Learnings From a Longitudinal Study of New Jersey Alternate Route and College-Prepared Elementary, Secondary English, and Secondary Math Teachers

During the past 30 years, the challenge of staffing public schools with highly qualified teachers to meet the needs of an increasingly diverse student population has intensified. Concerns about quantity and quality of teachers have led all states to develop alternate paths for people who did not choose to become teachers during their undergraduate years.

Documenting the first state effort, this study looked at an early cohort of 315 English, math, and elementary teachers prepared through New Jersey alternate route (AR) and college-based (CB) programs. Starting to teach in fall l987, the cohort was studied for three years, with follow-up surveys in the fourth and sixth years. A smaller group of 25 AR and CB exemplar teachers was followed for 11 years. As one of the few longitudinal and comparative studies, it contributes to the growing literature on different pathways to teaching.

Here we discuss our findings and the current literature on alternate routes, noting some emerging common themes and some unique contributions this study makes related to its context and availability of longitudinal and comparative data. This discussion is organized around recruitment, preparation, placement, and retention of teachers. The chapter ends with implications of what has been learned and still needs to be learned about different approaches in the face of the continued need for highly qualified teachers and in light of the contrasting policy agendas surrounding teacher education.

DISCUSSION OF FINDINGS

Reviews of research on AR programs are inconclusive (Cochran-Smith & Villegas, 2016; Cochran-Smith & Zeichner, 2005; Constantine et al., 2009; Dill, 1996; Grossman & Loeb, 2008; Wilson, Floden, & Ferrini Mundy, 2001; Wilson & Tamir, 2008; Viadero, 2010; Zeichner, 2016; Zeichner & Conklin, 2005; Zeichner & Shulte, 2001). Research has generally concluded that there is support for neither the definitive claims of AR proponents nor, for that matter, CB programs. The National Council on Teacher Quality (NCTQ), claiming that most CB programs are weak, concluded after its first analysis of 85 AR programs that “alternative certification is generally more broken than its traditional counterpart” (Greenberg, Walsh, & McKee, 2014, p. 2) in terms of admissions, preparation and support.

One problem with early studies is that findings were clouded by lack of clarity about what constituted an alternate certification program (Fenstermacher, 1990; Hawley, 1990; Humphrey & Wechsler, 2007; Wilson et al., 2001; Zeichner & Conklin, 2005; Zeichner & Shulte, 2001; Zumwalt, 1991). Various programs were lumped together as “alternatives.” Besides not distinguishing among the considerable variations of alternate programs, some researchers included any program that was different than a traditional four- or five-year undergraduate program. Hence, one- and two-year graduate preservice programs that adhere to similar professional standards as undergraduate programs and whose graduates complete certification requirements before becoming the teacher of record were grouped together with quick-fix alternate certification programs designed to get prospective teachers in front of the class as quickly as possible before completion of certification requirements (Darling-Hammond, 1990; Zeichner & Shulte, 2001). Recent researchers have been more careful in describing the characteristics of the alternate programs they are studying and are less likely to confuse programs that are alternatives to the traditional undergraduate model with alternate certification programs. However, comparing state databases, with their varying definitions, is still problematic.

As seen in this study, another confounding factor is treating teaching as a singular profession when looking at issues of recruitment, preparation, placement, and retention. Elementary teaching and secondary teaching provide different contexts, as do different subjects (Xu, Hannaway, & Taylor, 2011). In fact, the NCTQ has decided elementary teaching “requires too much specialized training in advance of teaching” to be a practical AR option (Greenberg et al., 2014, p. 55). Recent research on the impact of AR teachers on achievement has also pointed to subject matter differences (Boyd, Grossman, Lankford, Loeb, & Wycoff, 2006; Gansle, Noell, & Burns, 2012; Glazerman, Mayer, & Decker, 2006; Heilig & Jez, 2010; Ing & Loeb, 2008; Viadero, 2010).

In light of No Child Left Behind (NCLB) and increased attention to the achievement gap, especially among students of different ethnic and social class backgrounds, there have been attempts to prove the superiority or inferiority of alternate programs in terms of student outcomes and teacher retention. Again, such research usually provokes critique from those on opposing sides, leading to the conclusion that there is no definitive research.

Several studies have found large variation within and between AR and CB programs (Boyd et al., 2012; Humphrey & Wechsler, 2008; Ingersoll, Merrill, & May, 2012; Johnson, Birkeland, & Peske, 2005; Viadero, 2010). Given the complexities of conducting such research, some call for moving away from the “horse race” mentality of comparing AR and CB programs in terms of pupil outcomes and instead look at what program features and constellation of personal factors within specific contexts lead to competent teachers and their retention (Chin & Young, 2007; Cochran-Smith & Villegas, 2016; Grossman & Loeb, 2008; Haberman & Post, 1998; Humphrey & Wechsler, 2007; Zeichner, 2005, 2009; Zeichner & Conklin, 2008). Although this is a most productive focus for research, policy makers are going to continue to make comparisons. And interestingly, despite all the confounding variables at play, there are some common themes in such research to date. Some findings have been repeated in a variety of contexts, with some taking on a more conclusive nature than might be deserved. Hence, as we discuss the findings of this specific early study of learning to teach in New Jersey through AR and CB programs, we will look at the similarities and dissimilarities of findings from other studies in relation to recruitment, preparation, placement, and retention.

RECRUITMENT

Alternate teacher certification programs have attracted teachers with their easy entry, minimal preparation time, and “earn while you learn” feature. Most programs are an improvement over the days when teacher shortages were too frequently met by granting “emergency” licenses to get “warm bodies” into the classrooms.

One aim of New Jersey in initiating the alternate route was to increase the number of prepared and supported teachers, particularly for harder-to-staff urban and rural schools. In 1982, nearly 20% of new teachers entered on emergency credentials (Cooperman & Klagholz, 1985). Quality was also going to be improved by (a) recruiting the “best and brightest,” (b) creating a more diverse teaching pool, and (c) bringing “new enthusiasm and innovative methods of instruction” to stimulate change in New Jersey schools.

Increase Numbers of Teachers

In its third year, the AR program provided about one quarter of the new secondary English, secondary math, and elementary teachers who were prepared in New Jersey and ready to start teaching in fall 1987. Arguably, some people might have entered teaching by enrolling in undergraduate or master’s programs if the AR program had not existed. But the fact that school districts agreed to hire 129 English, math, and elementary teachers if they entered the New Jersey AR program indicates that they were likely worried about finding enough teachers from among newly graduated, certified teachers, out-of-state teachers, or reserve pool teachers. Hence, the AR program appears to have been a quick way to increase the number of teachers available for immediate classroom service. It was especially important in doubling the number of new New Jersey-prepared math teachers from 30 to 60 and directing the flow of teachers into harder-to-staff schools by basically giving principals the authority to admit prospective teachers to the AR program. Linking recruitment and placement, this policy initiative successfully increased the supply of teachers for New Jersey’s schools.

New Jersey’s initial experience increasing the number of teachers has continued over the last 25 years and has been replicated elsewhere, as the rapid spread of alternate routes indicates. Although there are continued definitional problems, Feistritzer (2005) estimated that alternate certification routes have provided at least 250,000 teachers since the mid-1980s. The numbers keep growing; about 50,000 teachers entered teaching in 2006 through alternate routes (Feistritzer & Haar, 2008). Reported percentages of AR teachers have varied; for particular years, 15% in New York City, 16% in Texas, 50% in Houston (Zumwalt & Craig, 2005a), and 20% in California (Chin & Young, 2007). Hence, the 25% reported for this early cohort in New Jersey seems relatively high, but it should be remembered that, for the purposes of this study, the cohort included only three teaching areas, including math, an area with more teacher shortages than other fields.

However, this early high percentage is not a fluke. The program has grown, accounting for over 23,800 teachers hired in New Jersey from 1985 to 2005 (Barclay et al., 2008). From 1992 to 2004, approximately 32% of New Jersey’s new teachers entered through the Provisional Teacher Program. About 2,800 teachers participated in the program in 2004–2005, accounting for approximately 38% of the new teachers hired that year. In 2013–2014, approximately 50% of new teachers entered through the AR (New Jersey Department of Education, 2014a). Interestingly, there is proportionately greater demand at the secondary levels, where about 70% were hired in 2005, compared with 60% AR teachers nationally and 42% of teachers employed at the secondary level. The extension of the New Jersey program to vocational education, found mostly at the secondary level, may explain this shift. The percentage of AR teachers hired each year in New Jersey still outpaces the nearly 40% of new public school teachers who enter though alternate routes nationally (Headden, 2014).

The local market for teachers is affected not only by the supply of teachers but also by the proportion of harder-to-staff schools in a particular locale or state. Additionally, the number of options for those who did not choose to become teachers in college affects the attractiveness of AR programs. In New Jersey, which does not require a master’s degree for certification, there were relatively few preservice graduate programs. Hence, for career switchers and other late-entry people, who chose not to enroll in an undergraduate program, the alternate route offers an option for them, as it does for experienced private and parochial school teachers. Added attractions are easy entry with minimal preparation, low cost, and the “earn as you learn” feature. Thus, not surprisingly, among our exemplar AR teachers, we saw more “pragmatic” choices to enter teaching compared with those who chose to prepare in a college teacher education program.

Improve Quality: Recruit the “Best and Brightest”

Using traditional quality measures, neither CB nor AR teachers had a superior quality profile. Although AR candidates had slightly lower grade point averages (GPAs) than CB candidates, they were more likely to have attended a higher status college, which might make the GPA differences meaningless. And although AR candidates had slightly higher National Teacher Examination (NTE) test scores, secondary AR teachers were less likely to have majored in the subject they were teaching. In addition, contrary to the expectation that the AR would attract candidates with stronger subject knowledge (Cooperman & Klagholz, 1985) the CB math (CBM) teachers had taken substantially more math courses in college than the alternate route math (ARM) teachers. At the elementary level, this was the last year that students could major only in education. Hence, although 69% of this CB sample had majored in education, the next year, none would be permitted to major only in education. Among the exemplar teachers nominated by college faculty, all the selected teachers had majored in a content area, not education. So any claimed academic advantages could be seen as a washout and could not legitimately be used to justify the “fast track” nature of the AR program.

Some of the lack of differences in quality measures resulted from tougher requirements for CB students than AR students (e.g., GPAs and college majors were more “relaxed”). Course work and acceptable college major were also more highly specified for CB teachers. The phenomenon of requiring higher standards for teachers graduating from CB programs than those entering AR programs has also been noted in other states, particularly where alternate routes have been developed primarily to address teacher supply issues. In this study, using traditional proxies for measuring entry characteristics, it could not be claimed that the AR program was bringing in an academically more or less able group of candidates.

Regardless of different state requirements, whether AR programs generally attract more academically able people remains inconclusive (Hammerness & Reinsinger, 2008; Humphrey & Wechsler, 2007; Wilson, Floden, & Ferrini-Mundy, 2001, 2002), as it was in our study. Most have found no difference in GPAs or NTE scores (Constantine et al., 2009; Zeichner & Schulte, 2001), although some reported higher or lower test scores for AR candidates, with very high pass rates for all (Zumwalt & Craig, 2005b). In a review of seven AR programs, including the New Jersey program in 2003–2004, Humphrey and Wechsler (2007) found that AR candidates were more likely to have graduated from more competitive colleges, but there was considerable variation depending on the different purposes and recruiting strategies. In New Jersey, 33% had attended competitive colleges, the second lowest of the seven programs, with the highest being TFA with 79%.

Improve Quality: Recruit More Diverse Teachers

Increasing the diversity of teachers was seen as another way to increase the quality of the teaching pool. In its third year, the AR program increased the number of new teachers who were older, had work experience in other sectors, grew up in urban environments, and spoke languages other than English. The program also attracted proportionally more males, more candidates from higher SES backgrounds, and more racial/ethnic minorities. However, most new teachers were still White middle-class suburban females in their 20s who spoke only English.

Age. As expected, the AR program attracted older new teachers who had not chosen to obtain certification during college. In this cohort, over half the ARM teachers had not even considered teaching as a career until after college. Not surprisingly, the average age of the AR teachers was 30, but, interestingly, the average age of the CB cohort, the majority of whom had thought about becoming teachers before college, was 25.

This phenomenon is not unique to this cohort; college graduates who go on to teach are older, on average, than other college graduates. Beginning teachers graduating from college programs are between 25 and 29 years old, not 22 (Zumwalt & Craig, 2005a). In 2011–2012, 30% of new hires in U.S. schools were 29 years or older (Ingersoll, Merrill, & Stuckey, 2014). In this study, the range of ages in both groups was quite similar, with CB teachers ranging from 21 to 51 years old, and AR teachers ranging from 21 to 54 years old. As we saw among our exemplars, some students started college late, and some took more than four years to finish their degrees. Because of the paucity of master’s programs in New Jersey and the reluctance of some college graduates to enroll in undergraduate programs again, the New Jersey AR program increased the range of possibilities for older people to become teachers. Hence, in New Jersey, as elsewhere, AR programs have generally increased the age of entering new teachers. But there are exceptions. Targeting new graduates, TFA recruits have an average age of 23, compared with late 20s for beginning teachers nationally, and 35 for the New Jersey AR program in 2004 (Humphrey & Wechsler, 2007).

The phenomenon of teachers being older when they first entered teaching created a cohort of beginning teachers different from the past, but with some commonalities regardless of their preparation route. For instance, the 25 exemplar teachers, at various points in their lives, chose to become teachers for a variety of common reasons. Like the teachers Johnson (2004) studied, they seldom provided a single reason for choosing teaching. In categorizing these multiple decisions as ranging from “passionate” to “pragmatic,” however, we found that younger teachers tended to express more passionate reasons and older teachers, regardless of preparation type, tended to express more pragmatic initial choices for becoming teachers. Although the majority had a combination of altruistic and practical reasons, CB teachers were more likely to make passionate/pragmatic choices, and generally, AR teachers were more likely to make pragmatic/passionate choices. Because the AR candidates were all postcollege age and in need of immediate employment, this difference is not unexpected. Given the very different retention rates 11 years later, it is interesting to note that the ARM teachers’ initial decisions to enter teaching were more pragmatic ones compared with the ARE teachers’ initial decisions, which were more altruistic.

