Updated July 8, 2011
Research shows that good teaching matters, but considerable debate continues over which policies and practices will help promote high-quality teaching.
In general, the preponderance of evidence concludes that effective teachers are capable of inspiring significantly greater learning gains in their students when compared with their weaker colleagues. Most of this evidence is based on “value added” analyses of large sets of data linking individual students’ test scores to their teachers.
Such studies determine students' average annual rates of improvement, as measured by test scores. They estimate how much value a teacher has contributed to student achievement, factoring in the gains the student was expected to make based on past performance, and in some cases, controlling for elements such as peer characteristics and background, including poverty level and family education.
Factors such as family background continue to predict a majority of the variation in student achievement, but scholars generally agree that teacher quality is probably the most important school-based factor affecting achievement. Specific estimates are difficult to arrive at because economists have been unable to link a portion of the variation in student achievement to any particular input (Sawchuk, 2011).
Back in 1998, several economists estimated that at least 7.5 percent of the variation in student achievement resulted directly from teacher quality and noted that the actual number could be as high as 20 percent (Hanushek, et al, 1998). A 1999 paper, meanwhile, puts all in-school factors, including school, teacher, and class level, at approximately 21 percent of the variation in 10th grade mathematics achievement. It further estimated that 8.5 percent was directly the result of teacher effectiveness (Goldhaber, et al, 1999).
Even so, the specific characteristics that constitute an effective teacher are hotly debated, in large part because teacher quality is extremely difficult to measure. As a result, most studies resort to measurable proxies, such as certification, academic degrees, and years of experience. Most of these characteristics bear some relationship to student scores, but on the whole, they explain only a fraction of teacher quality—perhaps as little as 3 percent of the overall variation in students’ test scores (Rivkin, et al, 2005; Goldhaber, et al, 1999).
Of the measurable characteristics isolated for study, teaching experience has consistently been linked to student scores. On average, beginning teachers produce smaller learning gains in their students compared with more seasoned teachers. Most of the studies show that teachers grow in effectiveness over at least the first five years on the job, though the benefits of experience are less clear after that point (Nye, et al, 2004; Clotfelter, et al, March 2007, October 2007; Harris and Sass 2007).
Deep content-area knowledge is also an attribute of teachers that seems to have a positive impact on student achievement. This appears especially true for mathematics teachers. A variety of studies have found that factors such as math-licensure test scores, math certification, a math undergraduate or graduate degree, and math-focused professional development for secondary educators bear a relationship to student scores (Hill, et al, 2005; Harris and Sass, 2007; Goldhaber and Brewer 1999; Clotfelter, et al, March 2007, October 2007).
In general, most teacher-quality issues, including preparation, certification, tenure, evaluation, and licensing, continue to be the provenance of states and districts. The first major federal foray into teacher-quality standards came with the passage of the federal No Child Left Behind Act of 2001. The law requires every teacher of a core academic subject defined in the law to be “highly qualified.”
To meet that designation, a teacher must be certified and have demonstrated proficiency in his or her subject matter by having majored in the subject in college, passing a subject-knowledge test, obtaining advanced certification in the subject, or using an alternate, state-determined method. The highly-qualified teacher rules have generally been criticized for having few effects overall on teacher practices (Keller, 2007).
At the time of this writing, education policy was beginning to undergo a sizeable shift in thinking about teacher quality. Through federal competitions, such as the $4.35 million Race to the Top program, and state legislation, policies have increasingly shifted away from investing in credentials and other “input”-based measures, toward policies designed to build teachers’ skill levels through observations linked to teaching standards.
Teacher evaluations, the thinking goes, can serve as a method of both identifying high- and low-performing teachers and making professional development more useful by identifying the specific areas in which teaches need help.
Early evidence suggests that teacher observations can lead to improvements in teaching quality, as measured by students’ scores. Observations, and especially conversations that result from them, can help teachers hone their craft and have been integrated as a key component of many schoolreform efforts. And building the professional capacity of educators was cited as one of five tenets that characterized school improvement in a 2010 report by the Consortium for Chicago School Research (Sawchuk, April 26, 2011; April 1, 2009; CCSR, 2010).
States that won grants in the Race to the Top competition pledged to make improvements to teacher evaluation and use the results to make decisions about teacher professional development, promotion, tenure, and compensation. Winning states are still in the beginning phases of doing so, and few extant examples on which to build currently exist. (Sawchuk, Dec. 16, 2009).
Among the most contentious issues, still under debate at the time of this writing, is the use of test scores as one of several measures of performance. It has been criticized by some teachers’ unions and by some researchers, but embraced by others (Economic Policy Institute, 2010; Brookings Institution, 2010).