one area (astronomy, geology, meteorology, or oceanography), plus 16 semester hours in other sciences. 4. General science: 8 semester hours each in biology, chemistry, physics, earth/ space science, and applications of science in society. Twelve hours in any 1 area, plus mathematics to at least the precalculus level. Physical science: 24 semester hours in chemistry, physics, and applications to society, plus 24 semester hours in earth/space science; also an introductory biology course. Physics: 32 semester hours in physics, plus 16 in other sciences. 5. 6. Source: Office of Technology Assessment (1988:64) .
STATISTICS RELATED TO QUAL17-Y 153 APPENDIX TABLE 5.3 States That Have Enacted Testing Programs for Initially Certifying Teachers: Fall 1987 State Enacted Effective Test Useda Alabama 1980 1981 State Arizona 1980 1980 State Arkansas 1979 1983 NTE California 1981 1982 CBEST Colorado 1981 1983 CAT Connecticut 1982 1985 State Delaware 1982 1983 PPST Florida 1978 1980 State Georgia 1975 1980 State Hawaii 1986 1986 NTE Idaho 1987 1988 NTE Illinois 1985 1988 State Indiana 1984 1985 NTE Kansas 1984 1986 NTE and PPST Kentucky 1984 1985 NTE Louisiana 1977 1978 NTE Maine 1984 1988 NTE Maryland 1986 1986 NTE Massachusetts 1985 b b Michigan 1986 1991 b Minnesota 1986 1988 PPST Mississippi 1975 1977 NTE Missouri 1985 1988 b Montana 1985 1986 NTE Nebraska 1984 1989 b Nevada 1984 1989 PPST and State New Hampshire 1984 1985 PPST and NTE New Jersey 1984 1985 NTE New Mexico 1981 1983 NTE New York 1980 1984 NTE North Carolina 1964 1964 NTE North Dakota 1986 b b Ohio 1986 1987 NTE Oklahoma 1980 1982 State Oregon 1984 1985 CBEST Pennsylvania 1985 1987 State Rhode Island 1985 1986 NTE South Carolina 1979 1982 NTE and State South Dakota 1985 1986 NTE Tennessee 1980 1981 NTE
154 PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS APPENDIX TABLE 5.3 Continued State Enacted Effective Test Useda Texas 1981 1986 State Virginia 1979 1980 NTE Washington 1984 b West Virginia 1982 1985 State Wisconsin 1986 1990 b a Tests: CAT = California Achievement Test; CBEST = California Basic Skills Test; NTE = National Teacher Examination; PPST = Pre-Professional Skills Test; b State = State-developed test. -To be determined. Source: National Center for Education Statistics (1988f:249-250~.
STATISTICS RELATED TO QUALITY APPENDIX TABLE 5.4 Comparison of Recommendations of Carnegie and Holmes Reports Pertaining to Preservice Education of Teachers Category of Recommenda- tion Fifth Year of Study Carnegie Report a Require bachelors degree in the arts and sciences as prerequisite of professional study of teaching. Require a master's degree for all teachers. Curriculum Develop new professional Revision curriculum in graduate schools of education leading to Master in Teaching degree based on systematic knowledge of teaching and including internships and residencies in schools. Coordination Connect institutions of higher education with schools through the development of professional development schools. Certification Create a national board for professional teaching standards to establish high standards for what teachers need to know and to be able to do, and to certify teachers who meet that standard. 155 Holmes Group b Make education of teachers more solid intellectually by pursuing an undergraduate major in an academic subject other than education, receive their professional training in a fifth year master's degree program, and complete a year-long supervised internship. Revise undergraduate curriculum in arts and sciences. Organize academic course requirements, including involvement of other departments in institutions of higher education. Need advanced studies inpedagogy (focus on human cognition, teaching and learning, and teaching), teachers' learning, assessment of professional performance, and evaluation of instruction. Need coherent program in schools and institutions of higher educa- tion that will support advanced study. Create professional development schools, similar to teaching hospitals, in which prospective teachers would receive their clinical training. Create 3-tier systems of teacher licensing: 0 Instructor--has BA degree, without year of supervised practice and study in pedagogy and human learning; has passed exams (see evaluation) Professional teacher--has MA in teaching; completed year of supervised practice; passed exams 0 Career professional--has completed all of the above plus further specialized study a
156 APPENDIX TABLE 5.4, continued PRECOLLEGE SCIENCE AND MATHEMATICS TEACHERS Category of Recommenda- tion Carnegie Report a Holmes Group b Evaluation/ Assessment Differential Restructure teaching force Staffing and introduce new category of lead teachers with proven ability to provide active leadership in redesign of schools and in helping colleagues to uphold high standards of learning and teaching. Use multiple evaluations o Test basic mastery of writing and speaking o Demonstrate mastery of subject, skill in lesson planning, and instructional delivery prior to clinical internship 0 Evaluate variety of teaching styles during internship-- including own--and present analytic evidence as part of professional portfolio for advancement Recognize differences in teacher's knowledge, skill, and commitment in their education, certification, and work. a Carnegie Task Force on Teaching as a Profession (1986) A Nation Prepared: Teachers for the 21st Century,. Washington, D.C.: Carnegie Forum on Education and the Economy. Pp. 55-56. b The Holmes Group (1986) Tomorrow's Teachers: A Group. East Lansing: The Holmes Group, Inc. Pp. 65-66. Report of the Holmes Source: Regional Laboratory for Educational Improvement of the Northeast and Islands (1987:15-17~.