Prior Work Experience. Being somewhat older, the AR candidates in this third New Jersey cohort were more likely to have had work experiences in other sectors of the economy. Interestingly, one third of them had held teaching positions at the same school level and subject, and 21%–37%, depending on the subject, had been substitute teachers. Hence, one of the consequences of establishing the AR program was enabling experienced, uncertified teachers to become certified quickly and move to the generally higher paying public sector. Having already taught, the “fast-track” preparation program was probably more justified for them than for the majority who had not had teaching jobs.

Perhaps because school districts control entry into New Jersey’s AR program, it attracts an unusually high number of experienced teachers, even higher than in the initial years. In their study of seven AR programs, Humphreys and Wechsler (2007) found that 50% of the New Jersey participants had been classroom teachers, with an average of 39 months on the job. This was considerably higher than the other six programs and higher than the national average of 17% in 2004–2005 (Feistritzer & Haar, 2008). Understandably, school districts prefer to hire experienced rather than inexperienced teachers.

The CB teachers were not without prior work experience. They were more likely to have been preschool or elementary teacher aides. While completing their certification requirements, 80% of them were working; nearly half of the elementary and one third of the secondary teachers had a non-education-related job. And state requirements meant all CB teachers had completed two practicums and a full semester of student teaching.

Among the exemplar teachers, 12 of the 25 entered teaching after brief to lengthy work in various areas, including parenting. Of the 12 teachers, seven chose to enter CB programs, and five chose the AR. This is another reminder that some of the desirable characteristics of AR candidates, such as maturity and prior work experience, also exist among older CB students who did not enter college right after high school.

Race/ethnicity. New Jersey was particularly interested in attracting more teachers who would feel comfortable and work effectively in harder-to-staff urban schools. Like elsewhere, the great majority of teachers were White females who grew up in the suburbs and spoke only English. And like elsewhere, AR programs, especially at the elementary level, have attracted higher proportions of minorities to teaching (Zumwalt & Craig, 2005a). In this early cohort, the New Jersey AR program was relatively successful in attracting more minorities, particularly Blacks, to elementary teaching (31% compared with 12% for CB programs). Attracting minorities to secondary teaching was a tougher sell, but they were somewhat more successful (13% compared with 3.3%). Although most teachers grew up in the suburbs, 20% of the AR teachers grew up in urban areas, compared with 12% of the CB teachers. Unfortunately, given the large number of Spanish-speaking students in New Jersey urban schools, with one exception, the majority of all teachers (65% to 88%) only spoke English. The exception was the ARE teachers, 60% of whom reported fluency in another language.

In 2003–2004, the AR program apparently was even more successful in recruiting minorities, with 42% of the three regional training centers sampled by Humphrey and Wechsler (2007) identifying as minorities, compared with 14% of New Jersey teachers. However, they noted that the proportions may reflect the local teacher market. “In Newark, for example, 59% of AR candidates are African American or Hispanic, the same percentage of minority teachers in the district as a whole” (p. 496). Perhaps for similar reasons, the 42% minority teachers they reported in 2004 is much higher than the 26% reported in the randomized stratified sample of New Jersey AR teachers in 2006 (Barclay et al., 2008). In 2013–2014, the New Jersey Department of Education (2014) reported that AR teachers were 31% minority compared with 15% of new CB teachers, which was the same as the statewide percentage of employed teachers. Similarly, nationally, 17.3% of teachers were minority in 2011–2012 (Ingersoll, Merrill, & Stuckey, 2014). Clearly the AR program is a source of minority teachers for New Jersey schools.

Gender. In this early cohort, the New Jersey AR program was even more successful in attracting males than it was in attracting minorities to teaching. Males comprised from 3% to 17% of the CB teachers, whereas they made up 24% to 62% of the AR teachers, being the majority for ARM teachers. The success may be related to the relative unattractiveness to an adolescent male of choosing a “female” career during college and, hence, the increased attractiveness of the later AR option. It may also be related to the admissions process that required prospective AR teachers to convince a school to hire them as a condition of entry. Schools may be more willing to take a chance on hiring a nonprepared male than a female because of stereotypes about who can handle tough situations, the desire to have more male role models, and the desirability of adding males to schools that are predominately female.

Nationally, approximately 75% of teachers are female, with slightly lower rates in CB programs. Although the majority of teachers are female, males are more likely to be teaching secondary students. Minority males choose teaching at higher rates than White males (Zumwalt & Craig, 2005a). The CB teachers in this study were more likely to be female than nationally, whereas the AR program attracted proportionately more males than nationally. Although many studies have found AR programs attracting proportionately more males than CB programs (Feistritzer & Haar, 2008), other studies have found little or no difference (Humphrey & Wechsler, 2007; Zumwalt & Craig, 2005a). Some authors have noted similar gender variation in comparing graduate and undergraduate programs, whereas others have found no difference (Zumwalt & Craig, 2005a).

Looking at 2006 data (Barclay et al., 2008), it appears that the New Jersey AR program continues to attract a disproportionate percentage of males to teaching—39%. However, it must be remembered that 70% of the 2006 cohort was teaching at the secondary level. In another recent study using 2003–2004 New Jersey data, percentages are broken down by level. Humphrey and Wechsler (2007) reported 10% males in self-contained classrooms and 43% males in single-subject classrooms, which is close to the national averages of approximately 12% at the elementary level and 41% at the secondary level. Data from 2013–2014 indicates that the AR teachers were 27% male compared with 22% of new CB teachers and 23% of all New Jersey teachers (New Jersey Department of Education, 2014a, 2014b). Thus, attracting more males may no longer be an advantage of New Jersey’s AR program, particularly at the elementary level, where it was particularly pronounced in earlier years. Perhaps the growth of the Teach for America (TFA) program—which has attracted an unusually high 22% males to elementary teaching, close to the 24% initially attracted to elementary teaching by New Jersey’s AR—explains the non-noteworthy proportions of males attracted to teaching by the New Jersey’s AR now. Humphrey and Weshsler (2007) have speculated that males are particularly attracted to TFA’s social change agenda and its focus on building future leaders rather than on a teaching career. Started in 1990, TFA was not available to the young men from highly competitive colleges who were attracted to New Jersey’s AR program in the early years

Social class. Attracting teachers from higher SES backgrounds who might have shunned teacher preparation programs in college or gone to colleges without such programs might appear counter to attracting candidates who would feel comfortable and relate well to students in urban schools. However, it was felt that attracting people who likely had had better educational experiences would give a boost to the quality of education provided in New Jersey schools. In general, this initial group of AR teachers, who had attended higher status colleges, was also more likely to have parents with higher educational and occupational status. Although the SES backgrounds of most teachers from both groups were relatively modest, their parents generally had higher educational attainment than teachers do nationally. Because data comparing SES backgrounds of new teachers are not readily available, this may represent the gradual increases seen across generations and/or higher educational attainment levels in the Northeast.

School level/subjects and diversity. Like teachers nationally, the elementary teachers were most likely to be female and have the highest proportions from racial/ethnic minorities (Zumwalt & Craig, 2005a). At 31%, the alternate route elementary (AREL) program attracted the highest proportion of racial/ethnic minorities to teaching.

English teachers were the most likely to be White, had attended the most selective colleges, were the most likely to have majored in the subject they were teaching, and had the highest GPAs, and a majority of them had at least one parent who had completed college. Additionally, unlike other groups, a majority of the ARE teachers spoke more than English and had fathers holding professional jobs. Generally, English teachers came from more advantaged backgrounds, socially and academically, than the other teachers.

Math teachers were more likely to be male, but only the ARM teachers were majority male. In terms of other background characteristics, however, the math teachers were the most different from each other. ARM teachers were considerably less likely than their CB counterparts to have majored in math in college, least likely to have thought of becoming teachers before or during college, and most likely to be from sectors other than education, and the least likely AR teachers to have had previous teaching experience. In contrast, the CBM teachers, most of whom majored in math, were the most likely teachers to have thought of teaching before college. In some senses, it was in secondary math that the AR program attracted the teachers who were least like those prepared in CB programs.

Summary. Although other AR programs may not have increased diversity on all five dimensions, New Jersey’s program in its third year was able to attract a more diverse group of new elementary, English, and math teachers. Given the needs of urban schools, its particular success in increasing the number of racial/ethnic minorities, teachers who grew up in urban areas and who spoke languages other than English, was particularly noteworthy.

Improve Quality: Recruit Status Quo Changers

Just before they were to assume responsibility for their own classrooms, 60% of CB teachers chose progressive ideals, and 70% of the AR teachers chose traditional visions of the ideal teacher. Depending on one’s view of whether schools are currently too progressive or too traditional, the views of these prospective teachers could lead one to be hopeful that the AR teachers would refocus schools on traditional goals or that the CB teachers would bring fresh ideas to schools.

Although the progressive orientation of CB teachers and the traditional orientation of those prepared in fast-track programs might be expected, looking at grade-level and subject differences reveals a more complex situation. The CBM teachers were the ones who deviated from the general trends, with slightly less than half choosing the progressive ideal. Like the AR groups, they were most likely to choose the traditional LESSONS as their ideal. However, the two progressive ideals (MYSTERY and NURTURING) were their second most likely choices, compared with the AR groups who chose BITE-SIZE. Perhaps the nature of school math led most to choose traditional lessons, but the progressive ideals promoted by many CB programs could also be seen. The AR teachers were most focused on what they saw as the fundamentals of teaching, clear lessons that enabled students to learn bite-size chunks of content.

This conventional notion was also seen among the AREL teachers, whom one might expect would be more oriented, like the stereotypical elementary teacher, to a nurturing role. Only 17% chose the NURTURING role, compared with 39% of the CB teachers. It is unclear whether their achievement focus represents varying motivations for entering teaching or different orientations between CB programs and those run by the state about the primacy of achievement as measured in traditional ways.

PREPARATION

Although prospective teachers were attracted to teaching for similar reasons, they had different motivations for gaining certification through a CB or AR program. The unique “earn while you learn” feature of the AR program was particularly attractive, whereas CB teachers attributed more importance to program features at specific colleges. While the Boyer Topics (Boyer,1984) were supposed to be in place everywhere, the CB programs, most having existed for many years, had reputations and noted features that may have distinguished one from the other. The AR’s newly established regional training centers, offering a generic curriculum and common price structure, did not have distinct reputations. Also, AR teachers did not pick a specific training center, but were assigned based on the school district that had hired them. So it is not surprising that they were more job oriented, whereas the CB teachers were more program oriented.

Most of the AR teachers, being more job and income oriented, said they would not have become teachers if they had to delay their entry into teaching by entering a CB program, most of which were undergraduate programs. Interestingly, once again it is the ARM teachers who stood out from their peers, with 71% of them saying they would not have entered a CB program, compared with 53% of the AREL and 59% of the ARE teachers. If more preservice graduate programs had been available in New Jersey, perhaps their response would have been different. These early New Jersey AR candidates were even less inclined to teach than a national sample of AR teachers in 2004–2005, 47% of whom said they would not have entered teaching if the alternate route had not existed (Feistritzer & Haar, 2008).

One could argue that this indicates a weaker commitment to the teaching profession given that these prospective teachers were not willing to make the financial and time sacrifices to become teachers. Or perhaps, being college graduates, they had financial and time constraints that made the AR program the only feasible option from their perspectives, having not made that choice as undergraduates. However, among our exemplars, CB elementary teachers Phyllis, Angie, and Teresa—who already had a bachelor’s degree—decided to enter a CB program, unlike the 11 other prospective teachers with bachelor’s degrees who entered the AR program. Four of the five other older prospective teachers, Beth (CBM) Connie (CBM), Johanna (CBM), and Diana (CBE), also chose a CB program rather than the AR because they needed to complete their bachelor’s degree. Only Linda (ARM), an older college student, who wanted to take more math courses, decided not to pursue teaching in college, but later in the AR program. Regardless of their motivation and commitments, the existence of the AR program opened up teaching to a number of people who would not have been available to teach in New Jersey schools in the past.

Program’s Vision of Ideal Teacher

Regardless of their own views, most of these prospective teachers entered CB and AR programs staffed by faculty and school personnel whom they reported were more likely to hold traditional views about the ideal teacher. The only exceptions were slightly more than half the CBEL teachers, who viewed their student teaching supervisors and the cooperating teacher as holding progressive visions of the ideal teacher. Once again, the most traditional visions were noted by the ARM teachers, 90% of whom said their supervising teachers held traditional visions. Although they too were the prospective teachers most likely to hold traditional visions (76%), the gap between the views of their supervising teachers and themselves was also the greatest of any group.

Curriculum

Although the Boyer Topics (Boyer, 1984) had been mandated as the curriculum for all CB and AR programs, the college’s curriculum was spread out over three years, whereas the AR teachers received the “same” curriculum, starting the summer before the school year, for some, and, for most, after school and on Saturdays, during their first full year of teaching. This organization, structure, and timing of engagement with the content undoubtedly led to different experiences for these AR teachers.

Besides the generic nature of the AR curriculum, the major difference was in the depth and breadth of their exposure to public school classrooms and opportunities to try out teaching before assuming full-time responsibility as a classroom teacher. With required sophomore and junior year practicums and a full semester of student teaching, the CB teachers had more practical experience in the classroom compared with Phase 1 for the AR teachers. The state required that student teaching be completed in the teachers’ certification field, but this was not necessarily so for the AR teachers, especially at the elementary level. This mismatch between Phase 1 and teaching subject and the generic nature of their preparation led to a rougher transition for AR teachers as they began their teaching careers. As well, AR teachers were supposed to be supervised intensively during Phase 1, but only 63%–64% of them reported that they were observed at least weekly, compared with 96% of CBEL and 79% of CB secondary teachers. AR teachers were more likely to say that the amount and quality of supervision was worse than expected compared with CB student teachers.

