Linda Benedict, Schafer, Mark J.
Don Asay and Mark J. Schafer
Louisiana Gov. Bobby Jindal signed into law Act 54 or House Bill 1033, which will implement value-added teacher assessment (VATA) in Louisiana during the 2012/2013 school year. Value-added teacher assessment is a relatively new approach to assessing teacher effectiveness. Most current systems use reviews by the principal and peers, even though typically neither principals nor peers observe much of the actual classroom instruction.
Some existing methods of teacher assessment use student test scores to evaluate teachers. In Texas and other places, for example, teachers receive grades based on how many of their students can meet minimum standards established with the No Child Left Behind (NCLB) Act of 2001. Jennifer Booher-Jennings refers to this system as educational triage because there is a strong incentive for teachers to focus on the children whose previous year’s test scores were close to the minimum standards. High-performing students will pass anyway and low-performing students have a low probability of passing.
Value-added methods change the incentive structure so that teachers get credit for all their students’ learning gains. William Sanders, a pioneer of this approach, argues that this method isolates the effects of teachers because it holds all other factors associated with the students as constant. Thus, a student’s race, socioeconomic background, home life, community and even their previous educational experiences and outcomes are all controlled for because each year’s tests are measured in comparison to each individual student’s scores on the previous year’s tests. In theory, all students learn in classes taught by effective teachers, irrespective of how much they had learned prior to that school year. By measuring individual students’ learning gains relative to previous year’s scores, all other factors wash out, and the only remaining factor is the effectiveness of the current year’s teacher.
VATA has proponents and detractors. The general approach has been implemented differently among states. Louisiana is in the process of implementing value-added teacher assessment, and debates about the system in Louisiana mirror debates surrounding the implementation of value-added teacher assessment across the nation.
History of Value-Added Teacher Assessment
In the 1970s, education economist Eric Hanushek introduced the VATA idea, which is that good, effective teachers add more value than average and ineffective teachers because their students learn more. Moreover, Hanushek argued that it should be possible to judge the effectiveness of teachers by measuring the learning gains of their students. William Sanders, an educational statistician, was the first to put Hanushek’s ideas into practice when he introduced the first value-added assessment models in Tennessee in 1992, the Tennessee Value-Added Assessment System (TVAAS). The TVAAS model defines effective teachers as those who produce increases in their students’ learning.
Two assumptions drive the TVAAS and other value-added models: (1) Student learning can be accurately measured through synchronized annual tests and (2) teachers matter. The first assumption that student learning can be effectively measured through synchronized annual tests has been widely accepted, but remains controversial. The NCLB signed into law by President George W. Bush in 2001 put mechanisms in place at the state level to measure student performance with the goal of each school reaching a level of adequate yearly progress. To ensure that students were making adequate yearly progress, each state was tasked with collecting measures of student performance through standardized tests. The NCLB requirement to test students produced quality data that can be used to answer a variety of questions about student learning at several levels of analysis.
As for the second assumption, debate continues among educators and social scientists over the extent to which schools and teachers matter when it comes to student learning and educational outcomes. Sociologists such as James Coleman, Christopher Jencks and others produced a large volume of research findings from 1970 to 2000 that identified factors outside of the school – including poverty, family and neighborhood characteristics – as the main causes of educational inequality. While sociological research downplayed the effects of schools and teachers, Hanushek, along with Harvard educational researcher Richard Murname, conducted a series of studies that showed large and consistent differences in student learning within schools, identifying effective and ineffective teaching as the primary cause of these intraschool differences. VATA models attempt to measure and assess teacher effectiveness broadly, both within and across schools, by measuring student learning in the target year against the previous year’s test scores, thus holding all other factors constant except for the current year’s class.
With NCLB, value-added assessment gained in popularity, and both researchers and educators recognized it as a new opportunity to ensure that all students get effective teachers. By 2010, some version of value-added assessment was being used – often as only one part of a broader teacher assessment effort.
Arguments in Favor of VATA
Proponents of value-added assessment point to the following primary advantages of this system of teacher assessment. First, proponents argue that value-added modeling isolates the effects of teachers from various other factors that could affect student learning, such as student intelligence, family socio-economic level, extent of parental involvement and neighborhood characteristics – all of which are beyond the teacher’s and school’s control. Because students’ actual scores are measured against expected scores based on previous years’ tests, the effects of nonschool factors are effectively controlled for in statistical analysis of student learning. Second, the value-added method provides strong incentives for teachers to teach to all students irrespective of variation in students’ individual-level abilities. Teachers will benefit just as much from learning increases in low-ability students as they will from high-ability students because students’ expected test scores are predicated upon the same students’ previous year’s scores. A learning trajectory will be established, and teachers will attempt to have each student meet or exceed their individually determined expected trajectory. The model encourages teachers to have high expectations for all students, including those who already meet the minimum standards set by NCLB. Finally, value-added modeling is focused more on individual-level learning as opposed to group-level learning (i.e., improvements in the class average or the percentage of proficient students) and, therefore, more closely meets the spirit of the NCLB act. By contrast, some group-based measures of school and teacher effectiveness reward schools when a relatively small proportion of students show improvement, even though the majority of students do not.
