Alternative Certification: A Review of Theory and Research

RESEARCH AND EVALUATION

Early Studies One of the first evaluations of AC training reviewed a program in Houston and reported three positive outcomes (Goebel, 1986). First, survey results of principals and administrators showed interns' abilities were perceived as equal to those of certified first-year teachers. Second, a high proportion of interns planned to return the following year. And third, student achievement scores were similar for interns and certified, first-year teachers. However, another study of alternative certification in Dallas showed mixed results, finding that, while the program would not have a significant impact on the teacher shortage, it did produce some qualified teachers in shortage areas (Lutz & Hutton, 1989). In addition, results showed that sufficient screening of applicants contributed to intern quality and that the participation of principals, supervising teachers, and advisors was crucial.

An early review of studies of AC programs provided further positive evidence of their effectiveness. Darling-Hammond, Hudson, and Kirby (1989) reviewed 64 math and science certification programs and found that these programs were successful in increasing the number of minorities and the number of female teachers of math and science. This review also found that a substantial portion of the participants in these programs had prior teaching experience of some kind, in contrast to the expectation that AC programs would recruit primarily candidates from outside the field of education.

Further empirical evidence to support AC programs was found in a Texas study that examined the GPAs of traditional and AC teaching candidates (Brown, Edington, Spencer, & Tinafero, 1989). This study found that the GPAs of alternative certification candidates were actually higher than traditionally prepared candidates or teachers with emergency permits, and that classroom performance of teachers in these three groups was similar.

A study of one of the first AC programs—in New Jersey—looked at recruitment, placement, and retention of alternatively and traditionally prepared teachers (Natriello & Zumwalt, 1992). The researchers used survey data collected over several years, and found that the AC program was able to maintain the same level of teacher quality at entry into teaching as traditional programs: "Although the New Jersey alternate-route teachers had lower grade point averages than their college-based counterparts, they came from more selective colleges and also scored higher on NTE tests despite being less likely to have majored in the teaching area" (Natriello & Zumwalt, 1992,p.76).

The researchers also found that the AC program contributed substantial numbers of teachers to urban schools. As with many studies of AC programs, however, other findings clouded the authors' conclusions. For example, even though the AC program helped fill positions in urban schools, over time many of these teachers left those schools, so that their placement levels (hiring and retention) ended up being similar to those of traditionally prepared teachers. In addition, a lower proportion of AC math teachers reported long-range interest in staying in teaching than did those teachers prepared in schools of education.

Another early study examined the assumption that teachers prepared through AC programs can learn on-the-job what they need to know about teaching (McDiarmid & Wilson, 1991). This research consisted of the analysis of survey data from AC and traditionally prepared teachers over the course of their first several years of teaching. In examining the extent to which teachers learned math concepts while teaching math, the authors found that AC teachers may learn some concepts, but that others are not easily learned through teaching. Moreover, results from the examination of a sub-sample of AC elementary teachers showed that these teachers possessed limited knowledge of math concepts upon entry into teaching, and their knowledge did not increase substantially by teaching it.

More Recent Studies As the number of alternative certification programs continued to increase during the 1990s, research on their structure and effectiveness became more intensive and pervasive. Today there exists a substantial body of literature on alternative certification that includes discussions of research, policy, and practice. Overall, the research to this point has been unable to clearly substantiate the effectiveness of these programs. Several studies have found positive or mixed results, but a significant portion of the research on AC programs has arrived at negative conclusions.

One recent study that compared alternatively and traditionally prepared teachers began with a review of research, and cited a substantial body of research that found either better or similar results for AC teachers when compared to traditionally prepared teachers (Miller, McKenna, & McKenna, 1998). However, the authors stated that, overall, the research on AC programs is "inconclusive and somewhat contradictory" (p.166). This is due to the fact that a variety of methodologies has been used to study AC programs, a variety of outcome variables has been examined, and a variety of operational definitions has been used to define variables.