Although most prospective teachers taught the full range of students during student teaching and Phase 1, we again see differences between the CBM and ARM teachers. Whereas the majority of both cohorts taught average/below-average students, the ARM teachers were more likely to teach remedial students than their CB counterparts, who were more likely to teach above-average and gifted students. Perhaps this helps explain why the ARM teachers, compared with others, were most likely to think their supervising teachers ascribed to traditional views of teaching.

In speaking with our exemplar AR teachers, although their assessment of the regional training sessions varied greatly, they generally found the ongoing opportunity to get together with peers from other schools several times a week particularly helpful because they faced their first teaching jobs without the usual student teaching experience and grade- and subject-specific methods courses. Hence, this structural feature creating collegial networks, intentionally or unintentionally, may have provided some compensation for what they had missed in the more intensive, specialized CB teacher education programs.

Assessment of Program

Six years later, most CB (78%) and AR teachers (69%) rated their respective programs as “good” or “very good,” unlike other studies indicating that AR teachers were more dissatisfied with their programs (Carter, Amrein-Beardsley, Hansen, 2011; Darling-Hammond, Chung, & Frelow, 2002). Although close to the same percentages said they would choose the same pathway to teaching again, about one fifth of each group said they would instead choose a graduate teacher education program, an option that was not widely available in New Jersey at the time. Some of these CB teachers might have been college graduates who would have preferred being graduate students rather than going to school with undergraduates again. Some of these AR teachers might believe that a master’s degree would have been better preparation and/or more portable to other states than AR certification. From their perspective of six years, few CB teachers (4%) would have chosen the AR program instead, whereas more AR teachers (17%) would have chosen to enroll in a CB program.

PLACEMENT

With the AR program, New Jersey aimed to increase the pool of teachers willing to work in urban and low-wealth districts and meet the needs of low-achieving students. Whether this early cohort met these goals is not as straightforward as one might expect. After discussing the findings about where teachers found jobs, we look at how their initial placements may have affected their views. The match between their preparation, interests, and placement may have been a critical factor in shaping the nature and length of their teaching careers. Last, we turn to the exemplar teachers for a better understanding of the many choices made by these teachers as they taught in different contexts.

Urban Schools

Because AR programs are often designed to staff urban schools, it is not surprising that many studies report that AR teachers prefer teaching in an urban setting and initially take jobs in urban schools (Hammerness & Reisinger, 2008; Zumwalt & Craig, 2005a). Likewise, in New Jersey, where AR teachers must find a teaching position as a condition of entry into the program, they were more likely to find these positions initially in harder-to-staff schools in urban areas than their CB counterparts.

Unexpectedly, however, the majority of AR and CB teachers found their first jobs in nonurban schools. Among the 25 exemplar teachers, only five AR teachers began their teaching careers in urban or low-wealth districts. This may be a consequence of many AR candidates having taught in the private or parochial sector, making them valuable hires even in schools without traditional shortages. Or it may be a reflection of districts’ perception that these candidates would bring needed diversity (e.g., race/ethnicity, gender, age, other experience) to their school staffs. Or maybe all schools were having staffing difficulties at this time. Interestingly, in 2006, although more new AR teachers (50%) were teaching in urban schools, 41% were still hired in suburban schools (Barclay et al., 2008).

Although data reported by the New Jersey Department of Education in 2014 do not differentiate placement by urban/suburban location, other related data indicate the same phenomenon—more new AR teachers in high-need schools, but the majority teaching elsewhere. In 2013–2014, 19% of new AR teachers were employed in “focus” and “priority” schools compared with 13% of CB teachers and teachers statewide. Likewise, 27% of new AR teachers taught in schools categorized by New Jersey as the lowest economic group—District Factor Group A—compared with 15% of new CB teachers and 17% of all teachers statewide (New Jersey Department of Education, 2014a, 2014b).

In our study, the initial differences in placements decreased over the next four years, primarily because AR teachers left urban schools or left teaching. However, by the sixth year, AREL teachers, having expressed the greatest preference for teaching disadvantaged children and working in urban schools, were still more likely to be teaching there than CBEL teachers. However, AR secondary teachers were the least likely to be teaching in urban schools. The AR program was successful in getting elementary teachers, but not secondary English or math teachers, to spend their first six years in urban schools. This differential success may be related to demographic differences; the AREL teachers were the group most likely to be non-White and to have grown up in urban areas.

Low-Wealth Schools

Somewhat similarly, only the majority of AREL and ARM teachers found their initial jobs in low-wealth districts. Again the differences between the groups diminished over the first six years. By then, half or more of the AREL and ARE teachers were teaching in low-wealth districts. The most striking change was among the secondary math teachers, with almost half of the CBM teachers in low-wealth districts but only a third of ARM teachers there. Hence, both the AR and CB programs were relatively more successful in attracting new teachers to low-wealth districts than to urban schools. The pronounced exodus of ARM teachers out of low-wealth districts is another indication of the different experiences of this group of teachers.

Types of Students

Over the first three years, as the types of schools converged, so did the types of students they taught, except for the math teachers. In the first year, the ARM teachers were more likely to teach remedial students, but by the third year, the opposite was the case. By the sixth year, CBM and CBE teachers were both more likely to be teaching remedial students than their AR counterparts. Some of the difference might be attributed to the fact that more CB secondary teachers were teaching in urban schools, and more CBM teachers were teaching in low-wealth districts than their AR counterparts. Some of the difference might be related to the fact that the CBM teachers were more likely to have majored in math and had taken more math and math methods courses that might increase their ability to address the needs of remedial students.

Views of the Ideal Teacher in Schools

CB and AR teachers’ views of the ideal teacher also converged. Before teaching full time, 60% of the CB teachers expressed progressive views, and 70% of the AR teachers articulated more traditional views of the ideal teacher. After the first year, their views were more evenly split between traditional and progressive views. With a few exceptions, the teachers found that their colleagues, supervisors, and administrators were more likely to hold traditional than progressive views of the ideal teacher. As the literature has described (Lortie, 1975; Zeichner & Gore, 1990), the conservative ethos of schools had the typical socializing impact on the CB teachers but seemed to have the opposite effect on AR teachers. The result was a convergence of views, with a roughly equal split between traditional and progressive ideals for both groups.

However, underneath the convergence were some differences by subject. CB teachers’ views seemed to stabilize after the first year, but the AR teachers’ views were more unstable, perhaps because they were less familiar with working in schools than were the CB teachers. By the end of second year, the elementary teachers were indistinguishable in their views, but the AR secondary teachers’ visions were more distinctive. The majority of ARE teachers went from expressing traditional to progressive to traditional visions by the end of the second year. Their views seemed to mirror their changing perceptions of administrators’ and supervisors’ views.

Before full-time teaching, the ARM teachers were most likely to express traditional views (76%). After the first year, like all CB teachers, they were evenly split (50%), but after the second year of teaching, they were the group most likely to express progressive views (60%). They consistently reported traditional views of the people they worked with in Phase1 and during the first two years of teaching. However, unlike the ARE teachers, whose views seemed to mirror the views they perceived around them, the ARM teachers seemed to take a counter-stance, expressing more progressive views each year. Some of this change may reflect their growing unease with the views found in schools, and some may be an artifact of selective attrition of those with more traditional views. (ARM teachers had the highest early attrition rates.) But whatever the reason, we see another indication of the different experiences of the ARM teachers.

Subject and School Level Match

Generally, CB teachers were more likely than AR teachers to find their first jobs at the level and in the subject for which they had prepared. CBM and CBEL teachers had the highest matches between preparation and teaching area. At the secondary level, there were some job market factors at play that cut across preparation programs. Whereas approximately two thirds of the English teachers had found jobs teaching English at the middle school or high school level, 91% of CBM and 80% of ARM teachers were teaching math at the secondary level. Although not teaching their subject, CBE teachers were more likely to be substituting, whereas ARE teachers, with their generic certification, were more likely to be teaching at the elementary level. They were also teaching math, drama, computer science, media, and gifted and talented classes. Hence, in the eyes of either the hiring school districts or the candidates themselves, the generic preparation of the ARE teachers made it easier to find employment as full-time teachers despite the tight job market for English teachers.

Exemplar Teachers

The exemplars give us more insight into the different placement history of CB and AR teachers. All 13 CB teachers found jobs in the subject and school level in which they had student taught. Five of the 12 AR teachers had initial placements that did not match their subject or grade-level interests or undergraduate major. This initial match or mismatch also affected later placements, as in Donaldson and Johnson (2010). Instead of adapting to schools that generally matched their interests and skills, the AR teachers were more likely to switch teaching jobs and/or leave teaching, temporarily or permanently. After six years, 10 of the 13 CB teachers were teaching in the same school or same school district, whereas only two of the 12 AR teachers were teaching in the same school district. These two AREL teachers were distinguished by their extensive personal experiences in the urban communities where they chose to teach (Randi, 2017).

The greater likelihood of CB teachers finding a good match in their first jobs may be related to both the content and length of their preparation program. They knew not only more about different placements but also more about themselves as teachers, their areas of strength, and their interests, having had field experiences, practicums, and student teaching prior to looking for a teaching job. As certified teachers, with actual experience in schools, they had a greater choice of jobs available to them than the AR teachers, who had to find a specific job in a district that had not filled the position with a certified teacher yet. Once in a job, the CB teachers seem to have had more understanding of how to negotiate the school culture in a way that worked for them and their school.

The advantage of this initial good match had personal consequences for them in terms of more job stability, but also consequences for the hiring district and, ultimately, for the profession. Their initial and subsequent teaching placements were a critical factor in their experiences in learning from these early years, in how long they stayed in a particular teaching position, and, ultimately, in how long they stayed as classroom teachers.

Exemplar teachers’ classroom practice. Although recruiting and retaining teachers, particularly in harder-to-staff schools, are clear measures of success, as important is the quality of the teaching practice of those hired and retained. At the most general level, the willingness of school principals to rehire teachers and, in the case of AR teachers, to grant them certification at the end of their provisional teaching is an indicator of success in the classroom. In this regard, the CB programs were more successful. Two ARM teachers, James and Jeff, did not receive certification from their principals. A third ARM teacher, Margaret, was so disturbed about teaching that she quit suddenly before Thanksgiving of her first year. An AREL teacher, David, left after one year, frustrated with classroom management.

The departures of other teachers were not directly initiated by quality issues, but rather by their attempts to find better matches for themselves, either in or outside the classroom. It is hard to read anything about the quality of practice from the greater mobility of the AR teachers who stayed in teaching. Their greater mobility, however, is likely to have had consequences for the schools they left behind.

Through interviews and observations, we did get a glimpse of their teaching practice as it played out. Initially, we saw some differences between the preparation routes, probably affected by both the contexts of where the teachers were hired and the length and nature of their preteaching experiences. Not unexpectedly, many of the AR teachers faced classroom management issues in their first few years. As described in Sawyer (2017), interacting with these management issues were their defined and relatively static conceptions of subject matter that tended to constrict their teaching and complicate the development of expertise in their teaching practice. However, for those who remained as teachers, the strength of their personal dispositions (e.g., imagination, flexibility, reflective stance, empathy, openness to diversity, and a sense of agency) propelled them forward despite their limited initial understandings of teaching and learning (e.g., view of content, knowledge of learning theory and learner-centered methods.) Many had the drive to pursue such knowledge on their own or by later attending graduate school.

Generally, the practice of AR teachers was initially marked by unevenness, reflecting more trial and error efforts than that of CB teachers, who had had many preteaching opportunities to try things out. CB teachers were consistently focused on student learning. They seemed to have familiarity with other-than-conventional visions of teaching, which jump-started most of them to take an active role in shaping the curriculum. At the elementary level, they quickly focused on the creation of integrated, interdisciplinary, learner-centered environments, while at the secondary level, they demonstrated their understanding of pedagogical content knowledge, adapting content to their learners. While at both levels, CB teachers demonstrated differentiation in some content areas and in some classes (e.g., basic skills classes), the differentiation was generally culturally neutral.

In contrast, the three AREL teachers, who were also the only teachers of color among the 25 exemplars, brought something others did not to their curricular decision-making. Kathy, despite remaining rather traditional in her approach to teaching, saw herself as a bridge between her students’ home culture and mainstream American culture. Gwen and Risa brought not only knowledge to support their diverse students but also the commitment to teach for social justice. Although their teaching skills often lagged behind their aspirations of creating curriculum to promote social change, they were the only ones who explicitly took this stance from the beginning. All three illustrate the contributions of the AR in diversifying the teaching pool.

Over time, most of the exemplar teachers who stayed with teaching, regardless of their preparation route, became active curriculum makers, seeing the curriculum as an evolving creation of teachers, students, content, and specific context. Instead of implementing the curriculum, they grew to see themselves and their students as actively enacting curriculum (Paris, 1993; Snyder, Bolin, & Zumwalt, 1992). Perhaps it was this active, student-oriented focus that sustained their learning and desire to stay as teachers.