Arguments Against VATA
Critics argue that VATA cannot accomplish its objective of distinguishing effective from ineffective teachers. First, VATA claims to isolate the effects of teachers by controlling for external factors by measuring actual student learning against an estimated amount of learning that takes into consideration each student’s educational history. Education policy researchers, including Haggai Kupermintz and Gerald Bracey, argue that this approach is insufficient because it does not deal with well-known selection biases – the fact that some schools attract more high-ability students with more supportive families and more positive neighborhood contexts than other schools.
Second, another aspect of the VATA approach is the establishment of learning trajectories for each student that allow for individualized projected or estimated learning. Even VATA supporters recognize that it is more difficult to establish learning trajectories and estimate how much they should learn for students with erratic experiences in the first years of schooling, such as multiple transfers or a mixture of good and bad teachers. Part of the problem is that the positive effects of good teachers or the negative effects of bad teachers are, in theory, supposed to have long-term effects that can last several years. But incorporating the long-term effects of good and bad teachers into the model that determines how much each student should learn in a given year is fraught with complexities.
Finally, some critics suggest that value-added assessment might work in theory, but it cannot deal with the practical reality of teaching in most public schools. For example, value-added assessment assumes one teacher per classroom, so it should theoretically work best in the lower grades where many students are being taught by one teacher throughout the day. But it is more difficult to apply in the higher grades when students switch from teacher to teacher throughout the day. On the other hand, value-added assessment does not work well in the lower grades because the variability in student test scores is much higher in the early years. Therefore, it is more difficult to establish a trajectory and expected test scores for the next year. Education researcher Richard Rothstein asks: “If a fifth-grade teacher gets students who were taught the previous year by an outstanding fourth-grade teacher, does this make her job easier because the fourth-grade teacher set the students on a fast-paced learning trajectory that will last several years? Or does this make her job more difficult because her students’ predicted outcomes will have been set on exceptional learning in fourth grade?” Rothstein argues that no one knows the answer.
Despite these criticisms, a growing practical consensus holds that VATA marks an improvement over the educational triage system of evaluating teachers based on how many students clear the minimum standard bar. But it cannot single-handedly separate the good from the bad teachers and, therefore, should be used in conjunction with other teacher evaluation methods. Ideally, VATA would be able to identify problem areas where a large percentage of students are not learning as much as they would be expected to learn, and then motivate further investigations to find out why.
VATA in Louisiana
Louisiana’s value-added teacher assessment methodology takes into consideration many of the pros and cons of value-added teacher assessment. Louisiana’s system was developed by George Noel, professor of psychology at Louisiana State University. It uses a combination of each student’s individual achievement history, along with other key factors, to first predict achievement in the target year and then measure actual achievement against the predicted level. Recognizing the need to establish learning trajectories, it begins with the fourth grade. It also excludes students for whom no prior achievement data is available, as well as those who move during the school year. The key factors include many variables that sociologists would consider in predicting achievement: gifted status, family income, disability status and discipline history, all of which are unequally distributed across the spectrum of public schools in the state. Teachers are assessed based on the average achievement of all the included students whom they taught during the school year.
Louisiana’s two major teachers unions, the Louisiana Federation of Teachers and the Louisiana Association of Educators, have criticized the new assessment method, preferring a mentoring program known as the Teacher Advancement Program (TAP). TAP also claims to add value to education not by using standardized tests in an incentive system for rewarding or punishing teachers based on their students’ test performance, but by using student results to help teachers identify areas of strengths and weaknesses, and develop methods for improving their craft.
A number of studies have shown a growing discontent among teachers everywhere about being evaluated solely on the basis of their students’ performance on standardized tests. VATA approaches generally exacerbate this problem, but over time the approach could be integrated into mentoring approaches, like TAP, and that might make them more palatable to teachers. However, teachers also want more recognition that they do a lot more than teach to the test. As Rothstein points out, teachers also help students develop reasoning skills and abilities to cooperate with others, which may have even more long-term benefits to individuals and society than improvement on standardized tests.
Don Asay is a graduate student in the Department of Sociology, and Mark J. Schafer is an associate professor in theDepartment of Agricultural Economics & Agribusiness.