Other problems that tend to confound studies of AC programs include inappropriate comparisons of AC teachers to non-equivalent groups from different regions or states. Comparisons of the qualifications of AC teachers to those of traditionally prepared teachers are somewhat disingenuous; teachers in AC programs may in fact, for example, have higher GPAs than traditionally certified teachers, but this is usually due to high screening standards and high ratios of applicants to openings. Studies that attempt to compare the teaching of AC teachers to that of regular teachers often employ weak or non-systematic approaches to the assessment of teacher performance, rely on state or district measures, or have extremely small sample sizes (Hawley, 1990). Reliance on district or state data to assess classroom teaching is inappropriate because these data are typically measures of student, not teacher, performance. In addition, these types of data are typically standardized, aggregated test score data that are difficult to use in assessing individual teacher behavior. Finally, studies that compare AC teachers to other, "traditionally prepared" teachers are problematic because the teachers used for comparison most likely graduated from schools of education before the recent shift toward an emphasis on standards and outcomes. Studies that conclude that AC teachers differ little from other teachers beg the question of the quality of those other teachers.

In attempting to address some of the methodological weaknesses of the AC research, Miller, McKenna, and McKenna (1998) designed a series of studies, both quantitative and qualitative, to examine the effectiveness of an alternative program. They used data from multiple sources, including observation of teaching behaviors, and the examination of student test scores and teachers' perceptions to determine if there were any differences between groups of alternatively and traditionally prepared teachers. The results showed that, after three years of experience and mentoring, the two groups were basically the same.

While this study provided solid evidence for the effectiveness of the AC program, it was unusual in several ways. First, it purposefully set out to address some of the weaknesses in the methodology that has been seen in studies of AC programs. Second, it compared the two groups after three years. Many critics of AC programs are more concerned about the impact on students of teachers with little training during the first year. The authors suggested that, in order to see similar results in other programs, the intensive three-year mentoring and support that were central to the program should be a model for other AC programs. This begs questions regarding the costs associated with such thorough mentoring, and is an example of a variable of interest in the study of AC programs that requires further research.

Data on the characteristics of AC teachers and their impact on the diversity of the teaching force are also mixed. An examination of survey data from over 14,000 teachers compared the responses of traditionally and alternatively prepared teachers (Shen, 1998). Results of this study showed that more minorities enter teaching through AC programs than through traditional ones, and teach in urban schools with high percentages of students of color. Results of this study also indicated that alternative programs are effective at recruiting math and science teachers who have degrees in those fields. On the negative side, the researcher found that while AC programs recruit and prepare minority teachers with high education levels they also bring in minority teachers without college degrees.

Studies of teacher performance are no more conclusive than the research discussed above. A review of the literature on the performance of alternatively certified teachers in comparison to graduates of teacher preparation programs found several studies that supported the idea that regularly certified teachers outperformed those from AC programs (Jelmberg, 1996). This review also found studies that provided evidence for the quality of AC teachers, and still others that had mixed results. In an attempt to sort out the issue, the author analyzed survey data from 136 recently certified teachers and ratings of performance by those teachers' principals. He found no differences between AC and traditionally prepared teachers in terms of academic credentials. The author did find that, even after three years of teaching experience, the overall ratings of teacher performance were higher for the graduates of traditional teacher preparation programs than for the alternatively certified teachers.

In terms of student outcomes, the effect of teacher ability on student performance has received a great deal of attention in recent years. Several studies have documented the relationship between teacher ability and student achievement (Ehrenberg & Brewer, 1995; Ferguson, 1998; National Commission on Teaching and America's Future, 1996). One study that researched the link between licensure and student outcomes examined data from the National Educational Longitudinal Study of 1988. Researchers employed statistical regression models to determine the effects on students of teachers with full, emergency, or out-of-subject certification. Results showed that students of teachers who are not certified or who hold private school certificates do less well than students whose teachers hold standard, probationary, or emergency certification (Goldhaber & Brewer, 2000). The fact that students of teachers with probationary or emergency licenses did no worse than students of teachers with standard certification provided support for the position that alternatively prepared teachers can be equal in quality to traditionally prepared teachers. In their conclusion, the authors state, "Although certification is pervasive, there is little rigorous evidence that it is systematically related to student achievement" (p. 141). It should be noted that this somewhat controversial finding has been called into question through a critique of the study's methodology (Darling-Hammond, Berry, & Thoreson, 2001).

Perhaps the most vocal critic of alternative certification has been Linda Darling-Hammond and her colleagues (Darling-Hammond, 1994; Darling-Hammond, Wise, & Kline, 1999; Darling-Hammond, 2000). Darling-Hammond has been a staunch supporter of the need for sound and thorough preparation of teachers that includes rigorous academic course work at the undergraduate level, highly structured and supervised internships, full licensure before a teacher candidate is given control of a classroom, and ongoing professional development.