RETENTION

Findings relating to retention of AR teachers have been mixed (Boyd, 2005, as cited in Feistritzer & Haar, 2008; Humphrey, Wechsler, & Hough, 2006; Zumwalt & Craig, 2005a, 2005b). Wilson et al. (2001, 2002) found little relationship between program and retention. Stoddart (1990) found lower rates of retention among AR teachers in Los Angeles, as did Boyd et al. (2012) for AR math teachers in New York City. Kirby, Darling-Hammond, and Hudson (1989) found that AR programs did not overcome the usual problems of retaining math and science teachers. Grissom (2008) argued that the higher attrition rate for AR teachers is primarily due to the characteristics of the schools in which they teach. Similarly, for TFA teachers, Donaldson and Johnson (2010) found that attrition was strongly influenced by the difficulty of the teaching assignment or misassignment. Others have suggested that higher attrition rates for AR teachers may reflect higher attrition rates of teachers from highly selective colleges (Kelly & Northrup, 2015). Some have suggested that generational attitudes about shorter and more varied careers may account for the unprecedented high level of teacher turnover (Headden, 2014; Peske, Liu, Johnson, Kauffman, & Kardos, 2001). Headden also speculated that the TFA two-year model may be prompting all teachers to think about the duration of their careers differently.

Several studies indicate that higher proportions of AR teachers are retained in the first years of teaching compared with teachers prepared in CB programs or compared with national averages (Adams & Dial, 1993; Zeichner & Schulte, 2001). This higher initial retention may be an artifact of the two- or three-year commitments required by some AR programs and to higher retention rates generally for older teachers (Zumwalt & Craig, 2005a).

More generally, school context, program goals, structure, content, and personal and life cycle factors contribute to mixed retention results (Borman & Dowling, 2008; Humphrey et al., 2006; Rinke, 2008; Tamir, 2013; Zeichner & Conklin, 2005). For example, TFA, which only requires a two-year commitment from its highly selective recruits, has relatively high attrition rates (Heilig & Jez, 2010). Rather than preparation route, Ingersoll et al. (2012) found that first-year attrition was highest for math and science teachers who had little or no pedagogical preparation and practice teaching. And in our study of the New Jersey AR program, a simple comparison between retention rates for CB and AR programs masked different experiences of ARE and ARM teachers (Natriello & Zumwalt, 1992). Also, looking at three, six, or 11 years postentry made a difference.

Retention Rates Through Third Year for Total Sample

In our study, entry rates for both AR and CB graduates surpassed national rates. According to Henke, Chen, and Geis (2000), only 41% of 1992–1993 newly graduated education majors nationally, who had not taught before obtaining their degree, entered teaching right after college. At the end of their program, 91% of the CB students said that they had found jobs or were looking for a teaching job in September. At least 80% of the CBM teachers, 74% of the CBEL teachers, and 58% of the CBE teachers were teaching the year after graduation. In contrast, 100% of the AR teachers had jobs in September, given that being hired by a school district was a condition of entry into the AR program. After their first year, approximately two thirds of the AREL (69%), ARE (67%), and ARM (60%) teachers were teaching the next year.

Although these hiring rates were relatively high for both CB and AR programs, we do not know if the number of CB teachers who actually found teaching jobs in their first year would have been higher if the AR had not existed. We know more CB graduates wanted teaching jobs than ended up being hired the year after their graduation. We do not know if AR teachers took jobs away from deserving CB graduates or replaced weaker CB graduates, or whether the jobs the AR teachers obtained were jobs that CB teachers would have accepted. Regardless of the reality, there was a perceived competition between CB and AR programs for teaching jobs because jobs were not as plentiful as they had been.

In their third year of teaching, comparisons between the CB and AR programs indicate similar retention rates for elementary teachers (72%–75%), the same but lower rates for English teachers (58%), and considerably higher rates for CBM teachers (73%) compared with ARM teachers (47%). Hence, in looking at short-term retention, the findings are mixed; this highlights the need to look at retention rates by subject rather than by program, echoing Murnane, Singer, Willett, Kemple, and Olsen (1991) and Zeichner and Shulte (2001).

National attrition rates are generally highest in the first three or four years of teaching, averaging about 10% annually (Zumwalt & Craig, 2005a). However, attrition rates can vary. For instance, nationally during the economic recession in 2007–2008, teachers had a third-year attrition rate of only 15% (Gray & Tale, 2015). In this study, third-year attrition rates for English teachers (42%), and especially ARM teachers (53%), was considerably higher than the expected 30%. For English teachers, the issue was finding a job; for the ARM teachers, it was staying in the job.

Retention Rates Through Sixth Year for Total Sample

At the end of six years, retention was highest for CBM teachers (63%), with 44%–49% of the elementary teachers and 42% of CBE teachers still teaching. The lowest rates were for the secondary AR teachers: 25% for ARE and 33% for ARM teachers. The higher retention rates for elementary teachers compared with secondary teachers fit national patterns (Zumwalt & Craig, 2005a). Retention rates for the New Jersey CB secondary teachers, especially in math, appear relatively high, whereas the AR secondary program retention rates are relatively low.

Although the AR was less successful in retaining secondary teachers, there were 10 math and six English teachers who might not have been teaching if the AR program had not existed six years earlier. Although these numbers might not seem noteworthy, only 19 CBM teachers and 15 CBE teachers were still teaching at the end of six years. Hence, despite the considerably lower retention rates, the AR was still providing 34% of the math teachers and 29% of the English teachers who had prepared to teach in New Jersey six years earlier.

Retention rates for minorities. The advantage of the AR program in attracting proportionately more minorities diminished over six years but continued for elementary and math teachers. The AR had also attracted minorities to secondary English teaching, but none was still teaching six years later. The AR’s success in attracting and retaining minority elementary teachers matches national data indicating that teachers of color are more likely to teach at the elementary rather than the secondary level (Zumwalt & Craig, 2005a).

The attrition rates for groups indicate some differences. Over the six years, the percentage of teachers who were minorities dropped for AREL teachers (31% to 24%), but increased for ARM teachers (13% to 20%). Two of the four minority ARM teachers were teaching six years later. Although small numbers, the relative success of retaining minority ARM teachers compared with ARM teachers generally is noteworthy.

Although minorities were a smaller proportion of CBEL than AREL teachers, New Jersey colleges actually produced more minority elementary teachers than did the AR program. Minorities who prepared through the CBEL programs also had a relatively higher retention rate, and the only minority CBM teacher was still teaching at the end of the sixth year.

Retention rates for males. The AR advantage in proportionately attracting more males into teaching was retained for all groups through the first six years. Compared with national data (Zumwalt & Craig, 2005a), the proportion of males in AR programs was higher than expected, and the proportion in CB programs was lower than expected.

These differences were accentuated by attrition. Retention rates for the AR were higher for males than for females, whereas for CB programs, it was the opposite. Over the six years, the percentage of male AR teachers increased from 24% to 30% for AREL teachers, from 42% to 67% for ARE teachers, and from 62% to 70% for ARM teachers. In contrast, the percentage of male CB teachers dropped from 3% to 2% for CBEL teachers, from 17% to 13% for CBE teachers, and from 17% to 5% for CBM teachers.

Because other studies report mixed results about gender and retention, it is hard to say whether the higher retention rate of AR males or the lower retention rates for CB males is more typical. One could speculate that males who choose a predominately female profession after college rather than in college may be less likely to doubt their choices later. Or perhaps it says more about the strength of commitment made by females who choose teaching in college than those who choose it after college.

When looking at the actual numbers at entry and six years later, we see that in these three fields, the AR added more males to the teaching pool (see Zumwalt, Natriello, Randi, Rutter, & Sawyer, 2017, Table 5) than it added minorities (Zumwalt et al., 2017, Table 4), but the New Jersey colleges added more minorities than males to the teaching pool. Likewise, after six years, the overall retention rates for the AR were higher for males (45%) compared with minorities (33%), whereas the overall retention rates for CB programs were higher for minorities (43%) compared with males (29%).

These differential retention rates for minorities and males are interesting in that, unexpectedly, in this cohort, the AR was more successful in increasing diversity by attracting and retaining more males than minorities into teaching. Although the American teaching pool is predominantly female and White, perhaps minority males still felt more comfortable choosing teaching during their college years as compared with other males. Hence, the AR provided a second chance for late entry nonminority males.                             

Former Teachers’ Reasons for Leaving Teaching

Asked why they left teaching, similar percentages from both programs, at the end of the second year (30%) and sixth year (15%), said they had decided they just did not want to teach. At both times, the AR teachers were more likely to have left because of an “unpleasant experience,” and the CB teachers because they could not find a job. The latter reason is ironic given that one reason the AR was initiated was to supply teachers. However, teacher shortages were not necessarily in schools where CB teachers wanted to teach. Unknown is whether the AR teachers were leaving because of unpleasant experiences that CB teachers were avoiding, or their short preparation left them unprepared to meet the challenges. The intensity of the AR program—after school and Saturday classes—may have also contributed to their “unpleasant experiences” and to the 35% who said they left because “they needed a rest.”

Attitudes and Beliefs of Teachers Related to Retention

The different reasons cited by former AR and CB teachers are also reflected in teachers’ expectations about teaching at the end of the second year. For CB teachers, the aspects of teaching that were worse than expected related to the job market, whereas the actual conditions of teaching were worse than expected for AR teachers. The expectations of the CB teachers might have been more realistic because of their three years of field experiences before assuming their own classroom or because they were less likely to find their first teaching jobs in low-wealth or urban schools. The AR teachers finding on-the-job supervision worse than expected could be related to greater needs resulting from their truncated preparation time and to actual differences in the quality of supervision in low-wealth and urban schools.

CB teachers were also more likely than AR teachers to cite factors that encouraged them to remain in teaching. At the end of the critical first year, CB teachers were more likely to mention colleagues, good teaching evaluations, and an invitation to stay at the school. These more positively perceived initial years likely had an effect on retention.

When asked about incentives that might keep them in teaching, all teachers most frequently indicated that a “substantial salary increase” would be helpful. Living in the Northeast, with its relatively high cost of living and greater range of salaries for entering professionals, may have accentuated the concern about salaries. They also thought that several other reforms would be significant in retaining them: opportunities for professional development and advancement, such as career ladders and master teacher status, smaller class size, and greater teacher autonomy in curricular and teaching decisions.

The only proposed reform initiative provoking a different response from the CB and AR teachers was mandatory continuing education. Perhaps CB teachers, having experienced more formal teacher preparation, were more predisposed to such education because they had found it useful and engaging, and/or because it signified professionalism. In contrast, the AR teachers might not have found their abbreviated formal teacher education, with its perceived emphasis on class participation rather than other goals, helpful or intellectually engaging. It should also be remembered that they were still involved in this aspect of the AR program, might have preferred options to anything mandatory, and/or might not have seen continuing education as essential to teaching or a professional identity.

Possible Explanations for Differential Retention Rates in First Six Years

Treating teaching as a singular profession is particularly misleading when looking at issues of recruitment and retention. The teaching fields included in this study—elementary teaching, secondary English, and secondary math—present different challenges for contextual reasons, as well as attract different types of prospective teachers and retain them at different rates.

At the end of three years, elementary teachers from both programs and the CBM teachers had the highest retention rates (72%–75%). The lowest retention rate was for the secondary ARM teachers (47%). The retention patterns across subjects and school level were somewhat different in the sixth year, but the trends had been set. The only group in which a majority of the teachers (63%) were still in the classroom was the CBM teacher group. The lowest retention rates were among the ARM (33%) and ARE (25%) teachers.

In looking for possible explanations, we note that some differences cut across preparation programs. For example, more minorities were attracted to elementary teaching, English teachers were from more academically and socially advantaged backgrounds, and males were more attracted to teaching math than English or elementary school. In each case, these differences were accentuated in the AR program, which attracted higher proportions of minorities to elementary teaching, more advantaged candidates to English teaching, and more males to teaching math.

Although these patterns across programs were clear, they do not explain differential retention rates for secondary teachers or the relatively similar rates for elementary teachers. Perhaps elementary AR teachers were more similar to CB teachers, or the short preparation program worked better at the elementary level. Or perhaps the AR program attracted a different pool of late-entry teachers at the secondary level.

Elementary teachers. Like teachers nationally, the elementary teachers were more likely to be female and minority than the secondary teachers. Compared with the secondary teachers, they had parents of lower educational and occupational backgrounds, had lower college GPAs, and were the most likely to report preteaching experiences working with children. By the end of the second year, the majority of elementary teachers expressed progressive visions of the ideal teacher. By the sixth year, elementary teachers were more likely than secondary teachers to prefer teaching disadvantaged children, and the majority reported teaching remedial students all six years, unlike the secondary teachers. The majority of elementary teachers from both groups planned to stay in teaching long term.

Although retention rates were similar for CB (49%) and AR (44%) elementary teachers, it should be noted that the AREL teachers were most likely to be teaching in urban and low-wealth schools. Perhaps some of this difference can be related to the fact that the AREL teachers were the group most likely to be minority, were most likely to have grown up in urban areas, and expressed the highest preference for teaching disadvantaged children at entry. In terms of the goals of the AR program and in terms of retention, perhaps unexpectedly, it was at the elementary level that the New Jersey Provisional Program was most successful.

English teachers. The English teachers were most likely to be White, were least likely to be married, attended more selective colleges, had higher GPAs, and were the most likely to have majored in the subject they prepared to teach, and a majority had at least one parent who was a college graduate. The ARE teachers attended the most selective colleges and had the highest GPAs, and the majority had fathers who held professional jobs. Additionally, a majority of them spoke more than one language. Although socially and academically advantaged compared with elementary and math teachers, the secondary English teachers had divergent retention rates at six years—42% for CBE teachers and 25% for ARE teachers.

In terms of their student teaching, Phase 1, and their later employment, the English teachers were less likely to be teaching in low-wealth and urban schools compared with elementary and math teachers. Although at the end of the second year, the majority of them planned to teach long term, labor market conditions were more challenging for them. English teachers had the most difficulty finding jobs in their field, either because English jobs were not as plentiful or because they were not as inclined to look in the low-wealth and urban districts. Because the AR teachers needed jobs to be in the program and had generic certification, they were initially more likely to be teaching at the elementary level, whereas the CB teachers were more likely to be substituting.

In the fall of 1987, when 100% of the ARE teachers were teaching in schools, only 58% of the CBE teachers were teaching full time. During the next three years, the percentage of ARE teachers dropped from 100% to 67% and then to 58%, matching the proportion of employed CBE teachers. In the next three years, retention rates for the ARE teachers (42%, 46%, 25%) decreased more than for the CBE teachers (53%, 44%, 42%). Hence, although the CBE teachers never reached the employment levels of the ARE teachers in the first two years, higher proportions were teaching in later years.

The higher attrition for ARE teachers may be related to their short preparation, the greater likelihood of reporting “unpleasant experiences,” more unmet expectations, or their later commitment to teaching. Perhaps later entry to a female profession had different meaning for female than male ARE teachers, who were more likely to stay. Possibly, as the most socially and academically advantaged group, ARE teachers could be expected to have a higher attrition rate, as some other studies indicate (Zumwalt & Craig, 2005a).

Math teachers. Math teachers were more likely to be male, attended less selective colleges than English or elementary teachers in their programs, and were the only teachers not reporting Spanish as the most common second language, and the majority expressed traditional visions of the ideal teacher before teaching. But there were greater differences in their backgrounds and their work experiences than in other CB and AR groups, which may account for greater differences in retention rates for math teachers. In the sixth year, 63% of the CBM teachers were teaching, whereas 33% of the ARM teachers were teaching.

ARM teachers were majority male at entry (67%) and at six years (70%), whereas the percentages were lower for CBM teachers (17% and 5%). ARM teachers, who had the highest average age (30 years), were more likely to be married (50%) than CBM teachers (20%), whose average age was 25. Unlike CBM teachers, the majority had at least one parent who was a college graduate. Compared with all groups, ARM teachers had the most other-sector work experience before teaching but the least teaching experience, were the least likely to have majored in their subject, were the least likely to have thought about teaching before or during college, had the largest gap between their own and their supervisors’ visions of the ideal teacher, were the most likely to express progressive visions after the first year, were the least likely to say they would become teachers without the AR option, were the least likely to have long-term plans to teach, and were the most likely to plan to work in higher education. The AR program recruited math teachers who were more distinct from their CB counterparts than the elementary or English teachers, which may help explain their higher attrition rate.

Their student teaching/Phase 1 experiences also differentiated them from each other. For Phase 1, the majority of the ARM teachers taught in low-wealth schools, and they were more likely to teach in urban schools than other groups. For student teaching, the CBM teachers, who were the least likely to have grown up in urban neighborhoods and expressed the least preference for teaching disadvantaged students of all groups, were the least likely of all groups to teach in urban schools. They were more likely to teach above-average and gifted students during student teaching than the ARM teachers, who were more likely to teach remedial students during Phase 1.

Initial teaching placements followed these patterns, but later teaching experiences diverged in unexpected ways (see Zumwalt et al., 2017, Tables 2 and 3). Despite the favorable job market for math teachers, there was a shift in where these differently prepared teachers ended up teaching. Over the six years, a declining proportion of ARM teachers were working in urban schools (40% to 3%) or in low-wealth districts (72% to 33%), whereas an increasing proportion of CBM teachers worked in urban schools (3% to 13%), particularly in low-wealth schools (18% to 47%). Thus, by the sixth year, it was the CB programs that were not only retaining considerably more math teachers than the AR (63% vs. 33%) but also providing more teachers for urban and low-wealth districts.

The greater attrition rate for the ARM teachers may have resulted from a variety of factors related to the adequacy of their preparation program, particularly in light of their limited relevant prior experience and the fact that only 40% had majored in math. Almost half of them reported they only received “satisfactory” evaluations at the end of the first year of teaching, considerably more than any other group. They were more likely to report unmet expectations and were least likely to think of factors encouraging them to remain in teaching. Former teachers also reported a relatively high level of “unpleasant experiences” as the reason they left. And opportunities—often more lucrative than teaching—were available to them, given that 60% of them had majored in some applied or mathematics-related field. Given that they came from higher SES backgrounds than the CBM teachers, perhaps their higher attrition might be expected. They were also the group that was least like their CB counterparts, perhaps making them more risky long-term hires. Given these factors, perhaps it is noteworthy that one third of them remained in teaching.

Equally noteworthy is the very high retention rate of CBM teachers—the highest of any group and higher than national norms. These predominantly female CBM teachers stood out as constituting the group that had most thought about teaching before college, had the easiest time finding jobs, and, six years later, was most likely to be teaching. New Jersey colleges were considerably more successful than the AR in recruiting, preparing, placing, and retaining secondary math teachers—both in terms of numbers and the relatively large proportion in low-wealth schools.

Similarly, in looking at 2013–2014 employment data from the New Jersey State Department of Education (2014a, 2014b), we see that 66% of eligible CBM teachers were working in New Jersey public schools, compared with 40% of eligible ARM teachers. So although the AR produced more teachers with middle school and/or secondary mathematics certificates of eligibility in 2010–2012, by 2013–2014, more math teachers prepared in CB programs taught in New Jersey public schools. Like in our study, the colleges have been more successful in getting their CBM teachers hired and retained than their CBE teachers (58%). However, changes in the job market were evident because secondary teachers from both routes were more likely to be working as teachers than were elementary teachers in 2013–2014. Only 33% of the AREL teachers and 47% of the CBEL teachers were employed in New Jersey public schools that year. That the AR had the lowest employment for its elementary teachers (compared with 40% ARM and 45% ARE) might also be related to administrators’ less positive judgments about the performance of AR teachers at the elementary level (Ing & Loeb, 2008).

Retention of Exemplar Teachers

Longitudinal data indicate retention differences among school level and subjects, but also highlight some differences for the 25 teachers chosen as exemplars by their respective programs (see Table 1). As might be expected, first-year retention rates for teachers identified by both programs as the kinds of teachers they were trying to attract were extremely high compared with national data and with the study sample. All the teachers, except for the ARM teacher who left in November, completed the first year of teaching. All CB graduates had found teaching jobs, in contrast to their peers in New Jersey and nationally who had trouble finding jobs or decided not to teach after college. However, this may be an artifact of the selection process, given that colleges may have nominated those who had or were likely to get jobs. Reasons for not teaching indicated similar differences found in the larger cohort. CB teachers, especially in the early years, were more likely to cite tight market conditions, whereas AR teachers were more likely to become disillusioned with teaching, go to graduate school, or seek other employment.

Table 1. Retention Rates for Survey and Exemplar Samples

English teachers. Retention rates for exemplar teachers were higher compared with the larger cohort, where job market conditions made it hard for many to find jobs. In fact, unlike the larger cohort, the highest retention rates for the exemplar teachers, in the long run, were for the English teachers, regardless of preparation route. Although there was more stability among the CBE teachers and more “in and out” movement among the ARE teachers, by the 11th year, all but one ARE teacher was teaching. It is noteworthy, however, that five of the seven employed English teachers in the 11th year were happily teaching at the upper elementary or middle school level. Hence, the job market conditions affected even the exemplars, not in finding a teaching job but in getting hired in high schools.    

Exemplar English teachers had the highest retention rate—a refreshing rebuke of the idea that the “best and brightest” exit teaching earlier. At the end of 11 years, the four CBE teachers had taught an average of 10 years; the four ARE teachers had averaged nine years.

Math teachers. Although the exemplar CBM teachers generally had similar or higher retention rates than the survey sample, the opposite was true of the ARM teachers. The AR retention record looks most dismal for the exemplar ARM teachers, none of whom were teaching after the fourth year. As in the larger cohort, however, the most divergent retention rates for exemplar teachers are between the low rates for the ARM teachers and the higher rates for the CBM teachers. Whatever was happening among the math teachers seems to be happening among the exemplar math teachers as well.

In contrast to the English teachers, the exemplar math teachers had the lowest 11th-year retention rates. While not having the same difficulty finding jobs as the English teachers, they did not stay in teaching as long, leaving for other opportunities. Perhaps the more favorable job market for those with mathematical backgrounds partially explains their lower retention rates. At the end of 11 years, the five CBM teachers had taught an average of 7.8 years. In sharp contrast, the four ARM teachers had averaged only two years.

Elementary teachers. The elementary exemplar teachers have higher or similar retention rates as those in the larger sample. However, the patterns for the exemplar elementary teachers are compounded by contextual factors—maternity leaves and administrative positions—and different timing of breaks from teaching. All the CBEL teachers taught for the first seven years, but in the eighth and ninth years, two teachers took maternity leave, and in the 10th year, one moved into an administrative peer coaching position in her school. This left one CBEL teacher, who was a parent before teaching, still in the classroom by the 11th year. In contrast, two AREL teachers left the classroom for graduate school, one after one year and one after four years of teaching. However, the two AREL teachers, who were parents before teaching, were still in the classroom by the 11th year. Besides all being parents before becoming teachers, the three remaining elementary teachers all entered teaching in their 30s, putting them in a different place in their career and life cycle.

The retention of the elementary teachers was not as low as the math teachers, nor as high as the English teachers. These rates, however, do not reflect the relatively long teaching careers of teachers from both programs. At the end of the 11th year, the four CBEL teachers averaged 8.75 years teaching, and the four AREL teachers averaged 6.75 years.

General findings. The longitudinal study of these 25 teachers not only provides texture but also dramatically illustrates the inadequacy of most follow-up studies that report retention rates for teachers in general. The superior retention for the CB teachers (averaging 8.8 years) over the AR teachers (averaging 5.9 years) masks the different rates for different subjects and levels. Both preparation routes were most successful in retaining the exemplar secondary English teachers (averaging nine and 10 years) and least successful in retaining the secondary math teachers (averaging 7.8 and two years). It is likely that factors in the school and general job market were similarly affecting teachers of different subjects and grades. However, most striking is the differential success of the AR and CB programs in retaining the exemplar math teachers, who perhaps had the most outside career options.

Retention rates also mask a complex phenomenon more reflective of the ebb and flow of the teaching profession. Following these 25 teachers reminds us that teaching is an “in and out” profession (Lortie, 1975), confounding easy linear analysis of recruitment and retention patterns and making definitive proclamations of success problematic.

Breaks in service, long part of the teaching culture, have confounded general retention rates, particularly as related to gender (Zumwalt & Craig, 2005b). Grissmer and Kirby (1987) found that 40%–60% of teachers who leave eventually return. National Education Association (1997) data indicate that one quarter of teachers reported taking a break of a year or more, the lowest percentage in 20 years. The 11-year study of these 25 teachers indicates more stability in the careers of CB teachers—they took fewer breaks and were less likely to move from one position to another. At the 11th year, two of the five remaining AR teachers (40%) had taught in the same school or district, whereas six of the eight CB teachers (75%) had taught in the same school or district. They seemed to have found good matches for themselves more quickly.

The greater movement of the AR teachers might be attributed to not having the three years of preteaching experiences that gave the CB teachers a better sense of what to expect and what they wanted. In fact, three AR teachers left after “trying out teaching” for one year or less. The movement also might be related to the kinds of jobs that were available to them as a condition of acceptance into the program. Unlike the CB teachers, they did not necessarily find good matches with their undergraduate majors and their teaching interests, and they were more likely to find jobs in harder-to-staff schools. After two to six years of teaching, another four AR teachers decided to leave after became increasingly disillusioned, particularly with the culture of schools. Having made a postcollege decision to pursue teaching, perhaps some AR teachers had less commitment to teaching as a career and, hence, less patience for negotiating the culture of schools.

Prospective teachers attracted to New Jersey’s AR, one of the first in the nation, were more likely to have grown up and gone to college in other states, unlike the CB teachers, who all enrolled in New Jersey colleges. Perhaps they were more inclined to move around, many having come to New Jersey to take advantage of an alternate way to try out teaching not available elsewhere at the time. During these 11 years, the 13 CB teachers collectively taught all but one year of their 114 years in New Jersey. In contrast, the 12 AR teachers taught 15 of their 71 years in other states or overseas, for approximately 21% of their collective teaching years.

Retention also was related to career and life cycles. Three parents who entered teaching in their 30s—two AR and one CB—were the only elementary teachers remaining in the classroom (and at the same school) at the 11th year. Two younger elementary teachers, who had taught seven and eight years, were taking maternity breaks. The parents had this stage behind them.

Age was even more predictive of retention than entry route. Of the eight teachers who entered teaching when they were 33–46 years old, 75% were teaching in the 11th year compared with 41% who were 22–26 years old when they entered teaching. Six of these older teachers were already parents and may have been more dependent on a steady income. They may have made a more informed career choice and/or felt that their years of career exploration and indecision were behind them. They may have been more careful in accepting particular jobs, expecting or desiring to teach for the rest of their career. The relationship between older age at entry and longer retention is not unique (Donaldson, 2012). Although the AR provides a pathway for older adults, attracting non–college graduates to CB programs is another promising source of long-term teachers.

In terms of career moves, two CB teachers used teaching as a “stepping stone” into school administration in their 10th and 11th year. In contrast, no AR teachers made such career moves within schools; rather, their moves were to nonschool education-related jobs or noneducation jobs. In looking at retention more broadly, all 13 CB exemplars (100%) were K–12 education oriented after 11 years, as were 67% of the AR exemplar teachers.

Those who stayed as teachers. The exemplars teaching in their 11th year—62% of the CB and 42% of the AR teachers—made improving their own teaching their major focus. These “career teachers” had a passion for teaching, a commitment to their students, and a belief that classroom teachers could make a difference.

The CB exemplars generally had easier and more consistent development as teachers when they assumed responsibility for their own classroom as compared with most of the AR teachers, who had an abbreviated preparation. CB teachers more consistently demonstrated adaptive expertise, articulated more consistent views of teaching and learning, and had a broader range of teaching approaches. They tended to have more sophisticated views of curriculum and their roles in constructing curriculum with their students. With a few exceptions, they entered the classroom trying to adapt content in support of student learning and developed more learner-centered approaches in their first three years.