Many critics of AC programs would contend that assigning students to teachers who have entered through shortened preparation programs is as inconceivable as entrusting them to the care of a doctor who has had six months of intensive course work and a mentor in the next room.

A review of one of the most well-known alternative programs, Teach for America (TFA), was scathing in its criticism (Darling-Hammond, 1994). The author cited several anecdotes of TFA recruits entering classrooms entirely unprepared to teach on their own, with substantial negative consequences for both the teacher and the students. She went on to criticize the costs, training, and quality control of the project, and argues that, far from being equal to teacher education programs, the TFA does not come close to adequately preparing participants to teach.

Quantitative research studies of TFA are not nearly as pessimistic. One study conducted by Kane, Parsons, & Associates examined the ratings of TFA teachers by the principals who supervised them. TFA teachers were rated as "good" or "excellent" on 23 indicators of successful teaching by over 90% of the principals surveyed (Kopp, 2000). Another study, conducted by CREDO (formerly the Center for Research on Education Outcomes), looked at the effects of TFA teachers in the Houston school system on student achievement as compared to other teachers. The results of this research suggested that students of TFA teachers achieved at slightly higher (although not statistically significantly higher) levels than the students of other teachers (Raymond, Fletcher, & Luque, 2001). One factor that was not addressed in this study was attrition. While the TFA program is specifically designed to recruit new teachers to teach for two years, analysis of attrition rates prior to completion of the two-year period would be informative. This issue is another area in which research on AC programs might be strengthened.

In a broader review of the general research on teacher preparation and its relationship to teaching and learning, Darling-Hammond, Wise, and Kline (1999) cited substantial research linking preparation in education and subject matter with a number of school-related outcomes including ratings by students, effectiveness of instruction in specific fields such as math and science, ratings of instructional effectiveness at the elementary level, and students' reading achievement. The authors also reviewed several studies that found that teachers from alternative programs have problems with curriculum development, pedagogical knowledge, classroom management, and attention to learning styles.

Summary So what can we learn from the divergent findings reviewed here? Some studies find positive results from alternative certification programs, some find negative results, and others find some positive and some negative. Positive results reviewed above included these findings:

Careful screening of applicants can contribute to intern quality.
AC programs can increase the number of minority teachers and increase the number of teachers in shortage areas.
The classroom performance and student outcomes of AC teachers can be similar to that of traditionally certified teachers in some programs.
Intensive mentoring and support can contribute to the development of alternatively certified teachers.

Some problems with AC programs have included the following:

No increase in the retention rate of alternatively certified teachers in comparison to traditional certification.
The inability of AC teachers to learn content knowledge "on-the-job."
Some programs that have allowed completely unqualified people to assume total responsibility for classrooms.
AC programs with inadequate attention to curriculum development, pedagogical knowledge, and classroom management.

To be sure, the issue of alternative certification is a political one. Like many controversial issues, ideologies and interests drive much of the debate—and subsequently the research questions and the ways in which they are studied (Haberman,2001; Darling-Hammond,Berry, & Thoreson, 2001). Those who have lost faith in the ability of higher education programs to adequately prepare teacher candidates (Haberman, 1991) look for evidence that supports alternative certification, while those with an interest in maintaining the traditional (but strengthened) model of teacher education (Darling-Hammond, Wise, & Klein, 1999) look for deficits in the alternative model. The only things that appear to be sure are that the number and size of alternative certification programs are growing, and that they will probably be with us for a while. It would seem that the best course of action is to learn from the research about the characteristics of successful and unsuccessful alternative programs, so that those that do arise or continue can be as effective as possible.

In addition, the evolution of AC programs has the potential to significantly affect teacher preparation programs based in colleges of education. Just as the charter school movement has informed improvement efforts in K-12 schools, much of the research on and knowledge about alternative routes that has been generated thus far provides lessons for college-based programs seeking to strengthen their approaches to teacher education. Research that explores the impact of mentoring, support, and retention on AC teachers can also inform college-based preparation programs. Similarly, new approaches to preparation in the areas of classroom management, subject/content, pedagogy, and child development have the potential to improve undergraduate instruction on these topics. As often happens with innovative approaches to traditional methods, the new techniques and lessons learned from AC programs have begun to influence college-based preparation programs. In fact, some schools of education have added alternative-route programs to their more traditional four-year programs.