Although these differences in trajectory might have been predicted by the length and depth of their preparation, there were some other patterns for teachers who stayed in the classroom beyond the first few years. CB teachers, who more consistently and immediately focused their teaching on student learning and creating integrated learning environments, were less focused on issues of diversity and social justice compared with some of the AR teachers, particularly the teachers of color. Although their classroom management and teaching skills often lagged behind their aspirations of creating curriculum to promote societal change, these AR teachers did bring a different dimension to their teaching. Interestingly, their strong subject matter preparation was a mixed blessing. For some, it became problematic as they struggled to translate their passion into teaching, and for others, it helped them adapt the curriculum to meet the needs of diverse learners.

By the 11th year, inspired and skilled teachers had emerged from both preparation routes. With few exceptions, these teachers had developed adaptive expertise and exhibited more complex, layered, “messier,” and idiosyncratic views of curriculum and their roles. They had a broad range of approaches, reflecting their growing need for self-critique and curriculum change to meet their students’ needs. Although subject matter remained important, as their careers developed, what became most important was “how their students inhabited and breathed life into their subjects” (see Sawyer, 2017).

Although clearly starting in a different place in terms of their professional knowledge, skills, and dispositions, AR teachers took advantage of a variety of opportunities to develop professionally over the 11 years, as did their CB peers. Formal options included advanced degrees, specialized certificates, workshops, seminars, and conferences. They also engaged in other activities in their schools, such as leading staff development, committee work, mentoring, and working with student teachers. More informal options included personal research and reading, action research, partnering and collaborating with peers, and personal reflection. For some AR teachers, their involvement might have been to compensate for their brief preparation; for others, it was for exploration, either to develop classroom skills or to prepare to leave the classroom (see Rutter, 2017).

Regardless of motivation, those who stayed were actively pursuing professional growth. After 11 years, it was hard to distinguish CB and AR teachers. Interestingly, it was the middle school teachers, regardless of preparation route, who had more opportunities to participate in a variety of collaborative, development, and leadership work, called for by many of the reform initiatives of the past 25 years.

IMPLICATIONS

There have always been alternate ways to staff our nation’s public schools when the supply of teachers is lacking. The plethora of alternate pathways today have roots in efforts to reform CB teacher education and to formalize AR paths in the mid-1980s to deal with issues of quality and quantity and break the cycle of just filling classrooms with “warm bodies.” Looking back at this early state effort, which still exists, through the experiences of this cohort of elementary, secondary English, and secondary math teachers gives us a unique opportunity to reflect on implications for teacher education.

With hindsight, the fears and promises of the politically contentious early years of formalized AR programs have moderated. CB teacher education, rather than disappearing, has become stronger. Alternate routes, although meeting some objectives, are not the panacea that some advocates had hoped. CB and AR programs are less sharply delineated as features of pathways to teaching are modified, requirements added, and providers diversified. And the research community, acknowledging the nonmonolithic nature of both CB and AR programs, has changed focus from a “horse-race” mentality to trying to understand the strengths and weaknesses of specific CB and AR programs, with the aim of improving the preparation of teachers and, ultimately, student learning.

Research can focus improvement efforts but also challenge assumptions. In this study, there were surprises. The AR teachers were not all novices; in this early cohort, one third had teaching experience at the grade and subject area in which they were doing their Phase 1 teaching. Although the AR, as planned, attracted older prospective teachers, the average age of the CB graduates entering teaching was 25, not the typical 21–22-year-olds. There were older adults who entered CB programs to obtain or finish their undergraduate degrees or who chose to enter CB programs rather than the AR. Age of entry into teaching was more predictive of retention than type of program. Graduation from a highly competitive college did not ensure that AR teachers would make it through the first year or stay in teaching. The CB graduates brought less conventional ideas of teaching, learning, and curriculum to the schools rather than the nontraditional AR recruits. In this early cohort, unexpectedly, most of the AR teachers found jobs in nonurban or non-low-wealth schools. Many of those who started teaching in urban or low-wealth districts, like teachers elsewhere, fled to more desirable positions over time. The AR program was more successful in attracting and retaining males than minorities. Although a higher proportion of AR teachers were minorities, the CB programs still supplied and retained more minority teachers for New Jersey schools than the AR program.

A major finding of this study is that neither preparation route should be viewed generically; there were important differences related to level (i.e., elementary, secondary) and to subject (i.e., English, math). The teachers who were least like their CB counterparts were the ARM teachers. They were the AR teachers who struggled the most and left sooner. In English, the existence of the AR exacerbated a tight job market, making it harder for CBE teachers to find jobs after college. ARE teachers, with generic certification, found jobs, but in other subjects and elementary teaching, which may have contributed to their higher attrition rates. Understanding such differences might have led to changes in the recruitment, preparation, and support for ARM teachers, and prioritizing subjects in which AR programs existed. The state, after all, was investing in CB programs at its public colleges/universities where most of the CB teachers were prepared. The AR was meant to supplement CB programs, not undermine the graduates of state-funded college programs, as was unexpectedly happening to CBE graduates and, in turn, the ARE teachers.

The English teachers provide an important lesson. Unfortunately, 20 years later, the initial polarization threatened to erupt again as economic conditions led to competition among the various routes when AR teachers were used to replace CB teachers rather than supplement them (Heilig & Jez, 2010; Toppo, 2009) Some states, such as California, use alternate options only in areas and subjects where immediate shortages exist (Chin & Young, 2007). Faced with teacher layoffs and minimal hiring needs, Los Angeles decided not to hire TFA teachers in 2009. According to Feistritzer (2009), this is what should happen because AR teachers are not needed when there is no market demand for teachers.

In contrast, other localities have replaced laid off teachers with less expensive AR substitutes. New York City froze teacher hiring for the 2009–2010 school year to force the hiring of unassigned teachers first. An exception was made, however, for TFA and for a reduced number of Teaching Fellows. Hence, AR teachers, who had not even started their fast-track summer training and for whom the city had to pay a recruiting fee to TFA, were given preference over fully prepared, certified teachers. In implementing mandated assessments in 2014, New York State required CB graduates to pass the edTPA during student teaching to obtain initial certification, whereas AR teachers are tested at the end of two years in their own classrooms. Such policies have the potential to inflame the political controversy that surrounded AR programs 30 years ago, and sidetrack policy makers and researchers from the common goal of preparing better teachers for all our schools.

It is hoped that recent events are a temporary aberration and not the return of polarized controversy. Our longitudinal study of learning to teach in New Jersey through CB and AR programs portrays the recruitment, preparation, and retention trade-offs that have arisen from one state’s implementation of an AR program. These trade-offs can provide a constructive frame rather than a return to the horse race mentality that dominated earlier years of alternate routes to teaching.

TEACHER EDUCATION AND ALTERNATE ROUTE TRADE-OFFS

Both CB and AR programs have provided considerable numbers of teachers for New Jersey schools. Between 1985 and 2005, public schools hired 23,800 teachers, about 32% through the AR and 68% through CB programs, according to the State Department of Education (Barclay et al., 2008). Over time, the AR option has grown as a source of New Jersey teachers. From 2010 to 2012, there were 20,757 teachers granted certificates of eligibility to teach in New Jersey, roughly half through the AR program and half through CB programs. In 2013–2014, 47% of these CB and 45% of these AR applicants were employed in New Jersey public schools (New Jersey State Department of Education, 2014a, 2014b). This relatively high proportion of teachers entering through the AR may be a reflection of a clearly structured statewide AR program, which puts admissions to the program in the hands of school districts that have had trouble finding suitable candidates. New Jersey may have a relatively high proportion of harder-to-staff schools. It may also be related to the fact that New Jersey, unlike neighboring New York, does not require a master’s degree for professional certification. Hence, there has not been a large number of institutions that offer one-year master’s degree programs, which essentially are an alternate to undergraduate teacher education programs. Hence, the existence of the AR, with its “earn while you learn” feature, is an attractive option for those who do not decide to become teachers as undergraduates and essentially can “try out” the job before making the investment in a graduate degree. However, that one fifth of this early AR cohort, at the end of six years, would have chosen to enter a graduate program instead indicates that it may be another viable pathway to attract teachers. In fact, as of 2003, New Jersey established an MAT option for the AR program.

The recent nationwide decline in newly prepared teachers is evident in New Jersey and has had an impact on the share of teachers prepared through AR and CB programs. Annual reports for 2015 and 2016 (New Jersey Department of Education, 2016) indicate a decline in the numbers of teachers prepared, particularly through the AR option. Compared with 2010–2012, in 2012–2014, the number of teachers prepared in CB programs (8,099) declined 21%, whereas the number of teachers prepared in AR programs (3,132) declined 70%. With less actual need for teachers, school districts hired fewer AR candidates, decreasing the numbers prepared in the AR. With less perceived need for teachers, prospective teachers were less likely to choose to become teachers, also decreasing the numbers prepared in CB programs. Thus, the feature of the AR program requiring candidates to be hired first appears to provide direct and more immediate market control. That 64% of the CB teachers and 59% of the AR teachers prepared in 2012–2014 were employed by New Jersey schools in 2015–2016 indicates that more teachers are still prepared than end up teaching in New Jersey. However, these percentages are considerably higher than the 47% and 45% employed two years earlier.

The AR program does meet the immediate and changing needs of supplying some teachers for public schools in a quicker way than an undergraduate or one-year master’s program. It has attracted more and different people to teaching. The question remains whether the trade-offs are worth it—in the short term and the long term. Two trade-offs involve the impact of an abbreviated preparation time on initial teaching performance and the increased teacher turnover associated with the AR program.

Recruitment Versus  Preparation

Generally, novice teachers who have had a CB program consisting of subject and elementary- or secondary-specific coursework, practicums, and student teaching are in a better position to assume full responsibility for classroom teaching their first year than AR candidates whose coursework and practicum preparation was largely completed (and sometimes started) while on the job (Darling-Hammond, 2008b; Grossman & Loeb, 2008; Headden, 2014; Heilig & Jez, 2010; Zeichner & Conklin, 2005). Among our exemplar teachers, we saw that the CB teachers more consistently demonstrated adaptive expertise, had more consistent views of teaching and learning, and had a broader range of teaching approaches. They tended to have more sophisticated views of curriculum and their roles in enacting curriculum with their students. With a few exceptions, they entered the classroom trying to adapt content in support of student learning and developed more learner-centered approaches in their first three years.

A 2008 evaluation of New Jersey’s AR program (Barclay et al., 2008) indicates that these initial differences continued to exist 20 years later. Although administrators were complimentary about the subject matter knowledge—content and life experiences—and the personal qualities of AR teachers, the vast majority agreed that the AR teachers needed more support than other novice teachers, particularly with classroom management, instructional planning, understanding of human development, accommodating students with special needs, and meeting the needs of diverse students. They were more positive about the benefits the AR teachers brought to secondary teaching and less positive about the performance of AR teachers at the elementary level (Ing & Loeb, 2008).

Additional requirements have been added to the AR program (New Jersey Department of Education, 2014a). Beyond the 200 hours of formal instruction required of all AR teachers, AREL teachers must take an additional 90 hours of formal instruction in language arts/literacy and mathematics. Since 2009, before getting their certificate of eligibility (CE) enabling a school district to hire them, all AR candidates “must demonstrate knowledge of basic pedagogical skills appropriate to their area of endorsement” by completing 24 hours of study at a state-approved provider or a New Jersey college.

In our study, before these changes were implemented, initial differences in classroom performance seem to lessen as AR teachers who stayed in the classroom gained experience and partook of opportunities for professional development, as did their CB colleagues, whose longer and more intensive preparation program gave them a head start. In essence, the front-loading of their preparation gave the CB graduates a jump-start on learning how to teach, but over time, their preparation differences seemed less consequential. So one of the trade-offs is whether the initial performance differential is acceptable given that the AR did attract people to teach in New Jersey public schools who might not have done so otherwise, either because they were not interested in teaching as undergraduates or their colleges did not have teacher education programs.

At one level, this question is one of values—whether it is acceptable that some children are taught by less prepared teachers, particularly if they are more likely to be minorities and from lower social classes. At another level, research on the nature of performance differences, when they dissipate, factors that could accelerate the learning of AR teachers, and the impact on student outcomes—not just achievement—would help decide whether this trade-off during the initial years of teaching is worth it.

In the meantime, there are structures in place, if appropriately and consistently used, that could help minimize some of the differences in initial performance. Since 2003, all novice AR and CB teachers in New Jersey are required to participate in a 30-week mentoring program during their first year of teaching, with AR teachers participating in an additional 20 days of preservice training and support. Like earlier issues with frequency and quality of supervision, a 2008 evaluation of the AR program (Barclay et al., 2008) indicates that both components could use more attention. The mentoring program could be a critical factor in lessening the impact of their abbreviated preparation. Good professional development programs tailored to the needs of individual teachers in their particular contexts should take teachers “from where they are,” quite different from the more common “one-size fits all” in-service or staff development programs.  As we have seen, there is much variability among teachers regardless of their entry route into teaching. Hence, as others have found, the impact of preservice and in-service teacher education depends on the interaction between the program, the teacher, the school context, and the supports for growth for the particular teacher (Chin & Young, 2007; Johnson, 2004). Regardless of where they are in their life cycle when they enter, most teachers have the potential for growth, given supportive opportunities—and all children can benefit from teachers who have not stopped learning.

Recruitment Versus Retention

Besides supplying needed teachers, New Jersey’s AR program aimed to recruit teachers who were different than the young White females who populated CB programs in New Jersey and elsewhere. In this early cohort, the most heralded success of the program has been to diversify the teaching force in terms of race/ethnicity, a particularly critical consequence in face of the growing diversity of the student population. Other aspects of diversity—gender, age, and experience—were also seen as positive outcomes in this early cohort, whereas academic differences were less conclusive.

Unquestionably, the AR is a successful recruitment strategy; however, as seen in the first six years for the whole cohort and 11 years for the exemplar teachers, the AR teacher retention rate was lower than for CB teachers. Not only were they more likely to leave teaching, but they also moved “in and out” of teaching and specific jobs more often. Teacher turnover affects the stability of schools, educational reform, student learning, district budgets, and teacher shortages (Headden, 2014; Ingersoll, Merrill, & Stuckey, 2014; Kini & Podolsky, 2016; Ronfeldt, Loeb, Wyckoff, 2013). Experienced teachers have been found to have positive effects on student achievement, on attendance, and on colleagues (Carver-Thomas & Darling-Hammond, 2017; Sawchuk, 2015). The question is whether the value added by a more diverse teaching staff is worth the consequences of increased teacher turnover.

How much of the AR retention problem is related to their short preparation, mismatch between placement and interests, less than promised initial supervision, movement to more desirable teaching jobs, the low cost of entry and exit, unrealistic expectations, or the constellation of personal factors related to postcollege career choices and exploration is unknown. Whatever the reasons, AR teachers did not stay in teaching as long, nor move as frequently to nonclassroom jobs in school districts. At the end of 11 years, 100% of the CB teachers were still committed to teaching or other related K–12 education careers compared with 67% of the AR teachers. Whether their longer term commitment is a result of their more extensive preparation, a reflection of their initial commitment, or some combination, we cannot be sure. However, it should be noted that both routes produced long-term teachers who were committed to their students and schools. Some of these long-term teachers might not have been there if the AR program had not existed.

And even those who left classroom teaching to pursue jobs in other fields undoubtedly took something with them about the challenges of classroom teaching that may affect their lives as parents and influence their attitudes and actions toward public education in the future. Depending on what they took from their experience, this could be beneficial or not. But it is an outcome worth thinking about.

Some experts now believe that retention, rather than recruitment, is the more critical issue in staffing our public schools (Costigan, 2005; Ingersoll, 2002, 2008). Issues of retention play out on several levels—who enters, how teachers are prepared before teaching, how they are matched with appropriate jobs, how they are supported, particularly during the early years, and whether they have opportunities for professional growth. Although conditions of teaching affect the retention of teachers generally, this study indicates that issues of retention vary for different subjects and different levels, different preparation routes, and teachers of different ages and genders. It also makes clear that reports of retention rates at any one point in time mask the individual journeys of teachers who take “breaks” from teaching for various reasons. While state and district polices obviously have an effect on retention, what happens at the local school and in their personal lives was critical for many of the exemplar teachers when making decisions about staying or leaving teaching.

WHAT’S NEXT?

Neither AR nor CB teacher education is going to disappear. Unless the conditions of teaching and schools change drastically, we are going to need multiple pathways to provide an adequate number of teachers for our public schools. Usually, AR programs are an improvement over “emergency licensing,” which affects schools for decades. And, in many cases, they have diversified (e.g., race/ethnicity, gender, age) public school teaching.

The professionalization versus deregulation battle in teacher education has moderated over the 30 years since New Jersey’s AR was instituted. The emerging policy focus on educational equity, embodied in NCLB, has placed an emphasis on having fully certified teachers in all public schools in an attempt to close the achievement gap. Although there have been various attempts to modify the definition of fully certified to include AR teachers, in essence, teacher certification has become the goal. The lines between those prepared in CB and AR programs have blurred; in many cases, most end up being CB graduates. For instance, in New York State, which has long required a master’s degree for certification, all AR teachers must now enroll in master’s programs during their first year of teaching. Even though New Jersey does not require a master’s degree for full certification, it now has an MAT option for AR teachers, and the primary provider of staff for the regional training centers is New Jersey colleges.

Also blurring the lines is the fact that many AR programs are now situated in colleges/universities. They operate in an institutional context that shapes their organization and practices, making CB and AR programs look quite similar (Grossman & McDonald, 2008). State policies differ, however, as do local and regional supply and demand issues, so it is likely the general trend for higher requirements for public school teachers will ebb and flow, state by state. However, what has not changed is that teachers who have met relatively higher professional standards are more likely to be found in public schools serving the most advantaged, rather than the least advantaged, students.

The equity imperative demands attention. AR teachers disproportionately enter teaching in public schools serving our most vulnerable students. Hence, it is imperative that we work to improve preparation, on-the-job support, and retention. In a 2008 evaluation of New Jersey’s AR (Barclay et al., 2008), when administrators were asked how the AR could be improved, the most consistent theme was adding a student teaching component, giving these teachers more “practice time” before assuming responsibility for their own classes.

Teacher educators have long been critical of the abbreviated content of the AR program, particularly the lack of pedagogical content knowledge inherent in its generic approach and the absence of graduated field experience and supervised, full-time student teaching. Supervised field experiences and student teaching permit prospective teachers to experiment and learn from mistakes in a safe environment, and provide the guidance and modeling of experienced cooperating teachers and supervisors. Theoretically, prospective teachers in such an environment learn to be reflective practitioners; they learn how to learn from experience rather than become overwhelmed by day-to-day survival. Developing a stance toward reflective practice can hopefully minimize the survival techniques of the first year that stay with teachers long past their usefulness (Zumwalt, 1984). Formal teacher preparation—content and fieldwork—helps novice teachers make sense of experience (Grossman, 1990). AR teachers have indicated they need many of the very things offered in CB programs (Cuddapah & Burtin, 2012; Hopkins, 2008). Analysis of recent SASS data indicates that teachers, regardless of preparation route, who have more methods-related coursework and practice teaching were more likely to stay in teaching (Ingersoll, Merrill, & May, 2014; Ronfeldt, Schwartz, & Jacob, 2014). The 2014 NCTQ report concluded that AR programs need “to provide some period of real teaching in a real classroom in advance of beginning the school year” (Greenberg et al., 2014, p. 79). Thus, it is not surprising that New Jersey administrators suggest that AR teachers have something resembling student teaching before assuming responsibility for their own classroom.

A promising model involving such change is teacher residency programs, which have similarities to a rejected proposal considered by the Newman Commission (Newman, 1981) in New Jersey that required internships before AR teachers assumed responsibility as the teacher of record. Teacher residency programs have been advanced as another way of recruiting, preparing, and retaining teachers for high-needs schools (Darling-Hammond, 2008a; Guha, Hyler, & Darling-Hammond, 2016; Headden, 2014, Solomon, 2009). These programs attempt to combine the attractive features of AR and CB programs. They provide prospective teachers a full-year residency coteaching with experienced, successful teachers in high-needs schools while taking coursework toward their master’s degree and teaching credential. The resident teachers get tuition and a stipend as they prepare to assume full-time teaching jobs in these schools the following year. In return, they make a commitment to teach several years in high-need schools. The teacher residency model was given a big boost in the 2008 reauthorization of the Higher Education Act, which funded a competitive grant program. Since then, residency-type programs have gained traction. For example, an alternate provider, Relay Graduate School of Education, has developed a program with a charter network for their graduates. It provides a stipend-paid, postcollege year of student teaching and course-taking, followed by a transition year of salary-paid teaching and courses (Ash, 2014). Even TFA has decided to experiment with providing “a year of upfront training to a subset” of recruits starting in 2015 (Sawchuk, 2014). Early results from the Boston Teacher Residency (Papay, West, Fullerton, & Kane, 2012) and others across the country (Guha et al., 2016) are promising. Perhaps melding of AR and CB elements will help fulfill promises of both approaches to provide well-prepared and effective teachers for all our nation’s public schools.  

Whatever options emerge, it is clear there are no simple solutions (Zumwalt, 1996). Changing conditions cyclically affect the supply and demand of teachers, particularly for harder-to-staff urban and rural schools. For a variety of reasons, as a nation we find ourselves once again facing a shortage of teachers, as we did when alternate routes to teaching were established in the mid-1980s. We have learned much in the last 30 years to help guide us through the emerging shortages and fight the temptation to develop “easy fix” state and local alternate options to provide classroom teachers. Alternate pathways can help address endemic issues in teacher supply, but “we need to be developing professionally defensible options for college graduates who did not make the decision to teach in their undergraduate years that do not undermine current efforts to upgrade and professionalize teaching” (Zumwalt, 1996, p. 42). We should be focusing on what kinds of preparation and support new and experienced teachers need to make sure all children have fully prepared teachers, no matter where their public schools are located.

References

Adams, C. J., & Dial, M. (1993). Teacher survival: A Cox regression analysis. Education and Urban Society, 26(1), 90–99.

Ash, K. (2014, February 26). Charter network finds new teachers among its graduates. Education Week, p. 7.

Barclay, R., Feistritzer, E., Grip, R., Haar, C., Seator, G., Sherman, S., & Stone, M. (2008). The New Jersey Alternate Route Program: An analysis of the perspectives of alternate route teachers, alternate route instructors and alternate route mentors. Retrieved August 18, 2008, from the New Jersey Department of Education website: http://www.state.nj.us/education/educators

Borman, G. D., & Dowling, N. M. (2008). Teacher attrition and retention: A meta-analytic and narrative review of the research. Review of Educational Research, 78(3), 367–409.

Boyd, D., Grossman, P., Hammerness, K., Lankford, H., Loch, S., Ronfeldt, M., & Wyckoff, J. (2012). Recruiting effective math teachers: Evidence from New York City. American Educational Research Journal, 49(6), 1088–1047.

Boyd, D., Grossman, P., Lankford, H., Loeb, S., & Wycoff, J. (2006). How changes in entry requirements alter the teacher workforce and affect student achievement. Education Finance and Policy, 1(2), 176–216.

Boyer, E. L. (1984). Report of a panel on the preparation of beginning teachers. Trenton: New Jersey State Department of Education.

Carter, H., Amrein-Beardsley, A., & Hansen, C. C. (2011). So NOT amazing! Teach for America Corps members’ evaluation of their first semester of their teacher preparation program. Teachers College Record, 113(5), 861–894.

Carver-Thomas, D., & Darling-Hammond, L. (2017). Teacher turnover: Why it matters and what we can do about it. Palo Alto, CA: Learning Policy Institute.

Chin, E., & Young, J. W. (2007). A person-oriented approach to characterizing beginning  teachers in alternative certification programs. Educational Researcher, 36(2), 74–83.

Cochran-Smith, M., & Villegas, A. (with Abrams, L. W., Chavez-Moreno, L. C., Mills, T., & Stern,  R.). (2016). Research on teacher preparation: Charting the landscape of a sprawling field. In D. H. Gitomer & C. A. Bell (Eds.), Handbook of research on teaching (5th ed., pp. 439–547). Washington, DC: American Educational Research Association.

Cochran-Smith, M., & Zeichner, K. (2005). Studying teacher education. The report of the AERA Panel on Teacher Education. Mahwah, NJ: Erlbaum.

Constantine, J., Player, D., Silva, T., Hallgren, K., Grider, M., & Deke, J. (2009). An evaluation of teachers trained through different routes to certification: Final report (NCEE 2009-4043). Washington, DC: National Center for Education Evaluation and Regional Assistance, Institute of Education Sciences, U.S. Department of Education.

Cooperman, S., & Klagholz, L. (1985). New Jersey’s alternate route to certification. Phi Delta Kappan, 66(10), 691–695.

Costigan, A. (2005). A “traditional” alternative route to certification: Narrative research and implications for teacher education and teacher retention. In J. R. Dangel & E. M. Guyton (Eds.), Research on alternative and non-traditional education (pp. 27–38). Teacher Education Yearbook XIII. Association of Teacher Educators. Lanham, MD: Scarecrow Education.

Cuddapah, J. L., & Burtin, A. S. (2012). What all novices need. Educational Leadership, 69(8), 66–69.

Darling-Hammond, L. (1990). Teaching and knowledge: Policy issues posed by alternative certification for teachers. Peabody Journal of Education, 67(3), 123–154.

Darling-Hammond, L. (2008a). A future worthy of teaching for America. Phi Delta Kappan, 89(10), 730–736.

Darling-Hammond, L. (2008b). Knowledge for teaching: What do we know? In M. Cochran-Smith, S. Feiman-Nemser, D. J. McIntyre, & K. E. Demers (Eds.), Handbook of research on teacher education: Enduring questions in changing contexts (3rd ed., pp. 1316–1323). New York, NY: Routledge.

Darling-Hammond, L., Chung, R., & Frelow, F. (2002). Variation in teacher preparation: How well do different pathways prepare teachers to teach? Journal of Teacher Education, 53, 286–302.

Dill, V. (1996). Alternative teacher certification. In J. Sikula (Ed.), Handbook of research on teacher education (2nd ed., pp. 932–960). New York, NY: Macmillan.

Donaldson, M. L. (2012). The promise of older novices: Teach for America teachers’ age of entry and subsequent retention in teaching and schools. Teachers College Record, 114(10), 1–37.

Donaldson, M. L., & Johnson, S. M. (2010). The price of misassignment. The role of teaching assignments in Teach for America teachers’ exit from low income schools and the teaching profession. Educational Evaluation and Policy Analysis, 32(2), 299–323.

Feistritzer, C. E. (2005). Profile of alternate route teachers. Washington, DC: National Center for Education Information.

Feistritzer, C. E. (2009, November 18) The impact of alternate routes: How teacher preparation has changed and why it may need to change more. Education Week, 29(12), 26, 32.

Feistritzer, E., & Haar, C. (2008). Alternate routes to teaching. Upper Saddle River, NJ: Pearson Prentice Hall.

Fenstermacher, G. (1990) The place of alternative certification the education of teachers. Peabody Journal of Education, 67(3), 155–184.

Gansle, K. A., Noell, G. H., & Burns, J. M. (2012). Do student achievement outcomes differ across teacher preparation programs? An analysis of teacher education in Louisiana. Journal of Teacher Education, 63(5), 304–317.

Glazerman, S., Mayer, D., & Decker, P. (2006). Alternative routes to teaching: The impacts of Teach for America on student achievement and other outcomes. Journal of Policy Analysis and Management, 25(1), 75–96.

Gray, L., & Tale, S. (2015). Public school attrition and mobility in the first five years: Results from the first through fifth waves of the 2007-2008 Beginning Teacher Longitudinal Study: First look (NCES 2015-337). U.S. Department of Education. Washington, DC: National Center for Educational Statistics.

Greenberg, J., Walsh, K., & McKee, A. (2014). 2014 Teacher Prep Review. Washington, DC: National Council on Teacher Quality.

Grissmer, D. W., & Kirby, S. N. (1987) Teacher attrition: The uphill climb to staff our nation’s schools. Santa Monica, CA: RAND.

Grissom, J. (2008). But do they stay? Addressing issues of teacher retention through alternative certification. In P. Grossman & S. Loeb (Eds.), Alternative routes to teaching; Mapping the new landscape of teacher education (pp. 129–156). Cambridge, MA: Harvard Education Press.

Grossman, P. (1990) The making of a teacher: Teacher knowledge and teacher education. New York, NY: Teachers College Press.

Grossman, P., & Loeb, S. (Eds.). (2008). Alternative routes to teaching: Mapping the new landscape of teacher education. Cambridge, MA: Harvard Education Press.

Grossman, P., & McDonald, M. (2008). Back to the future: Directions for research in teaching and teacher education. American Educational Research Journal, 45(1), 184–205.

Guha, R., Hyler, M. E., & Darling-Hammond, L. (2016). The teachers residency approach: An innovative model for preparing teachers. Palo Alto, CA: Learning Policy Institute.

Haberman, M., & Post, L. (1998). Teachers for multicultural schools. Theory Into Practice, 37(2), 97–104.

Hammerness, K., & Reisinger, M. (2008). Who goes into early-entry programs? In P. Grossman & S. Loeb (Eds.), Alternative routes to teaching: Mapping the new landscape of teacher education (pp. 31–64). Cambridge, MA: Harvard Education Press.

Hawley, W. D. (1990). The theory and practice of alternative certification: Implications for the improvement of teaching. Peabody Journal of Education, 67(3), 3–34.

Headden, S. (2014). Beginners in the classroom: What changing demographics of teaching mean for schools, students, and society. Stanford, CA: Carnegie Foundation for the Advancement of Teaching.

Heilig, J. V., & Jez, S. J. (2010). Teach for America: A review of evidence. Boulder, CO & Tempe, AZ: Education and the Public Interest Center & Education Policy Unit. Retrieved from http//epicpolicy.org/publication/teach-for-america

Henke, R. R., Chen, X., & Geis, S. (2000). Progress through the teacher pipeline: 1992-93 college graduates and elementary/secondary school teaching as of 1997. Washington, DC: National Center for Educational Statistics, U.S. Department of Education.

Hopkins, M. (2008). Training the next teachers for America: A proposal for reconceptualizing Teach for America. Phi Delta Kappan, 89(10), 721–725.

Humphrey, D., & Wechsler, M. (2007). Insights into alternative certification: Initial findings from a national study. Teachers College Record, 109(3), 483–530.

Humphrey, D., & Wechsler, M. (2008). Getting beyond the label: What characterizes alternative certification programs? In P. Grossman & S. Loeb (Eds.), Alternative routes to teaching: Mapping the new landscape of teacher education (pp. 65–98). Cambridge, MA: Harvard Education Press.

Humphrey, D., Wechsler, M., & Hough, H. (2006). Characteristics of effective alternative certification programs. Teachers College Record, 110(1), 139–146.

Ing, M., & Loeb, S. (2008). Assessing the effectiveness of teachers from different pathways: Issues and results. In P. Grossman & S. Loeb (Eds.), Alternate routes to teaching: Mapping the new landscape of teacher education (pp. 157–186). Cambridge, MA: Harvard Education Press.

Ingersoll, R. M. (2002). The teacher shortage: A case of wrong diagnosis and wrong prescription. NASSP Bulletin, 86, 16–31.

Ingersoll, R. M. (2008). The teacher quality problem. In M. Cochran-Smith, S. Feiman-Nemser, & D. J. McIntyre (Eds.), Handbook of research on teacher education: Enduring questions in changing contexts (pp. 527–533). New York, NY: Routledge, Taylor & Francis Group, and the Association of Teacher Educators.

Ingersoll, R., Merrill, L., & May, H. (2012). Retaining teachers: How preparation matters. Educational Leadership, 69(8), 30–34.

Ingersoll, R., Merrill, L., & May, H. (2014). What are the effects of teacher education and preparation on beginning teacher attrition? CPRE Report (#RR-82) Philadelphia, PA: Consortium for Policy Research in Education.

Ingersoll, R., Merrill, L., & Stuckey, D. (2014). Transformation of the teaching force, updated April 2014. CPRE Report No. RR-80. Philadelphia: Consortium for Policy Research in Education, University of Pennsylvania.

Johnson, S. M. (2004). Finders and keepers. San Francisco, CA: Jossey-Bass.

Johnson, S. M., Birkeland, S. E., & Peske, H. G. (2005). Life in the fast track: How states seek to balance incentives and quality in alternative teacher certification programs. Educational Policy, 19(1), 63–89.

Kelly, S., & Northrup, L. (2015). Early career outcomes for the “best and the brightest”: Selectivity, satisfaction, and attrition in the Beginning Teacher Longitudinal Survey. American Educational Research Journal, 52(4), 624–656.

Kini, T., & Podolsky, A. (2016). Does teaching experience increase effectiveness? A review of the research. Palo Alto, CA: Learning Policy Institute.

Kirby, S. N., Darling-Hammond, L., & Hudson, L. (1989). Nontraditional recruits to math and science teaching. Educational Evaluation and Policy Analysis, 11(3), 301–323.

Lortie, D. C. (1975). School teacher: A sociological study. Chicago, IL: University of Chicago Press.

Murnane, R. J., Singer, J. D., Willett, J. B., Kemple, J. J., & Olsen, R. J. (1991). Who will teach? Policies that matter. Cambridge, MA: Harvard University Press.

National Education Association. (1997). Status of the American public school teacher, 1995–96. Washington, DC: Author.

Natriello, G., & Zumwalt, K. (1992). Challenges to an alternative route to teacher education. In A. Lieberman (Ed.), The changing contexts of teaching (pp. 59–78). National Society for the Study of Education Yearbook. Chicago, IL: University of Chicago Press.

New Jersey State Department of Education. (2014a). Educator preparation provider 2014 annual state-wide overview: Alternate route. Trenton: New Jersey State Department of Education. Retrieved from State.nj.us/education/educators/rpr/preparation/providers/2014/Alternate Route/EPP.pdf

New Jersey State Department of Education. (2014b). Educator preparation provider 2014 annual state-wide overview: Traditional route. Trenton: New Jersey State Department of Education. Retrieved from http://state.nj.us/education/educators/rpr/preparation/providers/2014/AlternateRoute/EPP.pdf

New Jersey Department of Education. (2016). Educator preparation provider performance reports 2015 and 2016. Retrieved from http://www.state.nj.us/education/educators/rpr/preparation/providers/

Newman, D. F. (1981). Report of the Commission to Study Teacher Preparation Programs in New Jersey Colleges. Trenton: New Jersey Department of Education.

Papay, J., West, M., Fullerton, J., & Kane, T. (2012). Does an urban residency increase student achievement? Early evidence from Boston. Educational Evaluation and Policy Analysis 34(3), 413–434.

Paris, C. (1993). Teacher agency and curriculum making in classrooms. New York, NY:  Teachers College Press.

Peske, H. G., Liu, E., Johnson, S. M., Kauffman, D., & Kardos, S. (2001). The new generation of teachers: Changing conceptions of a career in teaching. Phi Delta Kappan, 83(4), 304–311.

Randi, J. (2017). Exploring options: From preparation to placements. Teachers College Record, 119(14).

Rinke, C. R. (2008). Understanding teachers’ careers: Linking professional life to professional path. Educational Research Review, 3(1), 1–13.

Ronfeldt, M., Loeb, S., & Wyckoff, J. (2013). How teacher turnover harms student achievement. American Educational Research Journal, 50(1), 4–36.

Ronfeldt, M., Schwartz, N., & Jacob, B. A. (2014). Does pre-service preparation matter? Examining an old question in new ways. Teachers College Record, 116(10), 1–46.

Rutter, A. (2017). Teacher choices: Teacher development. Teachers College Record, 119(14).

Sawchuk, S. (2014, March 12). Teach for America targets retention, longer preparation. Education Week, p. 6.

Sawchuk, S. (2015). New studies find that, for teachers, experience really does matter. Education Week. Retrieved from https://www.edweek.org/ew/articles/2015/03/25/new-studies-find-that-for-teachers-experience.html

Sawyer, R. (2017). Curriculum deliberations over time: A nexus of teacher dilemmas, questions, experimentation, and agency. Teachers College Record, 119(14).

Snyder, J., Bolin, F., & Zumwalt, K. (1992). Curriculum implementation. In P. W. Jackson (Ed.), Handbook of research on curriculum (pp. 402–435) New York, NY: Macmillan.

Solomon, J. (2009). The Boston Teacher Residency: District-based teacher education. Journal of Teacher Education, 60(5), 478–488.

Stoddart, T. (1990). Los Angeles Unified School District intern program: Recruiting and preparing teachers for an urban context. Peabody Journal of Education, 67(3), 84–122.

Tamir, F. (2013). What keeps teachers in and what drives them out: How urban public, Catholic, and Jewish day schools affect beginning teachers’ careers. Teachers College Record, 115(6), 1–35.

Toppo, G. (2009, July 7). Teach for America: Elite corps or costing older teachers jobs? USA Today. Retrieved from http://www.usatoday.com/news/education/2009-07-29-teach-for-america_N.htm

Viadero, D. (2010, May 10). Draw called over routes to teaching. Education Week, 29(31), 1–24.

Wilson, S., Floden, R., & Ferrini-Mundy, J. (2001). Teacher preparation research: Current knowledge, gaps and recommendations. Seattle, WA: Center for the Study of Teaching and Policy.

Wilson, S., Floden, R., & Ferrini-Mundy, J. (2002). Addendum to teacher preparation research: Current knowledge, gaps and recommendations. Seattle, WA: Center for the Study of Teaching and Policy.

Wilson, S., & Tamir, E. (2008). The evolving field of teacher education: How understanding challenge(r)s might improve the preparation of teachers. In M. Cochran-Smith, S. Feiman-Nemser, D. J. McIntyre, & K. E. Demers (Eds.), Handbook of research on teacher education: Enduring questions in changing contexts (3rd ed., pp. 908–935). New York, NY: Routledge.

Xu, Z., Hannaway, J., & Taylor, C. (2011). Making a difference? The effects of Teach for America in high school. Journal of Policy Analysis and Management, 30(3) 447–469.

Zeichner, K. (2005, September 16). The characteristics of candidates of alternative teacher certification programs. Paper presented at an invitational forum on alternative certification research sponsored by the Institute for Educational Sciences, Washington, DC.

Zeichner, K. M. (2009). Reflections of a university-based teacher educator on the future of college- and university-based teacher education. In K. M. Zeichner (Ed.), Teacher education and the struggle for social justice (pp. 146–160). New York, NY: Routledge.

Zeichner, K. (2016). Independent teacher education programs: Apocryphal claims, illusory evidence. Boulder, CO: National Policy Center.

Zeichner, K., & Conklin, H. (2005). Research on teacher education programs. In M. Cochran-Smith & K. Zeichner (Eds.), Studying teacher education: The report of the AERA Panel on Research and Teacher Education (pp. 645–735). Mahwah, NJ: Erlbaum.

Zeichner, K., & Conklin, H. G. (2008). Teacher education programs as sites for teacher preparation. In M. Cochran-Smith, S. Feiman-Nemser, D. J. McIntyre, & K. E. Demers (Eds.), Handbook of research on teacher education: Enduring questions in changing contexts (3rd ed., pp. 269–289). Association of Teacher Educators. New York, NY: Routledge.

Zeichner, K., & Gore, J. (1990). Teacher socialization. In R. W. Houston (Ed.), Handbook of research on teacher education (pp. 329–348). New York, NY: Macmillan.

Zeichner, K., & Schulte, A. (2001). What we know and don’t know from peer-reviewed research about alternative teacher certification programs. Journal of Teacher Education, 52(4), 266–282.

Zumwalt, K. (1984). Teachers and mothers: Facing new beginnings. Teachers College Record, 86(1), 138–155.

Zumwalt, K. (1991). Alternate routes to teaching: Three alternative approaches. Journal of Teacher Education, 42(2), 83–92.

Zumwalt, K. (1996). Simple answers: Alternative teacher education. Educational Researcher, 25(8), 40–42.

Zumwalt, K., & Craig, E. (2005a). Teachers’ characteristics: Research on the demographic profile. In M. Cochran-Smith & K. Zeichner (Eds.), Studying teacher education: The report of the AERA Panel on Teacher Education (pp. 111–156) Mahwah, NJ: Erlbaum.

Zumwalt, K., & Craig, E. (2005b). Teachers’ characteristics: Research on indicators of quality. In M. Cochran-Smith & K. Zeichner (Eds.), Studying teacher education: The report of the AERA Panel on Teacher Education (pp. 157–260). Mahwah, NJ: Erlbaum.

Zumwalt, K., Natriello, G., Randi, J., Rutter, A., & Sawyer, R. (2017). Recruitment, preparation, placement and retention of alternate route and college-prepared teachers: An early study of a New Jersey initiative. Teachers College Record, 119(14).